Ibm tivoli storage manager v5.4 implementation and administration part two tm526inst

IBM Tivoli Storage Manager 5.4 Implementation and Administration Part TwoInstructor’s Training GuideS150-2607-00

May 2007

Copyright NoticeCopyright © 2007 IBM Corporation, including this documentation and all software. All rights reserved. May only be used pursuant to a Tivoli Systems Software License Agreement, an IBM Soft-ware License Agreement, or Addendum for Tivoli Products to IBM Customer or License Agreement. No part of this publication may be reproduced, transmitted, transcribed, stored in a retrieval system, or translated into any computer language, in any form or by any means, electronic, mechanical, magnetic, optical, chemical, manual, or otherwise, without prior written permission of IBM Corpora-tion. IBM Corporation grants you limited permission to make hardcopy or other reproductions of any machine-readable documentation for your own use, provided that each such reproduction shall carry the IBM Corporation copyright notice. No other rights under copyright are granted without prior writ-ten permission of IBM Corporation. The document is not intended for production and is furnished “as is” without warranty of any kind. All warranties on this document are hereby disclaimed, including the warranties of merchantability and fitness for a particular purpose.Note to U.S. Government Users—Documentation related to restricted rights—Use, duplication or disclosure is subject to restrictions set forth in GSA ADP Schedule Contract with IBM Corporation.

TrademarksThe following are trademarks of IBM Corporation or Tivoli Systems Inc.: IBM, Tivoli, AIX, Cross-Site, NetView, OS/2, Planet Tivoli, RS/6000, Tivoli Certified, Tivoli Enterprise, Tivoli Ready, TME. In Den-mark, Tivoli is a trademark licensed from Kjøbenhavns Sommer - Tivoli A/S.Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the United States, other countries, or both.UNIX is a registered trademark of The Open Group in the United States and other countries.C-bus is a trademark of Corollary, Inc. in the United States, other countries, or both.Java and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other countries, or both.Lotus is a registered trademark of Lotus Development Corporation.PC Direct is a trademark of Ziff Communications Company in the United States, other countries, or both and is used by IBM Corporation under license.ActionMedia, LANDesk, MMX, Pentium, and ProShare are trademarks of Intel Corporation in the United States, other countries, or both.SET and the SET Logo are trademarks owned by SET Secure Electronic Transaction LLC. For fur-ther information, see http://www.setco.org/aboutmark.html.Other company, product, and service names may be trademarks or service marks of others.

NoticesReferences in this publication to Tivoli Systems or IBM products, programs, or services do not imply that they will be available in all countries in which Tivoli Systems or IBM operates. Any reference to these products, programs, or services is not intended to imply that only Tivoli Systems or IBM prod-ucts, programs, or services can be used. Subject to valid intellectual property or other legally pro-tectable right of Tivoli Systems or IBM, any functionally equivalent product, program, or service can be used instead of the referenced product, program, or service. The evaluation and verification of operation in conjunction with other products, except those expressly designated by Tivoli Systems or IBM, are the responsibility of the user. Tivoli Systems or IBM may have patents or pending patent applications covering subject matter in this document. The furnishing of this document does not give you any license to these patents. You can send license inquiries, in writing, to the IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, New York 10504-1785, U.S.A.Printed in Ireland.

• • • ••

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two i

Course Revision History

Introduction

The IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two course is a replacement for the IBM Tivoli Storage Manager Administration 5.3.2 course. Through information exchange there was a consistent opinion that the students needed a representative course for the added administration functionality that would be a natural follow-on to the IBM Tivoli Storage Manager 5.4 Implementation and Administration Part One course. The course design builds upon the previously learned concepts and adds to this knowledge base. This course is designed for users who are already experienced in administering Tivoli Storage Manager, or who have extensive knowledge of the product.

This course combines aspects of previous courses and the concepts from the new release. The course design returns to a flow where the student will use information from one unit to the next to learn and apply concepts and features for how they manage and administer their own Tivoli Storage Manager implementation.

About the IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two course:

• IBM Tivoli Storage Manager 5.3.2 Administration, released December 2005, was based on IBM Tivoli Storage Manager 5.3.2 code. This update was due to the significant changes to the Administration Center’s interface.

• IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two, released May 2007, is based on IBM Tivoli Storage Manager 5.4 code. This update explains the new data shredding feature and the new active-data storage pool configuration.

General Revisions of the May 2007 Version

Student’s Training Guide

• The instructor fix requests received from the GEO Delivery Managers were evaluated and applied where appropriate.

ii IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface

Course materials may not be reproduced in whole or in part without the prior written permission of IBM.

• Most of the material related to the introduction of the Administration Center has been removed.

• Diagrams that reference the Tivoli Storage Manager hierarchy have been changed to reflect a two-tiered hierarchy rather than a three-tiered hierarchy.

• Introduction of the new features in Unit 1 has been placed into slides rather than text.

• Unit 12 from the IBM Tivoli Storage Manager 5.3.2 Administration course has been moved to the Appendix A of the IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two course, and the exercise has been removed.

• Exercise 1 for Unit 9 has been removed.

Course Materials

The following table contains the updated part numbers for the May 2007 course materials.

Detailed Revisions of May 2007 Release

1. The modules for the IBM Tivoli Storage Manager 5.4 Implementation and Administration are listed below.

Unit 1 – Tivoli Storage Manager 5.4 Concept Review and New Features

IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two Part Number

Student’s Training Guide S150-2605-00

Student Exercises for Windows S150-2604-00

Student Exercises for AIX S150-2606-00

Instructor’s Training Guide S150-2607-00

Instructor Resources CD SK5T-7805-00

Student Resources CD SK5T-7804-00

Classroom Setup CD SK5T-7806-00

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two iii

• • •••

Preface


Unit 2 – Database and Log Manipulations

Unit 3 – Storage Pool Manipulations

Unit 4 – Database and Storage Pool Problem Determinations

Unit 5 – Macro and Scripts

Unit 6 – Network Configuration and Tivoli Storage Manager Performance Issues

Unit 7 – Reporting

Unit 8 – Problem Determination Concepts

Unit 9 – Backup and Restore Methodology

Unit 10 – Enterprise Administration

Unit 11 – Server-to-server Virtual Volumes

Appendix A - Related Tivoli Storage Manager Products

2. The Administration Center was new in the last release, and is no longer considered a new feature. Much of the material introducing this interface was deleted, but it is still referenced in several units of the course.

3. The administrative interfaces were described in Unit 2 of the IBM Tivoli Storage Manager Administration 5.3.2 course. This material was moved into Unit 1 of this course, and is considered to be a review of basic concepts.

4. Unit 2 of this course covers Tivoli Storage Manager database and log manipulations, and Tivoli Storage Manager storage pool manipulations has been moved to Unit 3.

5. Unit 3 describes the new data shredding, and the new active-data storage pool features.

6. A two-tiered representation of the Tivoli Storage Manager storage hierarchy is closer to the most common usage. All graphics representing the three-tiered hierarchy have been changed to reflect a two-tiered hierarchy.

7. The IBM Tivoli Storage Manager 5.3.2 Administration course offered an exercise for Lan-free backups in the Student Exercises for Windows guide. That exercise has been removed, but the information is still covered in the IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two course.

8. Unit 12 of the IBM Tivoli Storage Manager 5.3.2 Administration course discussed products related to IBM Tivoli Storage Manager. That unit has been deleted from this course, and the information has been moved to the Appendix A. The exercise for this unit relating to Tivoli Data Protection for Mail in the Student’s Exercise

iv IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface


Guide for Windows has been removed, but the information is still covered in the course. Also, some of the information for IBM Tivoli Storage Manager Space Management and IBM Storage Manager HSM for Windows has been removed, but the basic information is still covered. More information on this feature of the IBM Tivoli Storage Manager Extended Edition will soon be available in a separate IBM Tivoli Storage Manager course.

• • • ••

v

Instructor Preparation

Introduction

This section is a compilation of notes by the course developer to help instructors deliver IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two. It gives an overview of the course, instructions for setting up the classroom, and other information that instructors should have to make the class a success.

This course was developed to address the needs of Tivoli Storage Manager administrators requiring a deeper knowledge of some of the areas of Tivoli Storage Manager. The prerequisites for this course state the student should have taken the Implementation course or spent adequate time with the product to become familiar with the administrative utilities and the base concepts. There will be times when these prerequisites are not met and as an instructor you must be prepared for this occurrence.

Course Components

The following materials are needed to deliver the training.

For the instructor:

• Instructor’s Training Guide

• Student Exercises for Windows

• Student Exercises for AIX

• Classroom Setup CD

• Instructor Resources CD

• Student Resources CD

For each student:

• Student’s Training Guide

• Student Exercises for Windows

vi IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



• Student Exercises for AIX

• Student Resources CD

Instruction Preparation Requirements

Instructors should perform the following steps prior to delivering this course.

• Read the instructor preparation notes.

• Read the entire Student’s Training Guide to get familiarized with the content.

• Using the Classroom Setup Guides, set up a system for each platform on which the course will be delivered.

• Perform the exercises for each platform.

• Obtain the certification IBM Certified Deployment Professional - Tivoli Storage Manager. Latest information can be found at www.ibm.com/certify/.

Classroom Setup Instructions

To set up the classroom follow the appropriate Classroom Setup CD. The instructions are written to set up each system individually.

The Classroom Setup CD contains guides that have been written in detail to help those who are required to perform setup tasks and do not have the knowledge base of the products covered in this course material. Setup personnel can use IBM Tivoli resources on the Internet and product documentation to help with setup.

The Classroom Setup CD contains files needed for setup. These include files used in backups and archives, software installation files and operating system updates. The respective Classroom Setup Guide will inform you of software or files when needed to perform tasks. There are macros that have been created to aid in the correct setup of the Tivoli Storage Manager server and to save time in the process. Do not change these macros unless you understand the consequences and make sure they do not impact the system setup.

An image could be created for the initial system state and then installed onto each system for course use. This is a decision for each site presenting the course. There are a few settings that must be changed after an image has been installed. These settings are also detailed in the Classroom Setup Guides.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two vii

• • •••



The instructor should perform the setup of a system following the Classroom Setup Guide for each of the platforms they will present. This will provide a knowledge base for the way the labs will operate and help with any situations that may occur over the duration of the class.

Timing Guidelines

Reasonable target times for pacing the course are shown below. The times shown will differ with each presentation of the course. These times should help you to prepare for the course and understand the schedule that must be kept to complete the course.

Day 1

Unit Target Time

Introduction 9:00

Unit 1–Lecture 9:45

Break 10:15–10:25

Unit 1–Lecture continued 10:25

Unit 1–Exercises 10:55

Lunch 11:30–12:30

Unit 1–Exercises continued 12:30


Break 14:40–14:50


Unit 2–All completeEnd of Day 1 17:00

Day 2

Unit Target Time

Unit 3– Lecture 9:00

Break 10:15–10:25

viii IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



Unit 3 Exercises 10:25

Lunch 11:30–12:30

Unit 4 12:30

Unit 4 PD–Exercise #1 12:45


Break 14:45–14:55



Unit 5–Lecture CompleteEnd of Day 2 16:55

Day 3

Unit Target Time

Unit 5–Exercise 9:00

Break 10:25–10:35


Unit 6–Exercise 1 11:10

Lunch 11:30–12:30

Unit 6–Exercise 1 continued 12:30


Break 14:15–14:25




Unit 7–Lecture CompleteEnd of Day 3 16:50

Day 2

Unit Target Time

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ix

• • •••



Day 4

Unit Target Time


Break 10:10–10:20



Lunch 11:35–12:35

Unit 8–Exercises continued 13:05

Unit 9–Lecture and Exercises 13:10

Break 14:15–14:25

Unit 9–Lecture and Exercises continued 14:25

Unit 9–All CompleteEnd of Day 4 17:15

Day 5

Unit Target Time


Break 10:00–10:10



Lunch 11:30–12:30




Appendix A 14:40

Appendix A–All CompleteEnd of Day 5 15:30

x IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



Suggestions for Delivery

These notes provide an overview and insight into the Student’s Training Guide unit by unit. General suggestions for delivery of the course are also included.

The students will begin by using a preconfigured Tivoli Storage Manager server named Server1. This server will be utilized until the exercises for Unit 10. In the Unit 10 exercises Server1 will be halted and three Tivoli Storage Manager servers named Black, Cyan, and Grey will be used for enterprise administration. Unit 11 will call for the previous three servers to be halted and then two Tivoli Storage Manager servers named Source and Target will be used for virtual volumes. If the systems use an image then the accept date and enable session commands may need to be entered. This is noted in the Student’s Training Guide. In this version the servers must be halted and started again if these steps are required and this is also stated in the Student’s Training Guide.

The Classroom Setup CD has a folder of shortcuts that can be copied to the desktop and then arranged. These shortcuts provide icons to make the students access to the various server and clients easier. On AIX systems there are shell scripts provided to start all the needed aixterm windows for use with Unit 10 and Unit 11, Enterprise Administration and Server-to-server Virtual Volumes, respectively. Please reference the respective Classroom Setup Guide for more information on using these shortcuts and scripts.


This unit begins as an overview of features included with Tivoli Storage Manager 5.4 Extended Edition. The unit continues with a discussion of the new features available in Tivoli Storage Manager 5.4 Extended Edition. The instructor should take time to become familiar with these features and mention the ones that will have the most impact in a production environment. Two new features, data shredding and active-data storage pools, are covered more thoroughly in another unit.

Many students may have not upgraded from Tivoli Storage Manager version 5.2. Point out that they can migrate to Tivoli Storage Manager version 5.4. Also mention that there is an upgrade path from Tivoli Storage Manager Express to Tivoli Storage Manager 5.4 Extended Edition.

The rest of the unit is designed to be a quick review of the major components of Tivoli Storage Manager. This has been included for those students who may need a thorough overview. Some students who have experience with Tivoli Storage Manager may need to fill gaps in their knowledge. Many newer administrators are not aware and may have never used some of the options available from the command line. Some of these options are introduced in this unit, and are used later in the course. The instructor should not spend too much time here as it will make it difficult to cover the rest of the units and labs appropriately.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two xi

• • •••



The delivery of this unit will differ from class to class. It is very important for the instructor to determine the level of ability of the students and then spend the appropriate review time. This module needs to have moderation or timing will become an issue.

Unit 2 – Database and Log Manipulation

This unit will familiarize students with the aspects of the Tivoli Storage Manager database and recovery log. Many students are only familiar with features implemented in their environment. This information should be familiar to most students so it can be covered more quickly. Use your knowledge of the students’ levels to determine the amount of time to spend on the concepts. This unit is important to the student’s understanding and completion of the Problem Determination labs. The instructor should make sure students have a thorough understanding of the materials.

Unit 3 – Storage Pool Manipulation

Unit 3 will familiarize students with the aspects of the Tivoli Storage Manager storage pools. In prior courses, students were struggling with the exercise pertaining to migration, so some new slides have been added to further describe this feature.

Two new features, data shredding and active-data storage pools are introduced in this unit. Take time to explain these features thoroughly, as this is the first exposure the students have had to these features.

As in the unit for database and log manipulation, this unit is important to the student’s understanding and completion of the Problem Determination labs. The instructor should make sure students have a thorough understanding of the materials.

Unit 4 – Database and Storage Pool Problem Determination

The Problem Determination labs are critical to the students being able to manually recover the Tivoli Storage Manager server. These labs will allow the students to possibly perform their first Tivoli Storage Manager server recovery. There are three exercises to reinforce the ability of Tivoli Storage Manager to be recovered if the proper steps are taken to allow this process occur. It is imperative the exercises in Unit 3 have been completed successfully or there may be problems in these labs. Be prepared to help students if they have an issue with the server. The slide show for this unit enables the instructor to walk through the labs with the students. The slide show uses the same steps as the exercise guide.

xii IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



Unit 5 – Macros and Scripts

Macros and scripts are essential elements of an efficient Tivoli Storage Manager 5.4 implementation. Some students may not have any experience with scripting. This module has been included to give them the skills necessary to properly create a script. Although there are some good script examples in the text, the instructor should consider giving the students more script and macro examples and ideas to add value to this unit.


Proper performance tuning is essential to an effective Tivoli Storage Manager implementation. This unit has been expanded from previous courseware versions in order to give students a thorough understanding of performance issues and tuning parameters. The instructor should spend an appropriate amount of time in this unit. The instructor should read the Tivoli Storage Manager Performance Tuning Redbook as preparation for delivery.


This unit provides information on Tivoli Storage Manager Reporting. Much administration time is wasted in production environments due to poor reporting techniques. Though many students will be familiar with the Operational Reporting tool, there is usually more to learn. It is important that the instructor provide a number of examples of implementing SQL statements and queries in the Operational Reporting tool that will be of interest to students. It is important that the instructor have a good understanding SQL statements.


Unit 8 will aid students in the determination of problems they may encounter in a production environment. This unit will give students the skills and understanding necessary to perform traces. The instructor should be familiar with the Problem Determination Guide available on the Tivoli Web site and included on the Instructor Resources CD. Student knowledge and skill level will determine the way this unit is delivered. Expanding the discussions in this unit may not prove helpful for many students. A majority of students need a basic understanding of troubleshooting and a knowledge of how to implement tracing. They will benefit most from discussions that help them properly utilize Tivoli support to handle difficulties beyond their scope of knowledge.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two xiii

• • •••




This unit introduces advanced backup tools. These tools can also be used to address performance issues covered in Unit 6. Some of these tools can be handled lightly in order to make up time or make time for coming units. One example could be the lesson on Adaptive Subfile Backup used for the backup of mobile clients. There are only a few production environments that are candidates for this feature and the exercise is marked as optional.


This unit deals with Tivoli Storage Manager administration in multiple Tivoli Storage Manager server environments. Many students have a single server environment and may not have need of enterprise administration in the near future. This information is easily understood and there is no need for expanded discussions during the lecture. It would be better to move quickly through the discussions and allow the students to clarify their understanding while doing the exercise with the instructor assistance if needed.


This unit, like Unit 10, deals with Tivoli Storage Manager administration in multiple Tivoli Storage Manager server environments. Many students have a single server environment and may not have need of virtual volumes in the near future. This information is easily understood but some expanded discussions may be necessary during the lecture to address features that can be of help to students who have multiple server environments. Exploitation of virtual volumes for use with DRM should be highlighted for disaster recovery capabilities. One example would be the use of virtual volumes for database backups. That topic is covered in the DRM class but is important to discuss in this unit for the benefit of some students. For the most part, move quickly through the discussions and allow the students to clarify their understanding while doing the lab.

Appendix A – Related Tivoli Storage Manager Products

This is the last unit and can be tailored to meet the time requirements. It is meant to be a general overview. It is not possible to cover all of these products thoroughly in this class. It is important that the instructor be familiar with these products in order to explain their implementation. There are similarities between most of the Tivoli products for particular applications. It would be helpful for the instructor to mention that these products are generally implemented by consultants or partners. Consultants can assist customers in setting up several of their data protection agents, make sure they understand the process, and then allow them to complete the installation of the rest of the data protection agents.

xiv IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



This is the best way to learn to implement these agents in a production environment. The lectures will give the students the basic skills to implement the Tivoli Storage Manager side of the data protection agents. In most cases some knowledge of the applications is also necessary to effectively implement the data protection agents. The suggested policy settings have been included in the materials because this is critical for proper backup of applications and meeting the requirements of a service level agreement.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two xv

• • •••



Suggestions for Exercises

Introduction

The exercises are available for either Windows or AIX. There are exercises available for Windows that are not available if teaching a pure AIX version of the class. This is due to the functionality only being supported in a Windows environment. If Windows systems are also available then there are instructions for using the Windows systems to perform the exercises and utilize the AIX Tivoli Storage Manager server.

The Student Exercises for Windows or the Student Exercises for AIX contains the exercises without answers in the front, followed by an appendix that contains the answers. The exercises make use of the administrative command-line utility extensively in some areas. This has been done and should be pointed out by the instructor so they can learn commands that could be used in macros and scripts. The Administration Center is used for some of the exercises and is covered extensively in the Implementation course. The instructor can indicate, if the expertise with commands is not high, that the students can reference or even use the appendix for exercise completion. If this is done then it is important to make sure the students do more than type the commands; they should know why they are performing a task. They need to feel comfortable about asking for help. The exercises were written with this as a factor.

It is important for the instructor to have completed the exercises in preparation for delivery.

For release 5.4, the exercises were tested in a single machine environment. Those exercises requiring an AIX and a Windows machine should be tested for the Tivoli Storage Manager 5.4 version.


The exercise for this unit allows the student to investigate the setup of a Tivoli Storage Manager server to see what could be done to improve the current situation. It may be helpful, according to the level of expertise of the students, to help them see a few of the items and then give them time to use the query results to come up with the other suggestions.

After the students are finished with their evaluation, the instructor should lead a discussion to review the findings. Information from this exercise will be utilized in the exercises later in this class. The instructor needs to make sure suggestions made at the end of the exercise are understood by the students.

xvi IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



This unit allows the student to use the Administration Center to navigate and perform basic functions. The command-line interface (CLI) exercise gives the student the opportunity to use options they may have not known about or ever tried.

There is no difference between Windows and AIX exercises for this unit.

Unit 2 – Database and Log Manipulation

These exercises use the suggestions for improvement found in the Unit 1 exercise. These suggestions provide a basis for the manipulations performed on the database and recovery log. The completion of these labs is essential to the problem determination exercises in Unit 4. The instructor should either individually check the complete state of each student’s system or review this with the class as a whole.


Unit 3 – Storage Pool Manipulations

These exercises also use the suggestions for improvement found in the Unit 1 exercise. These suggestions provide a basis for the manipulations performed on the storage pools with their respective volumes. The completion of these labs is essential to the problem determination exercises in Unit 4.

Some problems will arise if the students use the up arrow to repeat commands and do not change all the values correctly. This may result in volumes assigned to the wrong storage pool or placed in the wrong location. Learn to spot these anomalies as you watch the students progress through the exercises.


Unit 4 – Database and Storage Pool Problem Determination

This unit is made up of a short introduction section in the Student’s Training Guide and then the exercises. The instructor slide show for this unit provides slides that mimic the exercises. This allows the instructor to lead the class through the recoveries. There is a slide show for Windows and another for AIX systems.

The problem determination exercises are set up by executing a PERL script that will create the anomalies on the Tivoli Storage Manager server. There are three exercises and the scripts are executed by starting a batch file from the root directory of the C: drive for

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two xvii

• • •••



Windows or a shell script in the root directory on an AIX system. It is preferred to have the students leave the room and for the instructor to execute these routines so the student does not see the results as the routine runs. There has to be a drm directory created per the Classroom Setup Guides.

c:\eod1.bat – Create problems–exercise 1 Windows commandsc:\eod2.bat – Create problems–exercise 2c:\eod3.bat – Create problems–exercise 3

cd /perlscripts AIX commandsperl -w /perlscripts/eod1.pl - Create problems - exercise 1perl -w /perlscripts/eod2.pl - Create problems - exercise 1perl -w /perlscripts/eod3.pl - Create problems - exercise 1

When prompted for the password, type in iluvtsm.

When asked to confirm the override, type in OVERRIDE (case sensitive).

Unit 5 – Macros and Scripts

The macro exercises have the student typing in macros with comments so they can practice documenting these helpful command sequences. The Instructor Resources CD has a directory, macro_scripts_lab_files, on it that contains two files for macros that are lengthy. The name of the file indicates the step number where the particular file is used. There is a file with the complete macro coded and a file where the comments are included and the commands are replaced with xxxxxxxxxxxxxxxxxxxx lines. You can choose to place these files on the student systems to expedite the exercises if needed. It is preferred that the files without the commands in them be used, so the students at least have to replace these lines with the correct commands. This will still eliminate a large amount of typing.

The scripting exercises have been expanded so the students can learn different script functionality and then see how a SQL select can be used in a script to help with scheduling. This is a forward reference since reporting has not been covered and will require the instructor to explain this concept using the slides available during lecture.


This exercise brings back an exercise from previous courses where the students must determine if a current topology can support the requirements for a Tivoli Storage Manager installation based on network and hardware criteria. It is sometimes better to split the class into groups and have them work together to find a solution. You may also have representatives present their conclusions or do this as a whole according to the time available. Make sure you have studied the results and are prepared to explain the issues.

xviii IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••




A series of SQL select statements are performed as the students learn about querying the Tivoli Storage Manager server while using the help select command to investigate select statements. The Operational Reporting tool is available on Windows systems only and will not be available in an AIX only lab environment.

To perform the ODBC labs will also require a Windows system.


The first exercise shows the student the use of querying the activity log for information and the different forms of that query. The client statistic information for the last session with the server is also viewed.

The second exercise covers tracing for the server and the client. When tracing the client, the last trace is performed on a single file selectively backed up with all the trace flags turned on. Explain that the rationale for this step is to show how much data can be generated from a trace and to only perform tracing when they know the resultant amount of data or have been guided by support personnel. The file generated by this trace will too large to view with Notepad on a Windows system, have the students use Wordpad.


The exercises for this unit cover some of the advanced features for backup and restore. Some of these are open file support, image backup and adaptive subfile backup. Some of the other features used in these exercises are only available on the Windows platform and can be accomplished in a mixed environment also. If the lab environment is AIX only, then the number of exercises will be reduced.


The students have been working with the Tivoli Storage Manager server, Server1. In this set of exercises the student will halt Server1 and use the servers Black, Cyan, and Grey to learn about server-to-server communications, command routing, and other facets of enterprise administration between multiple Tivoli Storage Manager servers. As mentioned in the introduction section there are shortcuts and shell scripts the students can use to speed up their access to the various server and clients (refer to the Classroom Setup Guides for more information). The students need to close all the servers and clients before moving on to the next unit’s exercises.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two xix

• • •••




The virtual volume exercises will make use of the Tivoli Storage Manager servers Source and Target. These servers can be accessed in the same manner as those in Unit 10. These labs will solidify the concepts of virtual volume storage between two Tivoli Storage Manager servers. The students will get to see how each Tivoli Storage Manager server stores the data, virtual volume on the source and archive data on the target.

Appendix A – Related Tivoli Storage Manager Products

As this is an appendix, there are no exercises for this unit.

xx IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



• • • ••

I

Table of Contents

Preface Course Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiAudience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiCurriculum Roadmap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

Role . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiCourse Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ivCourse Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vAudience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .viPrerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .viCourse Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiTypographical Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ixProduct Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ix

Software Release Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ixProduct Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xRecommendations in this Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x

Unit 1: Tivoli Storage Manager 5.4 Concept Review and New Features

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2

Lesson 1: Tivoli Storage Manager Products . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3Tivoli Storage Manager Express . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4Tivoli Storage Manager Basic Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4Tivoli Storage Manager Extended Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4

Tivoli Storage Manager Extended Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5Enterprise Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5Virtual Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5Tivoli Disaster Recovery Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6Tivoli Storage Manager Library Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6Tivoli Storage Manager with Storage Area Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6LAN-free Backup and Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6Additional Client Backup and Restore Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6Optimizing Performance for LAN and SAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7Tivoli Storage Manager Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7Tivoli Storage Manager for Space Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7Tivoli Storage Manager for Data Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7Tivoli Storage Manager’s System Backup and Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7NDMP Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-8

Tivoli Storage Manager Extended Edition 5.4 Product Enhancements . . . . . . . . . . . . . . . . . . . 1-9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-11

Lesson 2: Migration to Tivoli Storage Manager 5.4 . . . . . . . . . . . . . . . . . . . .1-12Migrating to Tivoli Storage Manager Version 5.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-12

Storage Agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13Device Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13Device Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-14Tivoli Storage Manager Express . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-14

II IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Lesson 3: Tivoli Storage Manager Components Overview . . . . . . . . . . . . . .1-15Database and Recovery Log Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-17Server Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-17Recovery Log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-18Server Options Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-19Communication Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-20

Volume History File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-20VOLUMEHISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-20Device Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-21DEVCONFIG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-21

Policy Management Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-22Tivoli Storage Manager Storage Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-24Daily Flow of Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-25

Lesson 4: Server Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-26Administrative Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-26Administrator Privilege Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-26

System Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-26Operator Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-27Analyst Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-27Unrestricted Policy Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-27Restricted Policy Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-28Unrestricted Storage Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-28Restricted Storage Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-28Node Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-28

Client Owner Authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-28Client Access Authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-29

The Integrated Solutions Console Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-30Integrated Solutions Console Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-30Console Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-30

The Administration Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-31Using Administration Center Portlets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-32

Administration Center Installation Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-32Installation Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-33

Server Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-34Administrative Client Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-35

Using Administrative Client Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-35Administrative Client Interactive Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-37Administrative Client Console Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-38 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-39Classroom Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-40Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-41Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-42Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-44Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-45

Unit 2: Database and Log ManipulationsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2

Lesson 1: Performance and Optimization of the Database and Log . . . . . . . .2-3Database and Recovery Log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5BUFPOOLSIZE and LOGPOOLSIZE Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-6Recovery Log Transaction Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-7

Normal Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-7Roll-Forward Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two III

• • •••

Table of Contents


Recovery Log Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-9Database and Recovery Log Mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-10

Lesson 2: Database and Recovery Log Space Allocation . . . . . . . . . . . . . . .2-11How Many Database Volumes Are Needed? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-11How Many Recovery Log Volumes Are Needed? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12

Database and Recovery Log Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-13Increase or Decrease the Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-14Space Trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-15Reorganizing the Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-16

Reorganizing the Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-16Using the DSMSERV Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-18

Lesson 3: Tivoli Storage Manager Server Backup Facilities . . . . . . . . . . . . .2-20Tivoli Storage Manager Database Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-21

Regular Only or Regular-plus-incremental Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-21Snapshot Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-22

Database Backup Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-23Backup Commands for Database Related Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-24

Database Backup-related Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-26query db . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-26query volhistory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-26

Syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-27delete volume history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-27

Lesson 4: Database Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-29Point-in-time Database Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-30Roll-forward Database Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-32Information Needed for Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-33Volume-related Information in the Tivoli Storage Manager Database . . . . . . . . . . . . . . . . . . . 2-34

Device-related Information in the Tivoli Storage Manager Database . . . . . . . . . . . . . . . . . . 2-34Examples of Restore Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-35

dsmserv restore db Utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-35Roll-forward Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-35Database Restore and Volume History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-36define dbbackuptrigger Command Syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-36

Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-37Lesson 5: Suggestions for Database Management . . . . . . . . . . . . . . . . . . . .2-38Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-41Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-42Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-43Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-44Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-45

Unit 3: Storage Pool ManipulationsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2

Lesson 1: Storage Pool Manipulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-3Tivoli Storage Manager Storage Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4Storage Area Network Tape Library Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5

Partitioning of Volume Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5Serialized Drive Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-6Serialized Mount Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-6

Storage Destinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8Migration for Disk Storage Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-9Migration Threshold Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10Manually Starting Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-11

IV IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Ending Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12Minimum and Maximum High-Migration Thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13Standard Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-14

How the Server Selects Files to Migrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-14Migration Using migdelay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-15

Enabling Cache for Disk Storage Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-15Using migdelay with migcontinue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16Migrate STGpool Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-17Collocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-19Collocation by Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20Data Shredding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-22Server Options for Data Shredding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-23Data Shredding Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-24Reclamation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-26Reclamation—Collocated versus Non-collocated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-27Reclaim STGpool Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-28Copy Storage Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-29Active-data Storage Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-30Devices for Active-data Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-31Copy Activedata Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-32Active-data Pool with Simultaneous Write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-33Basic Data Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-35Additional Data Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-36Moving Data in a Copy Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-37Disk Storage Pool Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-38Dynamic Disk Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-39

Lesson 2: Storage Pool Backup and Recovery . . . . . . . . . . . . . . . . . . . . . . .3-40Backup of Storage Pool Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-42Backup of Storage Pool Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-43Backup and Recovery Suggestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-45Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-47Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-48Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-49Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-50Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-51

Unit 4: Database and Storage Pool Problem DeterminationsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2

Lesson 1: Problem Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-3Exercise Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-6Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-8Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-9

Unit 5: Macros and ScriptsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2

Lesson 1: Coding Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-3Tivoli Storage Manager Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3Creating Tivoli Storage Manager Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two V

• • •••

Table of Contents


Committing a Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6Macro Example with Line Continuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-7Macro Examples—Task 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-8Symbol Substitution in Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9Executing a Tivoli Storage Manager Macro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-10Macro Examples—Task 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-11

Lesson 2: Coding Server Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-12Tivoli Storage Manager Server Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-12Script Flexibility and Usability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-13Administration Center Automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14

Syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15

Using the Administration Center to Define a Maintenance Script . . . . . . . . . . . . . . . . . . . . . . 5-16Administration Center Maintenance Script Wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-17Selecting Available Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-18Create a Database Backup Script Using the Wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-19Schedule Maintenance with Script Wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-20Define a Custom Script in the Administration Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-21Creating a Script in the Administration Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-22Conditional Logic Flow Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-23The if Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-24The exit Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-25The goto-label Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-26A Sample Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-27Server Processing Control in Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-28Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-30Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-31

Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-32Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-34

Unit 6: Network Configuration and Tivoli Storage Manager Performance Issues

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2

Lesson 1: Planning Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-3Network Capacity Planning Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3Application Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4Disaster Recovery Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-5Planning for Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-6

Lesson 2: Network Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-7Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-9

Lesson 3: Performance Tuning Overview . . . . . . . . . . . . . . . . . . . . . . . . . . .6-10Performance Tuning Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-11Performance Tuning Basics Continued . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-12Tivoli Storage Manager Performance Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-13

Lesson 4: Server Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-14Tivoli Storage Manager Server Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-14Tivoli Storage Manager Server Setup Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-15Performance Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-16

Back Up to Disk Storage Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-16Backupsets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-17Collocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-17Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-17Multi-threading Clients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-17

VI IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-17Storage Pool Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-18

Tuning the Tivoli Storage Manager Server with Options Overview . . . . . . . . . . . . . . . . . . . . . 6-19Logpoolsize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-19Selftunebufpoolsize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-20Expinterval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-20Maxsessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-20

Server Configuration – Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-22Server Configuration – BUFPOOLSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-23Server Configuration – USELARGEBUFFERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-24Server Configuration – I/O Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-25Server Configuration – Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-26Server Configuration – TCPWINDOWSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-27Server Configuration – TCPNODELAY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-28Server Configuration – Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-29Server Configuration – Tape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-31Server Configuration – Disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-32Server Configuration – Server Disk Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-33Server Configuration – Disk Write Cache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-34Database and Recovery Log Mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-35Server Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-36

Lesson 5: Client Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-38Tivoli Storage Manager Client General Tuning Considerations . . . . . . . . . . . . . . . . . . . . . . . . 6-38Client Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-39

dirmc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-39tapeprompt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-39include/exclude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-40

Client Tuning Settings – Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-41Client Tuning Settings – MEMORYEFFICIENTBACKUP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-42Client Tuning Settings – Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-43Client Tuning Settings – TCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-44

TCPWINDOWSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-44TCPBUFFSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-45TCPNODELAY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-45

Client Tuning Settings – DISKBUFFSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-46Client Tuning Settings – COMPRESSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-47Client Tuning Settings – Multi-session Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-48Client Tuning Settings – Multi-session Client Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-49Client Tuning Settings – Multi-session Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-50Client Tuning Settings – Multi-session Client Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-51Client Tuning Settings – Multi-session Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-52Client Tuning Settings – RESOURCEUTILIZATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-53Client Tuning Settings – QUIET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-55Client-related Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-56

Lesson 6: Transaction Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-57Errors and Interruptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-57Suggested Tivoli Storage Manager Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-58TXNGROUPMAX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-59TXNBYTELIMIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-60MOVEBATCHSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-62MOVESIZETHRESH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-63Overview of Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-64Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-65Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-66Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-68Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-69

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two VII

• • •••

Table of Contents


Unit 7: ReportingIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2

Lesson 1: Tivoli Storage Manager SQL Interface . . . . . . . . . . . . . . . . . . . . . . .7-3Accessing Information in the Tivoli Storage Manager Database . . . . . . . . . . . . . . . . . . . . . . . . 7-3Requesting Information from RDBMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4SQL Select Statement Syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5

SELECT and FROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5WHERE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6GROUP BY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6HAVING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6ORDER BY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-7

Server SQL Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-8Administrative Select Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9

SELECT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9System Catalog Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9Confirmation Message . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-10

Server SQL Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-11Virtual SQL Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-11

SQL Parameters and Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12SQLDISPLAYMODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12SQLDATETIMEFORMAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13SQLMATHMODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13QUERY SQLSESSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13

SQL Example in a Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-14ODBC Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-15

Using the ODBC Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-15ODBC Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-16

Defining Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-16Lesson 2: Tivoli Storage Manager Operational Reporting . . . . . . . . . . . . . . .7-18

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-18Tivoli Storage Manager Operational Reporting Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20

Minimizes the amount of time needed to administer Tivoli Storage Manager . . . . . . . . . . 7-20Easy to use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20

Operational Reporting Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-22Tivoli Storage Manager Management Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23Add Tivoli Storage Manager Network Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-24Add a Tivoli Storage Manager Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25Operational Report List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-26Add Report and Report Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-27Add Report Recipients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-28Add Report Missed Schedule Notifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-29View After Reports Have Been Defined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-30Operational Monitor List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31Add Monitor Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-32Add Monitor Recipients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-33Add Monitor Desktop Notify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-34Configure Reporting Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-35Configure Operational Reporting E-mail Account . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-36Configure Web Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-37Custom Summary Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-38Report Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-39Report Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-40Report Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41Notification Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-42Report Extensions Using SQL Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-43

VIII IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Test Extension Select Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-44Select Statement Builder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45Creates Interactive Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-46Interactive E-mail Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-47

Lesson 3: Tivoli Storage Manager Reporting with Tivoli Data Warehouse . .7-48Tivoli Data Warehouse: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-48

Tivoli Data Warehouse Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-49Tivoli Data Warehouse Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-50Tivoli Data Warehouse Software Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-51Tivoli Data Warehouse Report Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-52Tivoli Data Warehouse Report Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-53Tivoli Data Warehouse Report Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-54Tivoli Data Warehouse Report Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-55Tivoli Data Warehouse Report Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-56Tivoli Data Warehouse Reference Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-57Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-58Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-59Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-60Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-61Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-62

Unit 8: Problem Determination ConceptsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2

Lesson 1: Problem Determination Methodology . . . . . . . . . . . . . . . . . . . . . . . .8-3Steps for Problem Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-3Locate the Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4Hints and Tips for Libraries, Drives and Device Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6Diagnosing a Client Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-7Tivoli Storage Manager Client Error Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-8Diagnosing a Server Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-10Tivoli Storage Manager Server Error Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-11Problems with Data Flow in a Storage Area Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-13

Validating a Node's Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-14Lesson 2: Using Client and Server Tracing . . . . . . . . . . . . . . . . . . . . . . . . . .8-15Running a Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-15

Reporting a Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-15Running Trace Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-16

Server Trace Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17Diagnosing Server Problems with Trace Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17

Trace Enable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-18Trace Begin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-18Trace Flush . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-18Trace Disable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-18Trace End . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-18Trace List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-18Query Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-19

Lesson 3: Client Trace Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-20Client Trace Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22Client Trace Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-23

TRACESEGSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-23Trace Options - Continued . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-24

Lesson 4: Client and Server Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . .8-25testflag for Client Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-25

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two IX

• • •••

Table of Contents


Example Output for Client Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-26Client Instrumentation Detail Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27Client Instrumentation Detail Sections, continued . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-28Client Instrumentation Detail Sections, continued . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-29Server Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-30Server Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-31Diagnostics with Instrumentation Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-32Reporting a Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-33Providing Information to Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-34Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-35Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-36Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-37Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-38Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-39

Unit 9: Backup and Restore MethodologyIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2

Lesson 1: Backup and Restore Requirements . . . . . . . . . . . . . . . . . . . . . . . . .9-3Client Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4Data Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-5

Lesson 2: LAN-free Data Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-6LAN-free Data Functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-7Setup of the Storage Agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-8Components of the Shared Disk Pool Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-10The Shared File Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-11

Step-by-step Setup of the Shared Disk Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-11Total Overview of the Shared Pool Setup Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-12

Tivoli Storage Manager Administrator Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-12Overview of the Customized Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-14

Windows Step 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-14Windows Step 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-15Windows Step 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-16Windows Step 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-16Windows Step 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-17Windows Step 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-17AIX Steps 1 and 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-17AIX Step 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-17AIX Step 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-18AIX Step 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-18AIX Step 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-19AIX Step 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-19AIX Step 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-19AIX Step 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-19

Lesson 3: Multi-session Backup and Restore . . . . . . . . . . . . . . . . . . . . . . . .9-22Resource Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-22Regulating Backup and Archive Sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-22Regulating Restore Sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-23Multi-session Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-23Multi-session Backup and Restore Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-24Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-26

Lesson 4: Backup Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-27Instant Archive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-27Rapid Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-27Creating and Using Client Backup Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-28

X IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Generating Client Backup Sets on the Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-28Choosing Media for Generating the Backup Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-28Using Scratch Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-29Managing the Backup Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-30Restoring the Backup Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-31

Localbackupset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-31Restore Backupset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-32Restoring Backup Sets In a SAN Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-33

Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-34Lesson 5: Adaptive Subfile Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-35The Subfile Architectural Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-36

Setting Subfile Backup for Client Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-36Performing a Backup with Limited Bandwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-37

Adaptive Subfile Backup and Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-38Base File Backup Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-39Delta File Backup Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-40Adaptive Subfile Restore Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-41Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-43

Lesson 6: Image Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-44Logical Volume Backup (Image Backup) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-44

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-45Image Backup (Offline) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-45Image Snapshot Backup (Online) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-45Incremental Image Backup (Date-only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-46

Setting Up Online Image Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-46Policy for Logical Volume Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-48

Setting Up Third-party Snapshot Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-49Provide Support for External Snapshot Providers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-49Support for External Snapshot Providers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-50Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-50

Restore Image Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-51Logical Volume Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-51

Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-54Lesson 7: Setting Up Open File Support . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-55Open File Support for Backup Operations (Windows 2000, XP) . . . . . . . . . . . . . . . . . . . . . . . 9-55Configuring the LVSA to Enable Open File Support (Windows 2000, Windows XP) . . . . . . . . 9-56Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-56Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-57

Lesson 8: Backup Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-58The Backup Group Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-59

Filelist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-59Groupname . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-60Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-60

Backup Group Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-61Query Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-61Delete Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-62

Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-63Lesson 9: Moving Data By Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-64Moving Data for All File Spaces for One or More Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-65Moving Data for Selected File Spaces for One Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-65Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-66

Lesson 10: Administration through the Firewall . . . . . . . . . . . . . . . . . . . . . . .9-67Tivoli Storage Manager Firewall Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-67

TCP/IP Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-67TCP/IP Port for Administrative Sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-68

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two XI

• • •••

Table of Contents


New Administrative Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-68The Backup-Archive Client through the Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-70Tivoli Storage Manager Firewall Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-70

Method 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-70Method 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-71

Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-73Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-74Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-75Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-77

Unit 10: Enterprise AdministrationIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-2Enterprise Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-2

Lesson 1: Server-to-server Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-3Server-to-server Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-4Set Up a Remote Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-5

set servername . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-5set serverpassword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-5set serverhladdress xxx.xxx.xxx.xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-6set serverlladdress 1500 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-6set crossdefine on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-6Viewing the parameter settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-6

Set Up the Local Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-7Server Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-8

define server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-8ping server Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-9delete server Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-9

Lesson 2: Server Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-10Overview of Server Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-10Server Group Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-11

Server Group Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-11Lesson 3: Multi-server Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-13Multi-server Administration Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-13Multi-server Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-14

Lesson 4: Administrative Command Routing . . . . . . . . . . . . . . . . . . . . . . . .10-15Command Routing Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-15

Command Syntax for Performing Command Routing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-16Command Routing Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-16

Individual Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-17Server Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-17Combination of Servers and Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-17Command Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-17

Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-18Lesson 5: Enterprise Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-19Enterprise Configuration Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-19Configuration Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-20

Server Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-20Configuration Manager Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-21

Configuration Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-22Default Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-22Additional Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-23Managed Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-23Profile Subscribers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-24

XII IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Configuration Profile Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-25Configuration Profile Commands (continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-27Configuration Profile Commands (continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-28Managed Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-29

Define Managed Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-29Set Refresh Interval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-30Query Configuration Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-30Remove Profile Subscription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-30Configuration Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-31

Using the Administration Center for Enterprise Administration . . . . . . . . . . . . . . . . . . . . . . . 10-32Set Up the Configuration Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-33Beginning the Enterprise Configurations Wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-34Selecting the Role of the Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-35Set Up Sharing - Create a Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-36Select the Type of Objects for the Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-37Adding Administrators to the Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-38Adding Domains to the Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-39The Summary Page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-40View the Enterprise Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-41Working with Profiles for a Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-42Profile Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-43Managed Administrators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-44

Administrative Authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-44Lock Attribute . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-44Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-45

Managed Policy Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-46Policy Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-46Destination Storage Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-47Activated Policy Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-47

Managed Administrative Command Schedules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-48Administrative Command Schedules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-48Activate the Administrative Command Schedules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-48

Managed Server Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-49Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-49

Managed Client Option Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-50Node Association . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-50Client Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-50

Managed Servers and Server Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-51Default Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-51Server Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-51Server Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-52

Profile Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-53Profile Subscription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-53Managed Server Refresh Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-53Profile Subscription View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-54

Profile Subscribers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-55Configuration Manager Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-55The Current Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-55

Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-57Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-58Review answers begin on the next page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-59Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-60Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-61

Unit 11: Server-to-server Virtual VolumesIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-2

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two XIII

• • •••

Table of Contents


Lesson 1: Server-to-server Virtual Volumes Configuration . . . . . . . . . . . . . .11-3Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-3

Server Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-3Storage Pool Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-3TCP/IP Communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-4passwordaccess generate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-4DRM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-4

Prerequisites for Virtual Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-5Server-to-server Communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-5Archive Storage Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-5

Steps to Configure Virtual Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-6Configuration - Step 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-7

register node TYPE=SERVER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-7query node command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-8

Configuration - Step 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-9Define Server Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-9Query Server Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-10

Configuration - Step 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-11Query devclass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-12

Primary Storage Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-13Storage Pool Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-13

Database Backup - Disaster Recovery Manager Only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-14Copy Storage Pool - Disaster Recovery Manager Only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-15

Copy Pool Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-15Virtual Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-16Managing Virtual Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-19Server-to-server Virtual Volumes Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-21

Performance Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-22Monitoring Sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-22

Lesson 2: Using Server-to-server Virtual Volumes . . . . . . . . . . . . . . . . . . . .11-23Exploiting Server-to-server Virtual Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-23Site Disaster Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-24

Storage Pool Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-24Database Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-24Peer Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-24

Remote Disaster Recovery Manager File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-25Branch Office Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-26

Multiple Server Storage Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-26Policy-Based Data Placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-27Transparent Data Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-27Tape Management Minimized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-27Increased Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-27

Campus-wide Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-28Service Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-28Exploiting Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-28Volume Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-29

Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-30Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-31Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-32Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-33

Appendix A: Related Tivoli Storage Manager ProductsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-2Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-3

Lesson 1: Tivoli Storage Manager Data Protection Architecture . . . . . . . . . . A-4

XIV IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


Lesson 2: Database Backup and Restore Overview . . . . . . . . . . . . . . . . . . . A-5Tivoli Storage Manager for Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-6

Lesson 3: Database Backup Automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-8Backing Up and Restoring Database Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-9

Data Protection Application Client for Oracle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-9Data Protection Application Client for Oracle Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-10Microsoft SQL Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-10Data Protection Application Client for Microsoft SQL Server Policy Management . . . . . . . . A-10Data Protection Application Client for Microsoft SQL Server Administration . . . . . . . . . . . . . A-11Data Protection Application Client for Informix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-12

Lesson 4: Data Protection for Applications . . . . . . . . . . . . . . . . . . . . . . . . . . A-13Backing Up and Restoring Application Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-13

Tivoli Storage Manager for Application Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-13Overview for WAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-14

Lesson 5: Data Protection for Mail Servers . . . . . . . . . . . . . . . . . . . . . . . . . A-17Configuring the Data Protection Application Client for Microsoft Exchange . . . . . . . . . . . . . A-17Data Protection Application Client for Microsoft Exchange Policy Management . . . . . . . . . . A-18Data Protection Application Client for Microsoft Exchange Administration . . . . . . . . . . . . . . A-19Data Protection for Lotus Domino . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-19Data Protection for Lotus Domino Environmental Variables . . . . . . . . . . . . . . . . . . . . . . . . . A-19Data Protection for Lotus Domino–Policy Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-20

Lesson 6: Tivoli Storage Manager for Enterprise Resource Planning . . . . . A-21Data Protection Application Client for mySAP.com . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-21Incremental Backup (RMAN) Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-22Administration Assistant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-22

Lesson 7: Tivoli Storage Manager for Hardware . . . . . . . . . . . . . . . . . . . . . A-24Data Protection Application Client for EMC Symmetrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-24Data Protection Application Client for ESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-25

Functions of Data Protection Application Client for ESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-26Operating Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-26Active Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-27Passive Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-27

Data Protection for Hardware and DB2 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-28DB2 UDB Database Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-29DB2 UDB Database Quick Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-29Scheduled Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-29Integration with Tivoli Storage Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-29

DB2—UDB Backup Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-30How UDB DB2 Stores Data Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-30Define a New Management Class in an Existing Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . A-31

Backup Copy Group Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-31Special Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-31Archive Copy Group Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-32

Lesson 8: Tivoli Storage Manager for System Backup and Recovery . . . . . A-33Protect Your Most Vital Asset—Your Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-33Tivoli Storage Manager Integration with SysBack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-34

Comprehensive Backup and Recovery Options through a Simple, Efficient Interface . . . . . A-34Central Management and Automation Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-34Integrating with Tivoli Storage Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-35Integrating Network Boot and Installation Features in Support of RS/6000 SP Systems . . . A-35Provides Offline Mirror Backup Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-35

Two-way Pull Backup Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-36Three-way Pull Backup Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-37

Lesson 9: Hierarchical Storage Management Overview . . . . . . . . . . . . . . . A-38

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two XV

• • •••

Table of Contents


Tivoli Storage Manager for Space Management and Tivoli Storage Manager HSM for Windows A-38Lesson 10: Continuous Data Protection for Files . . . . . . . . . . . . . . . . . . . . . A-40Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-42Review Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-43Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-44

Tivoli Professional Certification Special Offer For Having Taken This Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IWhy Get Certified? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IWhat Is Role-based Certification? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IWhen Should I Attempt a Certification Exam? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IIWhere Can I Take a Certification Exam and How Much Do They Cost? . . . . . . . . . . . . . . . . . . . . IIWhere Can I Get More Information? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II

XVI IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Table of Contents


• • • ••

i

Preface

IBM Tivoli Storage Manager 5.4 Implementation and Administration

Part Two

ii IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface


Course Description

This five-day course is designed to provide in-depth administration training with IBM Tivoli Storage Manager 5.4 Extended Edition, at a skill level that is beyond the Implementation and Administrator Part One course. Through the use of reporting tools, administrators will be able to control their Tivoli Storage Manager environment, and determine which actions to take. These tasks will vary according to the action they are trying to accomplish. The classroom setup for this training includes installing and configuring multiple servers, server-to-server and virtual volumes. The students will participate in group problem determination exercises, and be presented with options for a tapeless backup, optimal data management, and tuning the backup-archive client and the Tivoli Storage Manager 5.4 server for optimal performance.

The Student Exercises are to be used with Windows 2003 and AIX 5.3. The lecture material covers all supported Intel and UNIX platforms.

Audience

This course is intended for System Administrators and Implementers who will be responsible for implementing a Tivoli Storage Manager 5.4 solution.

Curriculum Roadmap

Role

This role is focused on the day-to-day operations of Tivoli Storage Manager.

The latest information about IBM Tivoli education offerings can be found online at

http://www.ibm.com/software/tivoli/education/.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two iii

• • •••

Preface


If you have any questions about our education offerings, send an e-mail to the appropriate alias for your region:

– Americas: [email protected]

– Asia Pacific: [email protected]

– EMEA: [email protected]

3

Where to Find More InformationTraining Roadmaps Onlinehttp://www-306.ibm.com/software/tivoli/education/

iv IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface


Course Prerequisites

The following list contains the prerequisite knowledge or Tivoli product knowledge an attendee must have prior to taking the course:

• All students should have a working knowledge of Internet browsers.

• All students should have a basic understanding of client-server concepts.

• For classes taught on a Windows platform, the students should have the skills to navigate through Windows 2003 applications.

• For classes taught on a UNIX platform, the students should have experience using the UNIX command-line interface, allocating file system space, and connecting devices to a UNIX server.

• All students should have taken the IBM Tivoli Storage Manager 5.4 Implementation and Administration Part One class prior to taking this class, or have extensive knowledge of the basic concepts of the product.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two v

• • •••

Preface


Course Objectives

5

Course Objectives

Upon completion of this course, you will be able to:

• Identify storage solutions using Tivoli Storage Manager.• Describe the purpose of enterprise administration• Establish server-to-server communications.• Work with enterprise administration and server-to-server

virtual volume functions.• Configure single administrative logon.• Exploit server-to-server virtual volumes.• Design and create backup-archive solutions.• Configure and use multi-session backup and restore.• Configure and use subfile backup and restore.

6

Course Objectives - continued


•Design storage structure for tapeless solution using large disk pools.

•Configure and use online image backup and restore.•Enable and use file grouping and open file support.•Set up and use the move nodedata command for faster restores.

•Enforce data shredding.•Configure active-data storage pools.•Optimize Tivoli Storage Manager with performance tuning parameters.

•Prepare reports with Operational Reporting.

vi IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface


Audience

This course is designed for Tivoli Storage Manager (TSM) administrators who already understand the basic concepts of Tivoli Storage Manager. The course will provide some base knowledge while delving into some of the more complex issues that surround managing an installed Tivoli Storage Manager solution.

Prerequisites

The following skills are expected:

1. Knowledge of the base concepts of Tivoli Storage Manager

2. Attendance of the IBM Tivoli Storage Manager Implementation and Administration Part One course or working knowledge of the Tivoli Storage Manager product

3. Knowledge of file system concepts and sequential access library implementation

4. The ability to use the appropriate operating system utilities to perform exercises

7

Course Objectives - continued


• Invoke the trace function for both the client and the server for problem determination.

•Perform Tivoli Storage Manager performance tuning and determine network configuration issues.

•Perform problem determination to recover database mirror, database volumes and storage pool volumes.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two vii

• • •••

Preface


Course Outline

The following list gives a high-level description of the contents of this course. Each unit has an overview presentation, and all have a series of student exercises designed to reinforce the concepts presented. The units in this course build on each other in a progression of knowledge building. The course contains the following units:

• Unit 1: Tivoli Storage Manager 5.4 Concept Review and New Features

This unit presents the new features available for Tivoli Storage Manager 5.4 in an overview format. There is also a review of basic Tivoli Storage Manager concepts to make sure there is a base of knowledge for the course to proceed on. This unit also covers different administrative interfaces.

• Unit 2: Database and Log Manipulation

This unit explains the concepts dealing with the database and log and how they are sized and expanded.

• Unit 3: Storage Pool Manipulation

This unit explains the concepts dealing with the storage pools. The exercises for this unit provide the opportunity to learn how to manage the data storage used by the storage hierarchy.

• Unit 4: Database and Storage Pool Problem Determinations

This unit consists of exercises. There is a short presentation and then you will work with three different problem determination labs that help you learn how to recover the Tivoli Storage Manager server after anomalies have occurred.

• Unit 5: Macros and Scripts

In this unit you see how writing macros and scripts can help with the automation of common tasks. You also see that scripts can be scheduled, thus reducing the number of schedules required.

• Unit 6: Network Configuration and Tivoli Storage Manager Performance Issues

Performance and tuning parameters and options will be covered in this unit. Guidelines will be presented and you will also work with network backup capabilities when there are nodes that exist across the WAN environment from the Tivoli Storage Manager server installation.

viii IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface


• Unit 7: Reporting

Reporting tools are covered in the unit. The SQL interface along with the ODBC driver and connectivity are reviewed. The Operational Reporting tool is also presented so you see the capabilities of this included utility.

• Unit 8: Problem Determination Concepts

Since there may be times when Tivoli Storage Manager support needs you to run a trace, this unit covers what traces are and how to run them. Tracing will be examined from the server and client perspective.

• Unit 9: Backup and Restore Methodology

This unit covers many backup and restore methods not presented in the IBM Tivoli Storage Manager 5.4 Implementation and Administration Part One course. Some of the items examined will be multi-session backup and restore, open file support, image backup (static and online snapshot), backup groups and moving data by node.

• Unit 10: Enterprise Administration

Unit 10 establishes how to configure multiple instances of Tivoli Storage Manager servers in support of enterprise administration. Also covered in this unit is the establishment of a configuration manager server for propagation of objects in a controlled methodology from the manager to other Tivoli Storage Manager servers.

• Unit 11: Server-to-server Virtual Volumes

The capability to store backup and archive data for a Tivoli Storage Manager server on another Tivoli Storage Manager server is studied in this unit. This capability is available through the use of virtual volumes. The exploitation of these virtual volumes database backups and disaster recovery are explored.

• Appendix A: Related Tivoli Storage Manager Products

The add-on products for Tivoli Storage Manager, such as the Data Protection for Mail, Data Protection for Database, and others are described in this closing unit.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ix

• • •••

Preface


Typographical Conventions

In this course, the following typographical conventions are used.

Product Information

Software Release Numbers

The Tivoli software and release numbers referred to in this guide are:

Convention Usage

UPPERCASE Options and parameters, when not shown in examples, will appear in UPPERCASE.

BoldCommands, keywords, file names, authorization roles, URLs, or other information that you must use literally appear in bold.

ItalicsVariables and values that you must provide appear in italics. Words and phrases that are emphasized also appear in italics.

Bold Italics New terms appear in bold italics when they are defined in the text.

MonospaceCode examples, output, and system messages appear in a monospace font.

>>In this manual, the arrow character is used as a path arrow. The arrow indicates the path to the named windows.

Tivoli Software Release Number

IBM Tivoli Storage Manager 5.4.0

x IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Preface


Product Documentation

IBM Tivoli product documentation is available on the Student Resources CD. For access to documentation outside the classroom environment, please visit the IBM Web site.

Recommendations in this Course

Recommendations, suggestions, and best practices provided in this course are the results of product testing and monitoring in a variety of test and production environments. Results may vary in your environment. Refer to the documentation for your platform for more information.

• • • ••

1-1



1-2 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



Introduction

This unit reviews Tivoli Storage Manager functions and gives an overview of new features of Tivoli Storage Manager 5.4. Notable new features include data shredding and active data pools.

Objectives

2

ObjectivesUpon completion of this unit, you will be able to:

• Describe the changes in Tivoli Storage Manager 5.4• Understand migration issues from previous versions to 5.4• Identify the functions of Tivoli Storage Manager• Review Tivoli Storage Manager server components• Describe Tivoli Storage Manager 5.4 administrative functions andinterfaces

• Identify the Tivoli Storage Manager servers set up for this course

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two 1-3

• • •••

Unit 1: Tivoli Storage Manager 5.4 Concept Review and New FeaturesTivoli Storage Manager Products


Lesson 1: Tivoli Storage Manager Products

Tivoli Storage Manager Express

Tivoli Storage Manager Express is intended for small businesses, and is offered for Windows environments. It can be installed and configured in 30 minutes. Tivoli Storage Manager Express uses the familiar grandfather-father-son backup hierarchy. This product can be upgraded to Tivoli Storage Manager Extended Edition without loss of client data. Training for this product is offered in the Tivoli Storage Manager Express training course. You can find more information on this product at the following Web site:

http://www-306.ibm.com/software/tivoli/products/storage-mgr-express/

3

Tivoli Storage Manager Products•Tivoli Storage Manager Express

• Windows operating system• Intended for small businesses• Quick installation• Disk-disk backup or grandfather-father-son tape management

•Tivoli Storage Manager Basic Edition• Extensive operating system compatability• Intended for small enterprise• Archive and hierarchical storage management• Extensive policy management

•Tivoli Storage Manager Extended Edition• Extensive operating system compatability• Intended for large enterprise• Archive and hierarchical storage management• Extensive policy management• Advanced tape library support• Disaster recovery planning

1-4 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007• • • ••



Tivoli Storage Manager Basic Edition

Tivoli Storage Manager Basic Edition is intended for small enterprise environments. It is offered for a wide range of operating systems. Key features of Tivoli Storage Manager Basic Edition include:

• Progressive backup methodology

• Archive management

• Hierarchical storage management

• Tiered storage management of backup and archive data

• Extensive policy management

• Advanced tape management

• Tape library support

You can find more information about this product at the following Web site:

http://www.ibm.com/software/tivoli/products/storage-mgr/

Tivoli Storage Manager Extended Edition

Tivoli Storage Manager Extended Edition is intended for large enterprise environments. It encompasses all of the key features of the Tivoli Storage Manager Basic Edition, but expands for the following capabilities:

• Advanced tape library support

• NDMP (Network Data Management Protocol) backup for NAS (Network Attached Storage)

• Disaster recovery planning

Of these 3 products, Tivoli Storage Manager Extended Edition offers the richest set of features. This course will focus on the capabilities of Tivoli Storage manager Extended Edition. You can find additional information about this product at the following Web site:

http://www.ibm.com/software/tivoli/products/storage-mgr-extended/

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two 1-5• • •••




Enterprise Administration

The Tivoli Storage Manager server-to-server communication provides several single point-of-control management functions. Single logon is available for the administrator, and command routing is available to all the configured Tivoli Storage Manager servers. This allows centralized management of the configuration of the Tivoli Storage Manager servers.

Virtual Volumes

You can electronically transport data over a long distance to a remote Tivoli Storage Manager server with virtual volumes. This is the first step toward automating the operations of the Tivoli Storage Manager server. In addition, storing the data at a remote location is a necessary step for disaster protection.

4

•Tape Library Sharing•Disaster Recovery Management•Network Data Management Protocol (NDMP) Protection•Data Protection for:

• Mail• Application Servers• Databases• Enterprise Resource Planning• Copy Services, and Advanced Copy Services

•LAN-free Data Movement•Space Management•SANergy File Sharing





Tivoli Disaster Recovery Management

Tivoli Disaster Recovery Management automatically updates the disaster recovery plan, keeps track of the off-site data or cartridges, and records all necessary machine information required to recover the server and its clients.

Tivoli Storage Manager Library Sharing

Library sharing handles the cartridges and the tape drives in the tape library to create a consolidated storage environment. This feature allows for improved efficiency of storage use.

Tivoli Storage Manager with Storage Area Networks

In this course, there will be a brief overview of the components and management software used to set up a storage area network (SAN), as well as an explanation of concepts behind SAN implementation.

LAN-free Backup and Restore

LAN-free data movement allows backup-archive clients to utilize a SAN environment to send the bulk of the data to back up on the server. The LAN is still used for sending metadata. This implementation frees up bandwidth on the LAN for normal traffic and provides a high-speed solution for backup and restore processes.

Additional Client Backup and Restore Methods

Achieve the best backup and restore solutions with multiple sessions, subfile backups, active-data storage pools, image backups, differential backups, or incremental by date backups. Each of these methods serves a specific client need.

Tracing

Tracing helps find bottlenecks and other problems. It is important to note, however, that tracing is most useful when performed with the guidance of Tivoli support.




Optimizing Performance for LAN and SAN

Optimizing performance is an ongoing process requiring time and patience. This course is fortified with tips to help with this complicated task.

Tivoli Storage Manager Reporting

The Tivoli Storage Manager database contains a wealth of data accessed through SQL select statements. The challenge is to find the data you need for usable reports. Tivoli Storage Resource Manager and Tivoli Storage Manager Operational Reporting make this task more manageable.

Tivoli Storage Manager for Space Management

Tivoli Storage Manager for Space Management also known as Hierarchical Storage Management is available for those who require space management.

Tivoli Storage Manager for Data Protection

Tivoli Data Protection for Copy Services, Advanced Copy Services, Hardware, Databases, Mail, and Applications are available to back up and restore databases, mail server applications, and so forth. One Data Protection product will be discussed in detail in this unit and an overview of the other Data Protection products will be provided.

Tivoli Storage Manager’s System Backup and Recovery

To provide a more complete system recovery package, Tivoli Storage Manager 5.4 integrates with Tivoli Storage Manager for System Backup and Recovery (SysBack) for recovering AIX systems, and Cristie Bare Machine Recovery and Automated System Recovery (ASR) to create bare machine restores for Windows. These are especially useful when performing disaster recovery.




NDMP Support

Tivoli Storage Manager Extended Edition uses Network Data Management Protocol (NDMP) to perform high-performance, scalable backups and restores for Network Attached Storage (NAS) devices.




Tivoli Storage Manager Extended Edition 5.4 Product Enhancements

5Tivoli Storage Manager Extended Edition 5.4 Product Enhancements•New Administration Center functionality:

• Installation package updated• Performance fixes• Command-line applet usability fixes• More commands are now supported

•Active Data Pools:• Contain only active versions• Collocation of active data• Copy active data or simultaneous-write function• Sequential-access storage• Backup data only – no archive or Hierarchical Storage Manager data




6Tivoli Storage Manager Extended Edition 5.4 Product Enhancements, continued

•Backup Set enhancements:• Point-in-time file selection• New data types – file and image• Individual file restore

•Network Data Management Protocol changes:• Filer-to-server backup• Back-end data movement support for Network Data Management

Protocol images• Offsite vaulting

•Tivoli Storage Manager Express:• Upgrade from Tivoli Storage Manager Express to Tivoli Storage

Manager Extended Edition• Preserves backed up data and Tivoli Storage Manager Express

configurations

7

Tivoli Storage Manager Extended Edition 5.4 Product Enhancements, continued

•Backup-archive client updates:• New option to minimize memory utilization• Prevent unwanted schedule actions

• IBM Support Assistant Plug-in availability:• Tivoli Storage Manager Data Protection for Domino• Tivoli Storage Manager Data Protection for Oracle• Tivoli Storage Manager Backup-archive client

•New supported platforms and hardware:• Windows Vista client toleration• Mac OS X Intel support• Optical device support for Linux




There is a free Web-based training course which introduces in greater detail the new features for Tivoli Storage Manager 5.4. It is intended for Administrators, Operators, and Implementers who are experienced with Tivoli Storage Manager 5.3.x. Information about this course is available at the following Web site:

http://www-306.ibm.com/software/tivoli/education/E063615R01269G83.html

8

Tivoli Storage Manager Extended Edition 5.4 Product Enhancements, continued

• Improved Security:• Data Shredding

–Automatically after the data is deleted or manually by command–Random-access disk–Primary pools only–Sensitive data objects are destroyed (overwritten)

• Encryption at tape drive–Managed by the application–Encryption key is set on the device class–Applies to the IBM TS1120 tape drives

New features are detailed in the Differences for IBM Tivoli Storage Manager 5.4 training course


Unit 1: Tivoli Storage Manager 5.4 Concept Review and New FeaturesMigration to Tivoli Storage Manager 5.4


Lesson 2: Migration to Tivoli Storage Manager 5.4

Migrating to Tivoli Storage Manager Version 5.4

To migrate to the next version of Tivoli Storage Manager, first save all your recovery files such as dsmserv.opt, volhist.out, devconfig.out, and perform a full database backup. Then, uninstall any prior version of Tivoli Storage Manager before installing Tivoli Storage Manager Version 5.4. Delete all the license files, including the entire directory in which the previous Tivoli Storage Manager version was installed.

Resources to set up a Tivoli Storage Manager version 5.4 server in a test environment will also provide a way to test new functions before upgrading the production environment.

9

ADSM 3.1.X.XTSM 3.7.X.XTSM 4.1.X.XTSM 4.2.X.XTSM 5.1.X.XTSM 5.2.X.XTSM 5.3.X.X

TSM EXPRESS

TSM 5.4

Migrating to Tivoli Storage Manager 5.4


• • •••



Information for upgrading to the latest version is available at the following Web site:

http://www-1.ibm.com/support/docview.wss?rs=663&uid=swg21251665

Storage Agent

Beginning with Tivoli Storage Manager Version 5.2, the server and storage agent are no longer compatible with prior releases of the server and storage agent. If you have implemented library sharing, server-to-server, enterprise configuration or LAN-free, it is important to note that all Tivoli Storage Manager servers and storage agents must be upgraded to a compatible and supported level of Version 5.2 or later code before invoking any Version 5.2 library sharing, server-to-server, or LAN-free communication.

When upgrading multiple servers participating in library sharing to Tivoli Storage Manager 5.4, the servers acting as library manager must be upgraded first to maintain compatibility among the servers acting as library clients or storage agents. When upgrading multiple servers participating in server-to-server to Tivoli Storage Manager 5.4, all servers must be upgraded at the same time.

As a general rule, Tivoli Storage Manager server normally supports storage agents and library clients at the same version, release, and modification level, down to the modification level minus two. When the product release changes, Tivoli Storage Manager will normally support the last two maintenance releases of the previous release. When upgrading multiple servers participating in library sharing to Tivoli Storage Manager 5.4, the servers acting as library managers must be upgraded first to maintain compatibility among the servers.

Note: An exception to this rule is where a fix or a product enhancement requires concurrent code changes to server, storage agent, and library client. As a rule, read the readme files that come with the release or patch to check for contingencies that would affect your environment.

Device Support

Before migrating your Tivoli Storage Manager server, make sure the new server can support your current storage devices. Refer to the Device Support section of the Tivoli Storage Manager Web site at

www.ibm.com/software/sysmgmt/products/support/IBMTivoliStorageManager.html


• • • ••



Device Drivers

There are some changes to device drivers in Tivoli Storage Manager 5.4.

Note: Users of 3570, 3590 or LTO tape libraries must install the IBM tape device driver from IBM Storage Systems Group. Install the latest, most current driver from ftp://ftp.software.ibm.com/storage/devdrvr/.

Tivoli Storage Manager Express

You can install Tivoli Storage Manager 5.4 directly over Tivoli Storage Manager Express. You do not need to uninstall Tivoli Storage Manager Express.

Ensure that your Tivoli Storage Manager Express database is backed up before installing Tivoli Storage Manager 5.4. After installing the Tivoli Storage Manager server, you must issue the DSMSERV command specifying the UPGRADEDB parameter.

A special section of the online help in the Administration Center is available for users upgrading from Tivoli Storage Manager Express, and a set of FAQs (frequently asked questions) is available at

www-1.ibm.com/support/docview.wss?rs=3039&tc=SSRQGY& uid/docview.wss?rs=3039&tc=SSRGY&uid=swg21243392.


• • •••

Unit 1: Tivoli Storage Manager 5.4 Concept Review and New FeaturesTivoli Storage Manager Components Overview


Lesson 3: Tivoli Storage Manager Components Overview

Tivoli Storage Manager consists of the following basic components:

• Tivoli Storage Manager server – The Tivoli Storage Manager server is the program that provides backup, archive, space management and administrative services to clients. The server program must be at the necessary level to provide all these services. The Tivoli Storage Manager server contains the following two items:

– Tivoli Storage Manager database – A collection of information about all objects managed by the server, including policy management objects, users and administrators, and client nodes.

– Tivoli Storage Manager recovery log – A log of updates that are about to be written to the database. The log can be used to recover from system and media failures.

• Administrative client – A program that runs on a file server, workstation or mainframe that allows administrators to control and monitor the server through administrator commands. Compare with backup-archive client.

10

Tivoli Storage Manager Components

• Tivoli Storage Manager server• Database and recovery log• Administrative client • Backup-Archive client

– LAN managed node– SAN managed node

• Storage Pools– Primary– Copy

• Policy• Schedules

– Administrative command schedule– Client schedule


• • • ••



• Backup-archive client – A program that runs on a workstation or file server and provides a means for users to back up, archive, restore, and retrieve files. It includes LAN and SAN attached nodes.

• Storage pools – Named sets of storage volumes that Tivoli Storage Manager uses to store client data. Each storage pool represents a collection of volumes sharing the same media type. Clients’ backed up, archived, or space-managed data resides in storage pools.

– Primary storage pool – A named set of volumes that Tivoli Storage Manager uses to store backup versions of files, archive copies of files, and files migrated from hierarchical storage management client nodes. A primary storage pool may be backed up to a copy storage pool.

– Copy storage pool – A named set of volumes that contain copies of files residing in primary storage pools. Copy storage pools are used only to back up the data stored in primary storage pools. A copy storage pool cannot be a destination for a backup copy group, an archive copy group, or a management class (for space-managed files).

• Policy – A data storage management environment of three basic resources: client systems, rules, and data. The client systems contain the data to be managed. They also specify how management must occur, such as, in the case of backup, how many versions to keep and where to store them. Tivoli Storage Manager policies define the relationships among these three resources.

• Schedule – A database record that describes planned client operations or administrative commands.

– Administrative command schedule – A database record that describes the planned processing of an administrative command during a specific time period.

– Client schedule – A database record that describes the planned processing of a client operation during a specific time period. The client operation can be a backup, archive, restore, retrieve operation, a client operating system command, or a macro.

• Enterprise Administration – The functionality provided after two or more Tivoli Storage Manager servers are configured for server-to-server communications. The features available are command routing, servergoups, and centralized enterprise administration using the Administration Center.


• • •••



Database and Recovery Log Configuration

Server Database

The database manages information about client files residing in storage pools and provides for the other server components. The database contains, among other things:

• Access control information for administrative clients

• Information about registered client nodes

• Policies and schedules

• The activity log and event records

• Information about Tivoli Storage Manager volumes

• Data storage inventory

11Database and Recovery Log Configuration

TSM server database contains information about:• Administrative clients and registered client nodes• Policy and schedules• Activity log and events records• Tivoli Storage Manager volumes• Data storage inventory

TSM SERVERRecovery Log Database


• • • ••



Recovery Log

The server uses the recovery log to record all changes to the database. When a change occurs, the recovery log is updated with some transaction information before the database is updated. This enables uncommitted transactions to be rolled back or applied during recovery so the database remains consistent.

To support concurrent transactions, the recovery log holds log records in a log buffer. Transaction log records stay in the log buffer until it is full or until a transaction completes.

When the transaction log record is written to the recovery log, a recovery point is recorded and the data is committed to the database. If the database needs to be recovered, the server uses the recovery point in the recovery log to bring the database back to its last point of consistency. A point of consistency is a time when all recoverable information in the database matches the data managed by the server. If a failure occurs before a transaction is committed to the database, the server rolls back any changes made to the database pages.

The log is treated as a circular array of blocks with the head ( newest log records) always chasing the tail (oldest records). The server will never let the head overtake and overwrite the tail; it must take some other action. As transactions commit, they free up log space and allow the tail to move forward. The recovery log saves some records for transactions that have already been committed, but only to the extent necessary to redo processing on recovery. The recovery log can also be used for roll-forward recovery of the database during disaster recovery.


• • •••



Server Options Configuration

The default locations of the server options file are shown in the following table.

Platform File Location

AIX /usr/tivoli/tsm/server/bin/dsmserv.opt

HP/UX /opt/tivoli/tsm/server/bin/dsmserv.opt

MVS As specified by DD name options

Solaris /opt/tivoli/tsm/server/bin/dsmserv.opt

Windowsc:\Program Files\Tivoli\tsm\serverX\dsmserv.optwhere X is the number of the server instance you want to configure

12

Windows options:• Communication• Backup and archive processing• Restore and retrieve processing• Scheduling and language• Format• Command processing• Authorization• Error processing• Transaction processing UNIX options:

• Communication• Backup and archive processing• Restore and retrieve processing• Scheduling and language • Command processing• Authorization• Error processing• Transaction processing

Server Options File


• • • ••



Communication Options

You can choose one of the following communication methods:

NAMEDPIPES - Specifies the named pipes communication method option.

NONE - Specifies that no communication method is used. This option does not allow users to connect to the server and is useful for experimenting with policy commands.

SHAREDMEM - Specifies the shared memory communication method option. This method uses the same area of memory to send data between several applications at the same time. Both the server and the backup-archive client need to be configured to support the shared memory communication method, and they must be installed on the same machine.

SNMP - Specifies the SNMP communication method option.

TCPIP - Specifies the TCP/IP communication method option. This is the default.

Volume History File

Volume history information is automatically collected by the server and maintained in the Tivoli Storage Manager database. This information is used in database recovery. The Tivoli Storage Manager server collects the following types of information:

• Volumes used to back up a Tivoli Storage Manager database

• Volumes used to export administrator, node, policy, or server data

• Sequential access storage pool volumes that have been reused, added, or deleted

• Backupset

VOLUMEHISTORY

This option specifies the names of the files to be automatically updated whenever server sequential volume history information is changed. There is no default for this option.

You can include one or more volumehistory options in the dsmserv.opt file. When you use multiple volumehistory options, Tivoli Storage Manager automatically updates and stores a backup copy of the volume history information in each file specified.


• • •••



The syntax for this option is

volumehistory path/filename

where path is the name of the drive letter or file system and filename is the actual file name you want Tivoli Storage Manager to use to store a backup copy of the volume history information. For example,

volumehistory volhist.out

or

volumehistory /usr/tivoli/tsm/server/critfiles/volhist.out

Device Configuration

Device configuration information is also stored in the Tivoli Storage Manager database with a define devclass command. This information includes drive and library definitions created using the define drive and define library commands.

It is important to keep copies of devconfig.out/devcnfg.out (UNIX/W32) and volhist.out for database recovery purposes. This is done by updating the options file.

DEVCONFIG

This option specifies the name of a file in which you want Tivoli Storage Manager to store a backup copy of device configuration information.

Tivoli Storage Manager stores the following information in the device configuration file:

• Device class definitions created using the define devclass command

• Drive definitions created using the define drive command

• Drive definitions created using the define library command

You can include one or more DEVCONFIG options in the dsmserv.opt file. When you use multiple DEVCONFIG options, Tivoli Storage Manager automatically updates and stores a backup copy of device configuration information in each file you specify.

An example of the use of this option follows (as with the VOLUMEHISTORY option a path name can qualify the file name):

devconfig devcnfg.out


• • • ••



Policy Management Overview

Policy Domain - A group of nodes managed by the same set of policy constraints as defined by the policy sets. A node (workstation) may only be defined to one policy domain per server. A node may be defined to more than one Tivoli Storage Manager server.

Based on service level agreements with data owners, an Tivoli Storage Manager administrator can plan:

• What data to backup

• What data to archive

• Where to store the data

• Number of versions to retain

• The retention period

• Other specifications based on business needs

Policy Set - A collection of management class definitions. A policy domain may contain a number of policy sets, but only one policy set in a domain can be active at a time.

Management Class - A collection of management attributes describing backup and archive characteristics. There are two sets of management class (MC) attributes, one for

13

Policy Management Overview

Policy domain: Appserver

Policy set OLD

Policy set NEW

Policy set ACTIVE

MC1BackCG

ArchCG

MC2ArchCG

NODES:ACCTREP1ACCTREP2MKT1. . .

SCHEDULES:DAILYWEEKLYMONTHLY. . .

Grace backup retentionGrace archive retention

Include c:\taxes\* 7YEARSDSMC Archive -ARCHMc=MC

Server

Client


• • •••



backup and one for archive. These management attributes describe how the server is to manage backed up or archived files. A set of attributes is called a copy group. There are two types of copy groups:

• Backup copy group—Holds the rules for backup data

• Archive copy group—Holds the rules for archive data

While these two copy groups serve different purposes, they also share some common characteristics. They both specify where to store the data sent from backup and archive operations. The copy group destination parameter specifies a valid primary storage pool to hold the backup or archive data.

For Tivoli Space Manager clients only, there are parameters that affect space management.


• • • ••



Tivoli Storage Manager Storage Hierarchy

A storage pool is a named set of volumes. When a user backs up, archives, or migrates a file from a client node, the server looks at the management class that is bound to the file. The management class specifies the destination, the storage pool in which to store the file.

Storage hierarchy is a logical ordering of primary storage pools. In Tivoli Storage Manager, the storage hierarchy is defined by identifying the next storage pool in a storage pool definition.

Copy storage pools are used to back up the data stored in primary storage pools.

14

Tivoli Storage Manager Hierarchy

=Next Primary Storage Pool Copy Storage Pool

Clientdata

Destination

Primary Storage Pool


• • •••



Daily Flow of Operations

• Migrate disk storage pools to create space for the scheduled data backups.

• Back up clients. If you are backing up to disk storage pools, a greater number of simultaneous backups can take place.

• Back up storage pools.

• Back up database, volume history file, and device configuration file.

15

Daily Flow of Operations – Example Timing•20:00 Start client backups•03:00 Start storage pool backups•06:00 Start database backup•07:00 Script runs to backup volhist and devconfig•07:30 Prepare Disaster Recovery plan if utilized•07:45 Run reports•08:00 Check out TSM environment•09:00 Expire inventory•12:00 Start reclamation•18:00 Start migration•19:30 Stop migration•BACK to 20:00 and the cycle starts again


• • • ••

Unit 1: Tivoli Storage Manager 5.4 Concept Review and New FeaturesServer Administration


Lesson 4: Server Administration

Administrative Functions

Administrator Privilege Classes

To manage and monitor the Tivoli Storage Manager server, an administrator requires system, operator, or analyst privilege.

System Privilege

An administrator with system privilege can manage:

• System responsibilities

16

Administrative Functions

Administrators can perform the following activities when managing Tivoli Storage Manager security:

• Manage administrators• Manage levels of administrative authority • Manage administrator access to the server and clients • Manage password login procedures • Manage the server console


• • •••



• Enterprise management setup

• Tivoli Storage Manager security management

Operator Privilege

Administrators with operator privilege control the immediate operation of the server and the availability of storage media. An administrator with operator privilege can manage:

• The Tivoli Storage Manager server

• Tivoli Storage Manager sessions

• Tape operations

Analyst Privilege

An administrator with analyst privilege can issue commands that reset the counters tracking server statistics.

Unrestricted Policy Privilege

An administrator with unrestricted policy privilege can manage the backup and archive services for client nodes assigned to any policy domain. When new policy domains are defined to the server, an administrator with unrestricted policy privilege is automatically authorized to manage the new policy domains.

An administrator with unrestricted policy privilege can manage:

• Tivoli Storage Manager nodes

• Tivoli Storage Manager policy

• Tivoli Storage Manager schedules

Note: System privilege is required to copy, define, or delete the policy domains themselves.


• • • ••



Restricted Policy Privilege

An administrator with restricted policy privilege can perform the same operations as an administrator with unrestricted policy privilege but only for specified policy domains.

Unrestricted Storage Privilege

An administrator with unrestricted storage privilege has the authority to manage the database, recovery log, and all storage pools.

An administrator with unrestricted storage privilege can manage:

• The Tivoli Storage Manager database and recovery logs

• Tivoli Storage Manager devices

• Tivoli Storage Manager storage pool volumes

Note: An administrator with unrestricted storage privilege cannot define or delete storage pools.

Restricted Storage Privilege

Administrators with restricted storage privilege can manage only those storage pools for which they are authorized. They cannot manage the database or recovery log.

Node Privilege

The node administrator should have owner or client privileges for one client. Use the q admin command to query Tivoli Storage Manager administrators. To restrict the querying privileges of the client administrator, issue the q opt command from the server.

Client Owner Authority

• Access to the client through the Web backup-archive client or native backup- archive client.


• • •••



• Data ownership and the right to physically gain access to the data remotely. Back up and restore files on the same machine or a different one and delete file spaces or archive data.

• Access to the data from another machine using the NODENAME parameter.

• Ability to change the password of the client node under their authority.

• This is the default authority level for the client at registration. An administrator with system or policy privileges to a client’s domain has client owner authority by default.

Client Access Authority

• Access to the client only through the Web backup-archive client.

• Ability to restore data only to the original client.

• A user ID with client access authority cannot access the client from another machine using the NODENAME parameter.

• This authority level is useful for personnel who help users without system or policy privileges back up or restore data. The client data can only be restored to the original client. A user ID with client access privileges cannot directly access client data from a native backup-archive client.


• • • ••



The Integrated Solutions Console Interface

Integrated Solutions Console Infrastructure

The Tivoli Storage Manager’s Integrated Solutions Console interface assists the administrator in managing mutiple Tivoli Storage Manager servers within a single, integrated console.

The ISC builds on top of the WebSphere Application Server and WebSphere Portal base and includes lightweight versions of both in the ISC run times. It looks for common problems, actions, and subtasks across the range of ISC components in order to provide reusable services. Basing the ISC on a lightweight portal infrastructure provides the ability to aggregate independent tasks and information into a single organized presentation.

Console Components

In the case of Tivoli Storage Manager, an instance of an Integrated Solutions Console consists of the ISC framework hosting a set of console components, built by the Storage Group. The framework can also host console components from other IBM ISC framework- aware applications such as DB2.

17

The Integrated Solutions Console (ISC)• ISC Concepts

• Allows multiple Tivoli Storage Manager servers to be managed from the same console

• Built on WebSphere Application Server and WebSphere Portal Server

• Can host console components from other ISC aware applications such as DB2

• Installation Requirements• Can be installed on the same machine as the Tivoli Storage Manager server

• The Administration Center requires 512 MB of RAM not including overhead for Tivoli Storage Manager, the OS and other applications

• Pentium 4 processor at a minimum speed of 1.4 GHz• Consider installing on separate machine if other applications are managed from ISC


• • •••



The Administration Center

The GUI interface for Tivoli Storage Manager administration is called the Administration Center. The Administration Center is a plug-in to IBM’s Integrated Solutions Console (ISC). Use the Administration Center to centrally configure and manage your Tivoli Storage Manager environment.

Note: Commands to start, stop, and search fields are case sensitive even in Windows.

18

The Administration Center


• • • ••



Using Administration Center Portlets

The Tivoli Storage Manager Administration Center uses portlets to simplify administration and permit the Tivoli Storage Manager administrator to work on multiple tasks simultaneously. Click the X to close the portlet.

Administration Center Installation Location

A single Administration Center installation can be used for administration of many Tivoli Storage Manager servers. The Administration Center may be installed on the same machine as the Tivoli Storage Manager server if it meets the memory requirements and is being installed on one of the operating systems that the Administration Center supports. However, for availability reasons, you may want to install multiple Administration Center images. If you intend to use the capabilities of the Integrated Solutions Console for administration of additional products, you may want to consider installation on a system other than one with a Tivoli Storage Manager server.

Installing Administration Center locally may provide better performance if you intend to administer many remote Tivoli Storage Manager servers. For example, you are in Chicago and must administer Tivoli Storage Manager servers in Los Angeles, Paris, and Tokyo. There will be more network activity between the Administration Center and your browser than between the Administration Center and the Tivoli Storage Manager server.

19

Using Administration Center Portals

Select a work item on the left to work in a portlet on the right. Click

a portlet shortcut at the top to navigate between open portlets.


• • •••



Installation Requirements

The machine hosting the Administration Center and Integrated Solutions Console requires:

• Disk space

– Installation media location: 505 MB.

– 605 MB for the completed installation.

• An additional 75 MB of temporary space (which is also checked during the installation).

• Virtual memory or swap space: Equal to double the user's physical memory. At a minimum, this should be at least equal to your physical memory.

• Network adaptor.

For the latest recommendation on the Administration Center installation, read the Administration Center Readme located under Server Readmes at:

http://publib.boulder.ibm.com/infocenter/tivihelp/v1r1/index.jsp.


• • • ••



Server Console

At installation, the server console is defined with a special user ID that is named SERVER_CONSOLE. This name is reserved and cannot be used by another administrator. An administrator with system privilege can revoke or grant new privileges to the SERVER_CONSOLE user ID. However, an administrator cannot update, lock, rename, or remove the SERVER_CONSOLE user ID.

When you start or restart the Tivoli Storage Manager server, the server starts a server console session that is used to operate and administer the server until administrative clients are registered to the server. Upon starting, this server console session displays the following information:

• Product licensing and copyright information

• Processing information about the server options file

• Communication protocol information

• Database and recovery log information

• Storage Pool volume information

• Server generation date

• Progress messages and any errors encountered during server initialization

20

Server Console


• • •••



Administrative Client Options

Using Administrative Client Options

In all administrative client modes, the administrative client options modify your session responses.

Executing dsmadmc by itself or with the -ID and -PA options places you into interactive mode. You can also specify a single administrative command that will execute and then return you to the operating system prompt. This is referred to as batch mode. In the Macros and Scripts unit, you will learn to use batch mode to run macros from a system prompt.

The following options are valid from an administrative client session only:

• ID=userid—Specifies the Tivoli Storage Manager administrator’s user ID.

• PAssword=password—Specifies the Tivoli Storage Manager administrator’s password.

• NOConfirm—Requires Tivoli Storage Manager not to request confirmation before executing commands that affect the availability of the server or data managed by the server.

21

admin_commanddsmadmc

Administrative Client Options-ID=userid-Password=password-NOConfirm-Quiet-CONsolemode-MOUNTmode-Itemcommit-OUTfile=filename-COMMAdelimited-TABdelimited-DATAOnly=value-DISPLaymode=LISt or TABle


• • • ••



• Quiet—Requires Tivoli Storage Manager not to display standard output messages to your screen. When you use this option, certain error messages still appear.

• CONsolemode—In console mode, all unsolicited server console output is echoed to your screen. You cannot enter any administrator commands while Tivoli Storage Manager is running in console mode.

• MOUNTmode—In mount mode, all server removable-media mount messages are echoed to your screen. You cannot enter any administrator commands while running in mount mode.

• Itemcommit—Requires Tivoli Storage Manager to commit commands inside a script or a macro as each command is executed. With Itemcommit specified, each command is committed as script or macro runs.

• OUTfile=filename—Specifies a file to which Tivoli Storage Manager writes all output. In interactive or console modes, the output displays on your screen; in batch mode, the output is only written to the specified file.

• OUTfile—Requires Tivoli Storage Manager to write all output from a command to the standard output (usually your screen) in the same format that redirection characters (> or >>) format output in interactive mode. This option is only available in batch mode, and is primarily used to redirect output from a command.

• COMMAdelimited—Requires any tabular output from a server query to be formatted as comma-delimited strings rather than in readable format. This option is intended primarily to redirect the output of a Structured Query Language (SQL) query (select command). The comma-delimited value format is a standard data format that can be processed by many common programs, including spreadsheets, databases, and report generators.

• TABdelimited—Requires any tabular output from a server query to be formatted as tab-delimited strings rather than in readable format. This option is intended primarily to redirect the output of an SQL query (select command). The tab-delimited value format is a standard data format that can be processed by many common programs, including spreadsheets, databases, and report generators.

• DATAOnly=value—Specifies whether product version information and output headers display with the output. Acceptable values are YES or NO.

• DISPLaymode=LISt or TABle—Allows you to force the QUERY output to tabular or list format regardless of the command line window column width. If you want the output to go to a file and you are using the -DISPLaymode option, use redirection to write to the file. Do not use the -OUTfile option.


• • •••



Administrative Client Interactive Mode

Interactive mode is invoked by executing dsmadmc, after which you are prompted for your administrative ID and then your password. You are then given an Tivoli Storage Manager administrative command-line prompt at which you can enter administrative commands. After a command is executed, you are returned to the administrative command-line prompt. To terminate the session, enter quit.

The direct interactive mode is initiated by executing the administrative command-line program, dsmadmc, and passing in the -id and -password parameters. This places you at the administrative command-line prompt, at which you can enter administrative commands.

Note: Be careful passing the user ID and password in the interactive mode for accessing the administrative command line. If the operating system captures the commands and places them in a command log, the user ID and password for the Tivoli Storage Manager administrator could be compromised if the log is accessed.

22

Administrative Client Interactive Mode


• • • ••



Administrative Client Console Mode

Normally, you want to run Tivoli Storage Manager with the console secured, for example running it as a service. There are, however, situations when you would like to see the messages that are being sent to the console. Running an administrative client in console mode allows this but is also secure, because the session is read-only.

Note: Be careful using the batch mode for running commands if the operating system captures the commands and places them in a command log. This would mean the user ID and password for the Tivoli Storage Manager administrator could be compromised if access is gained to the log.

23

Administrative Client Console Mode


• • •••




• • • ••



Classroom Setup

Each student should have a standard Windows server or a standard AIX server. Each machine should have a C: drive through a K: drive for a W32 system and file systems for an RS6000 system with AIX.

A server named Server1, has been installed and set up to represent a system on-site that may not be set up according to suggested guidelines. Many of the guidelines will be provided during the course and must be tuned to each Tivoli Storage Manager server environment.

In preparation for the two units, Enterprise Administration and Server-to-server Virtual Volumes, the five required Tivoli Storage Manager servers have been installed on the student machines. These servers will have the names as listed on the slide. The Quick Start Guide contains all the setup information required to set up multiple server instances.

24

Classroom SetupPhysical Lab System

Server1 – This single TSM server is used for most of the exercises.

Server Black Server Cyan Server Grey

Servers Black, Cyan and Grey are configured for server-to-server communications and are used in the enterprise administration exercises.

Server Source Server Target

Servers Source and Target are used for the virtual volume exercises.


• • •••



Student Exercise

Open your Student Exercises book and perform the exercises for this unit.

25

Student Exercise


• • • ••



Review Questions

1. True or False: Previous versions of Tivoli Storage Manager must be uninstalled before upgrading to Tivoli Storage Manager 5.4.

2. True or False: Storage agents do not have to be upgraded in order to integrate with Tivoli Storage Manager 5.4.

3. True or False: Calendar-based scheduling is a new feature of Tivoli Storage Manager 5.4.

4. True or False: Tivoli Storage Manager for Databases is required for the operation of Informix backups directly to the Tivoli Server.

5. True or False: It is not necessary to read the readme file or check the IBM Web site to see that your devices are supported in Version 5.4 before upgrading your Tivoli Storage Manager server.


• • •••




• • • ••



Review Answers

1. True or False: Previous versions of Tivoli Storage Manager must be uninstalled before upgrading to Tivoli Storage Manager 5.4.

True.

2. True or False: Storage agents do not have to be upgraded in order to integrate with Tivoli Storage Manager 5.4.

False.

3. True or False: Calendar-based scheduling is a new feature of Tivoli Storage Manager 5.4.

True.

4. True or False: Tivoli Storage Manager for Databases is required for the operation of Informix backups directly to the Tivoli Server.

False. Data Protection for Informix provides centralized, online backup of Informix databases. The Tivoli Storage Manager for Databases component is no longer required.

5. True or False: It is not necessary to read the readme file or check the IBM Web site to see that your devices are supported in Version 5.4 before upgrading your Tivoli Storage Manager server.

False.


• • •••



Summary

26

SummaryYou should now be able to:

• Describe the changes in IBM Tivoli Storage Manager 5.4• Describe migration issues from previous versions to 5.4• Identify the functions of IBM Tivoli Storage Manager• Explain IBM Tivoli Storage Manager server components• Describe Tivoli Storage Manager 5.4 administrative functions andinterfaces

• Identify the IBM Tivoli Storage Manager servers setup for this course


• • • ••



• • • ••

2-1

Unit 2: Database and Log Manipulations

Unit 2: Database and LogManipulations


• • • ••

Unit 2: Database and Log Manipulations


Introduction

This unit reviews database concepts that set the groundwork for Tivoli Storage Manager database recovery to be covered later in the course.

Objectives

2


• Explain Tivoli Storage Manager database and recovery log concepts

• Perform Tivoli Storage Manager database backup and recovery


Unit 2: Database and Log ManipulationsPerformance and Optimization of the Database and Log


Lesson 1: Performance and Optimization of the Database and Log

3

Performance and Optimization of the Database and Log

•Database and log transactions•Database and log memory buffer pools •Space allocation•Space triggers•Database and log mirroring




Database and Recovery Log

The Tivoli Storage Manager database contains information that is needed for server operations and information about client data that has been backed up, archived, and space-managed. The database does not store client data. Instead, the database has pointers to the locations of the client files in the storage pools.

The database includes information about:

• Client nodes and administrators

• Policies and schedules

• Server settings

• Locations of client files on server storage

• Server operations (for example, activity logs and event records)

The recovery log contains information about database updates that have not yet been committed. Updates can include activities such as defining a management class, backing up a client file, and registering a client node. Changes to the database are recorded in the recovery log to maintain a consistent database image.

4

Database and Recovery Log




Transactions

A transaction is an exchange between a client and the server that accomplishes a particular action. Examples of transactions include:

• Registering an administrator or client node

• Defining a management class or copy group

• Associating a schedule with a client node

• Deleting a volume from a disk storage pool

When a transaction occurs, the server:

1. Reads a database page into the database buffer and updates it. A page is a 4096-byte block that is transferred as a unit between memory and disk storage.

2. Writes a transaction log record to the recovery log to describe the action occurring and associates it with the database page in case the page needs to be rolled back during recovery.

3. Writes the database page to the database, releasing it from the buffer pool. The page remains in the buffer pool until buffer space is needed for another page.

5

Transactions

Examples:• register node• define association• delete volume• backup-archive

Server

DBbuffers

LOGbuffers

A transaction is an exchange between a client and the server, or a server operation that is recorded in the database.

In-memorybuffer pools


• • • ••



BUFPOOLSIZE and LOGPOOLSIZE Parameters

The buffer sizes can be adjusted manually or Tivoli Storage Manager can dynamically adjust the size of the database buffer pool.

To have the server automatically tune the BUFPOOLSIZE option, set the SELFTUNEBUFPOOLSIZE option to YES. If you specify YES for the SELFTUNEBUFPOOLSIZE server option, the database buffer pool is dynamically adjusted. The cache hit ratio statistics for the buffer pool are reset at the beginning of expiration. After expiration processing completes, the buffer pool size is adjusted dynamically.

Server expiration processing resets the database buffer pool before the next processing starts and examines if the database buffer pool cache hit ratio is above 98 percent. If the cache hit ratio is lower than 98 percent, the database buffer pool will be increased; if it is higher, the buffer pool size will not change. Increasing the database buffer pool will not use more than 10 percent of available real memory.

Use the setopt BUFPOOLSIZE command to change the buffer pool size dynamically while the server is operational.

The LOGPOOLSIZE server option can be used to increase the buffer area for the recovery log. A large buffer pool may increase the rate by which recovery log transactions are committed to the database, but it also requires more memory.

6

BUFPOOLSIZE and LOGPOOLSIZE Parameters

Options that configure database buffer pool and log pool sizes are:

• BUFPOOLSIZE – The minimum size is 256 KB.– The maximum size is limited by available virtual memory

(the new default is 32756 KB).– Suggested: 131072 for a server with 1 GB real memory.

• LOGPOOLSIZE– The minimum size is 128 KB.– The maximum size is limited by available virtual memory

(the new default is 512 KB).

• SELFTUNEBUFPOOLSIZE

–Increasing the database buffer pool will not use more than 10 percent of available real memory.


• • •••



Recovery Log Transaction Modes

The recovery log is used by the server to keep a record of all changes to the database. When a change occurs, the recovery log is updated with some transaction information prior to the database being updated. This enables uncommitted transactions to be rolled back during recovery, so the database remains consistent.

The recovery log functions in two modes—normal mode and roll-forward mode. Use the set logmode command to specify which mode. The default is normal.

Normal Mode

When the transaction log record is written to the recovery log, a recovery point is recorded in it, and the data is committed to the database. If the database needs to be recovered, the server uses the recovery point in the recovery log to bring the database back to its last point of consistency. A point of consistency is a time when all recoverable information in the database matches the data managed by the server. If a failure occurs before a transaction is committed to the database, the server rolls back any changes made to the database pages. The log is treated as a circular array of blocks with the head (the newest log records) always chasing the tail (oldest records). The server will never let the head overtake and overwrite the tail; it must take some other action. As transactions commit, they free up log space and allow the tail to move forward. The recovery log saves some records for transactions that have already been committed, but only to the extent necessary to perform redo processing.

7

Recovery Log Transaction Modes

Normal Mode: New transaction overwrites old.

Roll-Forward Mode: Logs all transactions between database backups.


• • • ••



Roll-Forward Mode

In roll-forward mode all changes made to the database since the last backup are saved in the recovery log. All changes are cleared when the database is backed up.


• • •••



Recovery Log Consumption

The number of transactions between database backups determines how large the recovery log should be.

In roll-forward log mode, the recovery log could be significantly larger than in normal log mode. With the recovery log in roll forward mode, the transactions are cached after they have been committed to the database.Tivoli Storage Manager deletes cached transactions only after a successful database backup.

In normal log mode, you need less space for the recovery log since Tivoli Storage Manager deletes transactions once they have been committed to the database.

If you want to gather statistics about the recovery log consumption, you can reset the cumulative log consumption by using the command reset logconsumption. You can view the value by using the command q log f=d.

8

Recovery Log Consumption

backup db

Cached transactions

Cached transactions are cleared after successful database backup


• • • ••



Database and Recovery Log Mirroring

The availability of the database and recovery log is ensured, by providing protection from logical errors and from device or media failures.

Mirroring offers additional flexibility to that of using hardware dual copy. Mirroring has the following benefits:

• Three copies can be supported.

• You have finer control over what is mirrored.

• Copies may be on physical volumes that reside on different storage control units.

• Support for multiple and different direct access storage device (DASD) device types, including 9340.

Read performance improves with mirroring because the server will read from the device with the best response time.

9



• • •••

Unit 2: Database and Log ManipulationsDatabase and Recovery Log Space Allocation


Lesson 2: Database and Recovery Log Space Allocation

How Many Database Volumes Are Needed?

In general, database access is predominately read-oriented. Placing the database volumes over multiple physical volumes may improve performance because this allows the logical volume manager to spread the I/O load over more volumes. This is also true when mirroring is used because the logical volume manager schedules read operations to the least busy volume in a mirror set. However, keep the number of volumes reasonable (less than 12), to reduce space used for logical volume manager overhead. The maximum database size limit is 530 GB. Although the maximum size of the database is quite large, it is suggested that the database size be kept to a manageable size for the environment.

10

Database and Recovery Log Space Allocation

• Database:• Predominately read-oriented.• Spread database across disks for better performance.

• Maximum size limit is 530 GB.

• Recovery log:• Predominately write-oriented.• No performance gain from spreading recovery log across disks.

• Log file maximum size limit is 13 GB.

• Allocate DB and Log on separate volumes.


• • • ••



How Many Recovery Log Volumes Are Needed?

In general, access to the recovery log is predominately write-oriented with the writes and the few reads clustered together for the most part. The writes are done in a moving cursor format, which does not lend itself to multiple volume optimizations. Therefore, fewer recovery log volumes are appropriate. Mirroring has little effect on the performance of the recovery log. The maximum log file size limit is 13 GB.


• • •••



Database and Recovery Log Space

Volumes used to contain the database and the recovery log must be disk volumes.

Tivoli Storage Manager treats all volumes associated with the database or with the recovery log as a single logical volume. The logical volume manager maps data between logical and physical storage, allowing database and recovery log data to span physical disks. Reorganization of the database or recovery log is not required after adding a volume.

The amount of available space for the database or recovery log equals the combined space of all volumes defined to the database or recovery log. As data is added, Tivoli Storage Manager tracks the percentage of utilization, which is the amount of space used at a specific point in time. Be aware that the maximum amount of space used by the recovery log can vary significantly throughout the day, as it is proportional to the transaction load on the system. The maximum amount of space used by the database is more consistent with the utilization percentage, because the amount of database space consumed grows in proportion to the number of objects inserted into the database.

11Database and Recovery Log Space Considerations

All volumes treated as a single logical volume.

available space

assigned capacitymaximumextension

maximum % usedmaximum reduction

% utilization


• • • ••



Increase or Decrease the Size

Monitoring the database and log regularly will help you determine if you need to increase or decrease space. You might want to reset the maximum utilization counters each day. Utilization statistics are reset in 2 ways:

• Automatically when the server is started

• By issuing the reset dbmaxutilization or reset logmaxutilization commands

To display information about the database or recovery log, issue the q db or the q log command. As your requirements change, you can manually increase or decrease the size of the database or recovery log. To change the assigned capacity of the available space, issue the extend db megabytes, reduce db megabytes, extend log megabytes, or reduce log megabytes command, where megabytes is the value, in 4 megabytes increments, of the space to be increased or decreased.

These commands take advantage of the available space, they do not delete or add volumes to the space. If you need more space, and your maximum extention is exhausted, you will need to add more volumes.

12

Increase or Decrease the Size


• • •••



Space Trigger

A space trigger can be defined for both the database and recovery log with the define spacetrigger command. When the database or recovery log reaches a defined utilization threshold, Tivoli Storage Manager can automatically extend the database or recovery log into any area of the database that has available space. In the case of UNIX and Windows, a space trigger can allow additional space to be allocated.

The amount of space allocated is based on a defined percentage of the existing database or recovery log space. If any volumes are mirrored, the newly allocated space is also mirrored. Database Management

13

Automatic expansion of the database or recovery logwhen the trigger is reached

The Space Trigger

threshold80%

20% more

db or log space

20% more

mirrored space


• • • ••



Reorganizing the Database


Periodically reorganizing the server’s database can improve performance. Use the estimate dbreorgstats command to determine the estimated space saved by reorganization of the database. The process can be queried to monitor the command’s progress and, when finished, the results can be viewed by querying the database in detailed format.

To see the results of the command query the database in detailed format. These two lines from the query db f=d command provide an example of the details of recoverable database space by reorganizing after using the estimate dbreorgstats command:

Estimate of Recoverable Space (MB): 81Last Estimate of Recoverable Space (MB): 03/26/2007 16:06:35

Over time, database volumes become fragmented. You can restore the efficiency of the database and improve database performance by reorganizing the database using database unload and reload processing. By reloading the database, you compress and reorganize it.

The procedure includes unloading the database, formatting database volumes and recovery log volumes to prepare for loading, and then loading the database. The operations read device information from the device configuration file, not from the server’s database.

14


dsmserv unloaddbdsmserv loaddb


• • •••



You can use a device class of FILE for the DSMSERV UNLOADDB and DSMSERV LOADDB operations. If you use any other type of device class for the operations, you must use a drive that is assigned to a manual library (library type of MANUAL). If the drive that you want to use is not assigned to a manual library, you must edit the device configuration file to temporarily change the definition so that it appears to be in a manual library. Refer to the Administrators Guide for Unload and Reload procedure instructions.


• • • ••



Using the DSMSERV Command

Administrative commands can be issued in an interactive mode using the dsmserv command. These commands are typically used in an emergency to correct a situation with the server halted.

dsmserv

-k specifies the name of the Windows registry key from which to retrieve information about the server. The default is server1.

runfile file_name specifies the name of a text file to be run on the server that contains a list of server commands.

upgradedb specifies that the server database is to be upgraded.

auditdb resolves inconsistencies in the database. Contact Tivoli Storage Manager Software Support before issuing this command if you are using this command in situations other than when a message indicates it is necessary.

display dbbackupvolume displays database backup volumes when volume history information is not available.

display dbvolumes displays information about all database volumes.

display logvolumes displays information about recovery log volumes.

15

Using the DSMSERV Command

dsmserv-krunfileupgrade dbauditdbdisplay dbbackupvolumedisplay dbvolumesdisplay logvolumesextend logformatrestore dbdumpdbunloaddbloaddbcocorrupted recovery log


• • •••



extend log extends the size of the recovery log.

format initializes the database and recovery log. No other server activity is allowed while initializing the database and recovery log.

restore db restores the database using previously created database backups.

dumpdb is used as the first step in dumping, reinitializing, and reloading the database if a catastrophic error occurs (recovery log corruption, for example). No other server activity is allowed during the processing of this command.

unload reorganizes the contents of the database when the server is offline.

loaddb reloads a dumped database. No other activity is allowed while reloading the database.

Exercise caution, and consult the Tivoli Storage Manager Administrator’s Guide before using these commands.


• • • ••

Unit 2: Database and Log ManipulationsTivoli Storage Manager Server Backup Facilities


Lesson 3: Tivoli Storage Manager Server Backup Facilities

Objectives of a Tivoli Storage Manager server database backup include:

• Back up without interruption of services

• Multiple options to initiate a DB backup

• Disaster recovery support

• Point-in-time restores

• Individual database volume restores

• Database snapshot backup

• System object support

16

• Database backup options

• Database backup commands

• Database recovery

• Restore commands

• Considerations

Tivoli Storage Manager Server Backup Facilities

RecoveryLog

TSM server

The database and log can be mirrored to protect data in case of media failure. Storage pool copies can be kept on-site for recovery, and off-site for disaster recovery.

Database

Mirror Recovery

Log

Mirror Database




Tivoli Storage Manager Database Backup

Tivoli Storage Manager can perform regular and incremental database backups to tape while the server is running and available to clients. With the Tivoli Storage Manager recovery log in normal mode, the backup media can then be stored on-site or off-site and can be used to recover the database up to the point of the backup. You can run regular or incremental backups as often as necessary to ensure that the database can be restored to an acceptable point in time.

You can provide even more complete protection if you specify that Tivoli Storage Manager run in roll-forward mode. With Tivoli Storage Manager in roll-forward mode and with an intact recovery log, you can recover the database up to its most current status.

Regular Only or Regular-plus-incremental Backups

Administrators may choose to run regular or incremental backups or both, whichever combination best suits their business needs. The option to choose one or both methods helps administrators fit the backups into the time frames allotted.

A regular backup takes longer to run than an incremental backup because it copies the entire database. However, recovery time is faster with a regular backup because only one set of volumes needs to be loaded to restore the entire database. A regular backup is required

17

Tivoli Storage Manager Database Backup

Types of database backupsFull backup

Full backup required when– Database has never been backed up.– Maximum incremental backups are reached.– Database has been extended or reduced.– Recovery log mode changed to roll-forward (default = normal).

Incremental backups (default)Database out-of band (snapshot) backup

+ +1 full

Database backup series Up to 32 incremental backups

TSM allows for online database backups.

TSM DB


• • • ••



under specific conditions, but an administrator can choose to run as many as 32 incremental backups between regular backups.

An incremental backup takes less time to run because it copies only those database pages that have changed since the last time the database was backed up. However, incremental backups increase the time it takes to recover a database, because a regular backup must be loaded first followed by some or all of the incremental backups in the same database backup series.

Snapshot Backup

A snapshot backup is a full backup that does not interrupt the full-plus-incremental backup series. In other words, it is an out-of-band database backup. This backup can be carried off-site for disaster recovery purposes.

A database snapshot backup is tracked by the volume history and can be used for a restore of the Tivoli Storage Manager database to the point in time when the snapshot was performed.


• • •••



Database Backup Command

Use the BACKUP DB command to back up the database.

Parameter Description

DEVclass=devclassname

Specifies the name of the sequential access device class to use for the backup. Be sure that you have used the DEVCONFIG option in the dsmserv.opt file to specify an external file in which to store a backup copy of device class definitions. If you do not have this file and your Tivoli Storage Manager database is damaged or lost and must be restored, the definitions created by using the DEFINE DEVCLASS command will not be available and must be recreated manually. This parameter is required.

Type=typevalueSpecifies the type of backup to run. This parameter is optional. The default value is INCREMENTAL.

18

Database Backup Command

BAckup DB DEVclass=devclassnameType=Incremental / Full / DBSnapshotVOLumenames=volname(s)Scratch=Yes/No Wait=No/Yes

devconfig

vol history

DB

Storage Managerserver recovery log


• • • ••



Backup Commands for Database Related Files

Use the backup devconfig command to specify one or more files in which to store a backup copy of the following definitions:

• Device class definitions created using the define devclass command

• Library definitions created using the define library command

• Drive definitions created using the define drive command

VOLumenames=volname

Specifies the volumes to use for the backup. You can specify more than one volume by separating each volume name with a comma, with no intervening spaces.

Scratch=scratchvalueSpecifies whether scratch volumes can be used for the backup. This parameter is optional. The default value is YES.

Wait=waitvalue

Specifies whether to wait for the server to complete processing this command in the foreground. The default value is NO. The first backup of your database must be a regular backup. You can run up to 32 incremental backups between regular backups. To perform a regular backup of your database to the TAPECLASS device class, for example, enter:

backup db type=full devclass=tapeclass

In this example, Tivoli Storage Manager writes the backup data to scratch volumes. You can also specify volumes by name. After a regular backup, you can perform incremental backups, which copy only the changes to the database since the previous backup. Run an incremental backup of the database, using a scratch volume. Assume a device class of FILE for the backup:

backup db devclass=file type=incremental



• • •••



Use the backup volhistory command to request that the server save sequential volume history information to one or more files. Volume history information includes data such as date and time of use for the following volume types:

• Volumes used to back up a Tivoli Storage Manager database (regular or incremental)

• Volumes used to dump a Tivoli Storage Manager database


• Sequential access volumes that have been added to storage pools

• Sequential access volumes that have been reused in reclamation of storage pools or move data operations

• Sequential access volumes that have been removed using the delete volume command or removed during reclamation of scratch volumes


• • • ••



Database Backup-related Commands

query db

To help you determine how much storage space a regular or incremental backup will require, use the query database command. This command displays the number of changed megabytes in the database.

query volhistory

Use the query volhistory command to display sequential volume history information that has been collected by the server. Volume history information includes data such as date and time of use for the following types of volumes:

• Volumes not in storage pools

• Volumes used to back up a Tivoli Storage Manager database (regular or incremental)

• Volumes used to dump a Tivoli Storage Manager database

19Tivoli Storage Manager DB Backup-related Commands

query volhistoryDate/Time: 04/22/2007 13:04:01

Volume Type: BACKUPFULL

Backup Series: 3

Backup Operation: 0

Volume Seq: 1

Device Class: TESTBUDB

Volume Name: D:\tsm\server\9326461.dbb

delete volhistory todate=today–7 type=dbbackup

query db format = detail

Pct. Util: xx%

Max Pct. Util: xx%

Incremental since last full: x

Changed since last backup (MB): xxx


• • •••




• Backup sets

Syntax

query volhistory type=typevalue

Typevalue specifies the type of records to display from the volume history file. This parameter is optional. The possible values for it are listed in the following table; the default value is ALL.

delete volume history

Use the delete volhistory command to delete sequential volume history information collected by the server when that information is no longer needed. For example, you may want to delete information about volumes used for obsolete database backups.

When volume history information about volumes not in storage pools is deleted, the volumes return to scratch status if they were acquired by Tivoli Storage Manager as scratch volumes. For scratch volumes with device type file, the files are deleted.

Do not delete sequential volume history information until you no longer need it. Do not delete the volume history information for database dump, database backup, or export volumes that reside in automated libraries unless you want to return the volumes to scratch

Typevalue Description

All Specifies that all history information is displayed.

DBBackupSpecifies that only records containing information about volumes used for backups (regular or incremental) are to be displayed.

DBSnapshotDatabase out-of band (snapshot) backup is a full backup that does not interrupt the current full+incremental backup series.

DBDumpSpecifies that only records containing information about volumes used for online database dumps (done before installation of Version 2.1) are to be displayed.

EXPortSpecifies that only records containing information about volumes used for exporting data are to be displayed.


• • • ••



status. When the delete volhistory command removes volume information for database dump, database backup, or export volumes, the volumes are automatically returned to scratch status if they reside in automated libraries. These volumes are then available for reuse by the server, and the information stored on them may be overwritten when the server reuses the volume for some other purpose, such as storage pool volumes or other database backups. When determining how many days you want to keep the volumes before deleting them, you should also consider the reusedelay parameter value that is set on the storage pool of which the volume is a member. Define or update the storage pool so that the reusedelay parameter value is equal to or greater than the number of days you want to keep the volumes before returning them to scratch status.

For users of Disaster Recovery Manager, the database backup expiration should be controlled with the set drmdbbackupexpiredays command instead of the delete volhistory command. Using the delete volhistory command can cause volumes to be lost that were managed by the move drmedia command.


• • •••

Unit 2: Database and Log ManipulationsDatabase Recovery


Lesson 4: Database Recovery

Tivoli Storage Manager provides three types of recovery, all initiated with the dsmserv restore db command:

• Point-in-time recovery

• Roll-forward recovery to most recent state

• Single database volume recovery

20

Database Recovery

Point-in-time recovery

Recovery to most current state

Single database volume recovery

1 full Up to 32 incrementals



Log roll-forward

Log roll-forward

dsmserv restore db

X




Point-in-time Database Recovery

Point-in-time recovery is normally used for exceptional situations such as disaster recovery or removing errors causing an inconsistency in the database. If an administrator enables roll-forward recovery, point-in-time recovery is also available.

For disaster recovery purposes, an administrator can run regular or incremental backups as often as necessary to insure that a Tivoli Storage Manager database can be restored to an acceptable point in time, and can send those backups to an off-site location for safekeeping. It is very important to perform database backup with storage pool backup to provide disaster recovery services for the server.

To restore a database to a specific point in time to remove an error causing an inconsistency in a database, an administrator can use some or all of the backups contained in a database backup series.

An administrator can also choose to enable point-in-time recovery only. Take the following considerations into account when making this choice:

• When only point-in-time recovery is enabled, the recovery log can be defined with a smaller size, and mirrored copies of the recovery log consume less disk space.

• Because there is no requirement for auditing the database after a database restore, recovery time is much faster compared to the database salvage utility.

21

Point-in-time Database Recovery

+ Smaller size for recovery log+ No auditing of database required after restore

– Cannot restore to most recent state– Single database volume cannot be restored– Auditing of storage pools required after

restore


Point-in-time recovery




• The database cannot be restored to its most current state. It can only be restored to the point in time when the particular backup series used to restore the database was created.

• Because restoring a single database volume requires roll-forward recovery, the entire database must be restored even if only one volume is damaged.

• After the database has been restored, you must audit the sequential storage pool volumes based on the information in the volume history file. You also must audit the disk storage pool volumes.

Note: It is suggested that you mirror the recovery log even if you only use point-in-time recovery.




Roll-forward Database Recovery

Roll-forward recovery to most recent state is only possible if the recovery log is available. Together, mirroring and roll-forward recovery provide the most comprehensive protection for a Tivoli Storage Manager database. Considerations when enabling roll-forward recovery include the following:

• Restoring the database to its most current state allows an installation to protect against loss of client files due to a hardware failure on the Tivoli Storage Manager server.

• Because there is no requirement for auditing the database and storage pools after a database restore, the recovery time is much faster.

• If there is a media failure on a database volume only that volume needs to be restored, which reduces Tivoli Storage Manager database recovery time.

• The recovery log must be defined with a larger size, because Tivoli Storage Manager must save additional recovery log records. The size of the recovery log depends on the frequency of required backups, how the database backup trigger is set, and the volume of Tivoli Storage Manager transactions.

• Mirrored copies of the recovery log consume additional disk space because of the larger recovery log size. It is suggested you mirror your recovery log if you have enough disk space.

22

Roll-forward Database Recovery

+ Database can be restored to most recent state.+ No auditing of database and storage pools required

after database restored.+ Single database volume can be restored.– Larger recovery log.– Mirrored recovery log consumes additional disk space.

Roll-forward recovery to most current state

Mirroring + Roll-forward = Comprehensive Protection

1 full Up to 32 incrementals Log roll-forward




• It is also suggested that you mirror your database.

Information Needed for Restore

23

Preparation for Database Restore

volume history files

device configuration

To prepare for database recovery:

1. Find the name of the database volume

2. Find the size of the database volume

3. volhist

4. devconfig

5. restore

TSM database

backup volhist

backup devcnfg

Current copies of the following items enable you to restore your server.

• volhist.out

• devcnfg.out

• dsmserv.opt

• dsmserv.dsk

• Database and storage pool backup




Volume-related Information in the Tivoli Storage Manager Database

Volume history information is automatically collected by the server for use in performing database recovery and maintained in the Tivoli Storage Manager database. The Tivoli Storage Manager server collects the following volume history information volume types:

• Volumes used to back up a Tivoli Storage Manager database


• Sequential access storage pool volumes that have been reused, added, or deleted

Device-related Information in the Tivoli Storage Manager Database

Device, drive, and library definitions are stored in the Tivoli Storage Manager database.

24

Examples of Database Restore Commands

Restore to Current State for:• Database:

dsmserv restore db preview=no|yes

• Single database volume:dsmserv restore db dbvolume=volname preview=no|yes

Restore Database to Specific Point in Time When:• History file is available:

dsmserv restore db todate=date totime=time source=dbbackup|dbsnapshotpreview=no|yes

• History files are not available:dsmserv restore db devclass=devclassname volumename=volname

commit=no|yes




Device configuration information consists of device class definitions created using the define devclass command. On an AIX server, it also includes drive and library definitions created using the define drive and define library commands.

Examples of Restore Commands

dsmserv restore db Utility

The dsmserv restore db utility uses backup versions of either of the following database backups:

• The database you create using the backup db command.

• Backup versions created automatically by Tivoli Storage Manager based on the settings you specify with the define dbbackuptrigger command.

Roll-forward Recovery

You can perform a roll-forward recovery to most current state for an entire database, if the following conditions are met:

• The log mode was set to rollforward continuously, from the time that the last backup series was created until the time that the database was damaged or lost.

• The recovery log is available and intact.

• A volume history file is available and intact.

You can perform a roll-forward recovery for a single database volume if both the following conditions are met:

• The log mode was set to rollforward continuously, from the time that the last backup series was created until the time that the database volume was damaged or lost.

• The recovery log is available and intact.




Database Restore and Volume History

If a backup copy of volume history information is available, you can restore a database to a specific point in time. The date to which to restore the database is specified using the todate=date parameter. Possible values for this parameter are:

• mm/dd/yyyy —Specifies that you want to restore a database using the most recent backup series created prior to a specific date.

• today—Specifies that you want to restore a database using the most recent backup series created prior to the current date.

If the volume history file is unavailable, you can still restore a Tivoli Storage Manager database to a specific point in time by using one or more dsmserv restore db utilities. For example, if you need to load a regular backup and one or more incremental backups, you can issue a dsmserv restore db command to restore the regular backup, then issue an additional dsmserv restore db command for each incremental backup.

When you use multiple dsmserv restore db utilities, you must specify commit=no for each command except the last one you issue. For the last dsmserv restore db command, you must specify commit=yes to place the database in a consistent and usable state.

If the volume history file is not available you need to use the VOLUMENAME option to tell restore db which volumes to restore. This parameter specifies the names of one or more volumes to use to restore the database. You can specify more than one backup volume by separating each volume name with a comma, with no intervening spaces. You must list the volumes in the order in which the backup was stored. For example, if you want to use the backup volumes tape01, tape02, and tape03, specify:

volumename=tape01,tape02,tape03

define dbbackuptrigger Command Syntax


devclass=devclassname

Specifies the name of the sequential access device class to use for backups. This parameter is required. If the mount limit for the device class specified has already been reached when the backup runs (all drives are busy), Tivoli Storage Manager automatically cancels lower-priority operations, such as reclamation, to make a mount point available for the backup.


• • •••



Example

Define the setting of the database backup trigger at 80 percent. Run two incremental backups to every regular backup of the database. Assume a device class of file for the backup. Use the following command.

define dbbackuptrigger devclass=file logfullpct=80 numincremental=2

Run the query status command to view the mode the recovery log is in:

• Normal—Specifies that Tivoli Storage Manager is to save only those recovery log records that are needed to recover the database in case of a system failure. This is the default for the log mode.

• Roll-forward—Specifies that Tivoli Storage Manager is to save enough log records to ensure that the database can be restored to its most current state using the most recent database backup series and recovery log records.

logfullpct=percentage

Specifies when Tivoli Storage Manager automatically backs up the database, based on the percentage of available space used for the recovery log. When the percentage of space used for the recovery log exceeds this value, Tivoli Storage Manager performs an automatic backup of the database and removes unnecessary records from the recovery log.

numincremental=number

Specifies the maximum number of incremental backups that can be run before a regular backup is required. You can specify a value from 0 to 32. 0 specifies that Tivoli Storage Manager runs only regular backups automatically. This parameter is optional.



• • • ••

Unit 2: Database and Log ManipulationsSuggestions for Database Management


Lesson 5: Suggestions for Database Management

• In order to restore the database, you must have the following items:

– A full database backup

– Any incremental database backups between the last full backup and the point in time to which you are recovering

– Copies of storage pool volumes

– Server options file

– Volume history file

– Device configuration file

– Database and recovery log setup (the output from detailed queries of your database and recovery log volumes)

• Before your first backup, you must do some or all of the following steps:

– Define device classes for backups

– Set the recovery log mode

25

Suggestions for Database Management

• Keep current copies of these files off-site to ensure your ability to recover the server efficiently:

volhist.out, devcnfg.out, dsmserv.opt, dsmserv.dsk• Before the first backup, there are several tasks that need to be performed to

configure the environment.

• Performance can be increased by:• Defining more than one database and recovery log volume—allows simultaneous access• Allocating a few large disk volumes instead of many small ones—minimizes overhead

requirements

• For better reliability allocate:• Database and recovery log volumes in an NTFS file system, not FAT• Database and recovery logs on disk drives that reside on the TSM server machine, not on a

remotely mounted file system

• Mirror the database and recovery log to ensure availability.• You can provide even more complete protection if you use roll-forward mode.


• • •••



– Schedule database backups

– Estimate the recovery log size

– Automate database backups to occur according to a defined schedule or when the recovery log utilization reaches a specified percentage

• Improve performance and reliability by doing the following tasks:

– Define more than one volume for the database and for the recovery log and keep them on separate disks. This enables simultaneous access to different parts of the database or recovery log.

– Define and use a few large disk volumes rather than many small ones. This saves space by minimizing overhead.

– Allocate database and recovery log volumes in an NTFS file system, not a FAT file system. NTFS takes advantage of Windows ability to recover from problems that can occur during I/O to a disk.

– Allocate database and recovery log volumes on disk drives that reside on the Tivoli Storage Manager server machine, not on remotely mounted file systems. Doing so can also avoid database and recovery log corruption.

• Mirror the database and recovery log volumes to protect data from loss due to media failure. By separating volume copies on different physical devices, you protect the server from media failure and increase the availability of the database and recovery log. To create mirrors of the database and recovery log you must do the following tasks:

– Allocate disk volumes to mirror the database and recovery log

– Define mirrored volume copies of the database and recovery log

– Specify server options for mirroring and database page shadowing

– Request information about mirrored volumes

If you cannot allocate each volume copy to its own physical disk, use the method shown below:

Physical Disk Database Volume Recovery Log Volume

Physical Disk 1 Database Volume Copy 1 Recovery Log Volume Copy 1




• • • ••



• Use roll-forward mode if you have enough space, as it offers the greatest protection for your data. However, there are costs to roll-forward mode. If the log mode is set to roll-forward after a point-in-time database restore, a database backup starts when the server is brought up for the first time. This can cause loss of data: a tape can have current data on it, but because of the point-in-time restoration, it can be marked as scratch. When the server starts for the first time Tivoli Storage Manager may use this tape to write the database backup, thus destroying the original data on this tape.


• • •••



Student Exercise


26

Student Exercise


• • • ••



Review Questions

1. The roll-forward recovery log mode provides:

a. An extra level of database security

b. The ability to restore a Tivoli Storage Manager database to the last transaction

c. Quicker and easier restoration of client data

d. Backup of Tivoli Storage Manager database files

2. Database backups should occur (choose three):

a. Before storage pool migration

b. After activities that change the database

c. Daily after backup stgpool and migration operations

d. After all client backups have been completed

3. True or False: There can be 64 incremental database backups between full database backups.

4. True or False: If the Tivoli Storage Manager database is corrupted, the server must be rebuilt.

5. Which is the best sequence of commands to use in creating off-site backups?

a. Back up clients, volume history, database, storage pools

b. Back up clients, storage pools, database, volume history

c. Back up volume history, clients, database, storage pools

d. Back up volume history, database, storage pools, clients

6. True or False: Both database backups and database snapshots can be used to restore to the current state.


• • •••



Review answers begin on the next page.


• • • ••



Review Answers

1. The roll-forward recovery log mode provides:

b. The ability to recover a Tivoli Storage Manager database to the last transaction

2. Database backups should occur (choose three):

b. After activities that change the databasec. Daily after backup stgpool and migration operationsd. After all client backups have been completed

3. True or False: There can be 64 incremental database backups between full database backups.

False, only 32

4. True or False: If the Tivoli Storage Manager database is corrupted, the server must be rebuilt.

False

5. Which is the best sequence of commands to use in creating off-site backups?

b. Back up clients, storage pools, database, volume history

6. True or False: Both database backups and database snapshots can be used to restore to the current state.

False. Only database backups can be used in conjunction with the recovery log to recover to the current state. Database snapshots are designed for disaster recovery.


• • •••



Summary

27

Summary

You should now be able to:• Explain Tivoli Storage Manager database and recovery log concepts

• Perform Tivoli Storage Manager database backup and recovery




• • • ••

3-1

Unit 3: Storage Pool Manipulations





Introduction

This unit reviews storage concepts for Tivoli Storage Manager. It also introduces active-data storage pools, and the concept of data shredding.

Objectives

2


• Describe storage pool concepts• Perform Tivoli Storage Manager storage pool backup and recovery


Unit 3: Storage Pool ManipulationsStorage Pool Manipulations


Lesson 1: Storage Pool Manipulations

Data Storage Pools

The server stores backed up and archived files in data storage pools. The database serves as the index of client files within data storage. Data storage is defined as a collection of storage pools.

Data storage may be composed of optical media, direct access storage, and sequential tape media. Files may initially be placed on different storage pools according to the desired storage management policy. They are automatically moved to other devices to satisfy requirements for space utilization, performance, and recovery.

Data storage can be managed by an administrator with system, storage, or operator privileges. Management of data storage includes planning, preparing, monitoring, and deleting storage volumes and storage pools, depending on the level of privilege.

3

Data Storage Pools

Hierarchy of server storage holding client files

Systemstorageoperator

Administrator

backup data backup data

migration





A storage pool is a named set of volumes, belonging to the same device class, that is the destination of backed up or archived data. A device class is a grouping of like devices, such as disk or tape.

The purpose of storage pools is to match user requirements for data with the physical characteristics of storage devices. For example, if users need immediate access to certain data, you can define a storage pool consisting of storage volumes residing on a high performance direct access storage device (DASD). Another possibility would be to define a storage pool consisting of storage volumes residing on a storage area network, or SAN. Users can then use this storage pool as a destination for their files by binding the appropriate management class.

4


Storage Pools• Collections of like media

(same device class)

• Storage for backed up, archived, migrated files

• Variety of devices supported

• Chained to create storage hierarchy

UNIX NTOS/390

Application Server

Backup Server

FC FC

FC

Storage Area Network (SAN)

100 MB/s

LAN/WAN

TCP/IP

FC FC


• • •••



Storage Area Network Tape Library Sharing

When Tivoli Storage Manager servers share a storage device, one server—the library manager—controls device operations. These operations include mount, dismount, volume ownership and library inventory. Other servers and other library clients use server-to-server communications to contact the library manager and request device service. Data moves over the SAN between the servers and the storage device.

When sharing an automated library device, Tivoli Storage Manager servers use features discussed in the following subsections.

Partitioning of Volume Inventory

The inventory of media volumes in the shared library device is partitioned among servers. Either one Tivoli Storage Manager server owns a particular volume, or the volume is in the global scratch pool. No one server owns the scratch pool at any given time.

5Storage Area Network Tape Library Sharing

Storage Area Network (SAN) ConfigurationThe servers communicate over the LAN. The library manager controls the library over the SAN. The library client stores data in the library devices over the SAN.


• • • ••



Serialized Drive Access

Only one Tivoli Storage Manager server accesses each tape drive at a time. Drive access is serialized and controlled so that servers do not dismount other servers’ volumes or write to drives where other servers mount their volumes.

Serialized Mount Access

The library device’s autochanger performs a single mount or dismount operation at a time. A single server (the library manager) performs all mount operations to provide this serialization.

SAN data movement by a client requires the installation of a storage agent on the client machine. The Tivoli Storage Manager server maintains the Tivoli Storage Manager database and recovery log and controls device operations. The storage agent on the client handles the data transfer to the device on the SAN. This implementation frees up bandwidth on the LAN that would otherwise be used for client data movement.

In a typical backup scenario for a Tivoli Storage Manager client using SAN data movement

1. The client begins a backup operation. The client and server exchange policy information over the LAN to determine the destination of the backed up data.

6

Storage Area Network Data MovementTSM clientstorage agent installed

TSM serverLibrary controlclient metadata

LAN

SAN

Tape library

Client data Library

control


• • •••



2. For a client using LAN-free data movement, the destination is a storage pool that uses a device on the SAN. That device must also be mapped for the client.

3. Because the destination is on the SAN, the client contacts the storage agent, which handles the data transfer. The storage agent sends a request for a volume mount to the server.

4. The server contacts the storage device and mounts the appropriate media.

5. The server notifies the client of the location of the mounted media.

6. The client, through the storage agent, writes the backup data directly to the device over the SAN.

7. The storage agent sends file attribute information to the Tivoli Storage Manager server, and the server stores the information in its database.


• • • ••



Storage Destinations

A storage destination is the storage pool where client data is sent when it is backed up or archived. It is specified in the backup and archive copy group definitions included in the management class. Data placement is also influenced by storage pool settings, which may restrict read and write access, and limit the size of files placed in the storage pool.

7

Storage Destination


• • •••



Migration for Disk Storage Pools

Migration processing maintains free space in a primary storage pool. The server can move data stored in a random access disk storage pool to a slower but less expensive sequential access storage pool.

When you define or update a storage pool, you can set migration thresholds to specify when the server should start and stop migrating data to the next storage pool in the storage hierarchy. Migration thresholds are defined in terms of a percentage of total data capacity for the disk storage pool. The server starts a migration process when the storage pool has been filled to the capacity defined by the high-migration threshold. The migration process continues to run until the low-migration threshold has been reached.

The default high-migration threshold value is 90%, and the default low-migration threshold value is 70%.

8

Migration for Disk Storage Pools

STORAGE POOL

HIGH MIGRATION

LOW MIGRATION

90%

70%

NEXT STORAGE POOL


• • • ••



Migration Threshold Values

Setting migration thresholds for disk storage pools ensures sufficient free space on faster-speed devices, which can lead to better performance. Choosing thresholds appropriate for your situation takes some experimenting. You need to ensure that migration occurs frequently enough to maintain some free space but not so frequently that the device is unavailable for other use.

When choosing the high-migration threshold, consider the amount of storage capacity provided for the storage pool, and the amount of free storage needed to store additional files without having migration occur.

Keeping the high-migration threshold at a single value means that migration processing could start at any time of day, whenever that threshold is exceeded.

9

Migration Threshold Values

STORAGE POOL

NEXT STORAGE POOL

HIGH MIGRATION

LOW MIGRATION

90%

70%


• • •••



Manually Starting Migration

You can control when migration occurs by using administrative commands or schedules to change the thresholds. You can set the high-migration threshold high during the night when clients run their backup operations, and lower the high-migration threshold during the day when you want migration to occur. Choose a time when your tape drives and mount operations are available for the operation. Use the UPDate STGpool command to change the high-migration and low-migration threshold values.

You can specify an integer from 0 to 100 for the high-migration threshold. You can specify an integer from 0 to 99 for the low-migration threshold.

Multiple concurrent processes let you make better use of your available tape drives. Specify the number of processes to be used for migrating files from the storage pool with the MIGPRocess optional parameter in the DEFine STGpool or the UPDate STGpool command.

10

Manually Starting Migration

fileA.ext

LOW MIGRATION

60%

20%

HIGH MIGRATION

fileB.ext

fileD.ext

fileC.ext

UPDate STGpool STORAGEPOOL HIghmig=60 LOmig=20

STORAGE POOL

NEXT STORAGE POOL


• • • ••



Ending Migration

Migration will continue to run until it reaches the value set on the low-migration threshold. The smaller the low threshold, the longer time that a migration runs. Set the low-migration threshold to 0% if you want all the data to migrate to the next storage pool.

After the migration processing completes, reset the thresholds back to the desired values for the nightly backup processing.

Note: The command cancel process will not end migration processing unless the thresholds have been reset to a value that will allow migration processing to end.

11

Ending Migration

STORAGE POOL

NEXT STORAGE POOL

HIGH MIGRATION

LOW MIGRATION

60%

0%

UPDate STGpool STORAGEPOOL HIghmig=60 LOmig=0


• • •••



Minimum and Maximum High-Migration Thresholds

Setting the high-migration threshold to 100% prevents migration processing. If the storage pool is at 100% capacity, and the high-migration threshold value is 100%, then new files will by-pass the storage pool and will be directly written to the next storage pool.

Setting the high-migration threshold value to 0% will cause all files in the storage pool to be migrated. If new files attempt to write to the storage pool, they will be directly written to the next storage pool.

Caution: Performance may be affected as a result of files being written directly to tape instead of disk.

12

Minimum and Maximum High-Migration Thresholds

HIGH MIGRATION100%

CAPACITY FULL

HIGH MIGRATION0%

NEXT STORAGE POOL

NEXT STORAGE POOL

STORAGE POOL

STORAGE POOL


• • • ••



Standard Migration

How the Server Selects Files to Migrate

When data in a storage pool uses a percentage of the pool’s capacity that is equal to the high migration threshold, the server migrates files from that pool to the next storage pool. The server checks for the node of the client that has backed up or migrated the largest single file space or has archived files that occupy the most space.

13

Standard Migration

Mark /apps 175 A Mark /home 100 B

node file spacebackup or

archive

tapepool

High80%

low20%

MB

Node: ___________ MB ________

Question:If you employ a standard migration, whose data will be migrated first?

Manfred C: 80 A

Ramon G: 160 A

Claire E: 140 B

Harley /home 100 A

Simon E: 125 B

Manfred D: 150 B

Claire D: 120 A

Manfred C: 200 B

high90%




Migration Using migdelay

You can ensure that files remain in a storage pool for a minimum amount of time before the server migrates them to another pool, by setting a migration delay period for a storage pool. The file must remain in the storage pool at least as long as the migration delay period. For disk storage pools, the last time the file was accessed is also considered for migration delay.

Enabling Cache for Disk Storage Pools

If you decide to enable cache for disk storage pools, files can temporarily remain on disks even after migration. You may want to set migration thresholds lower when you use cache.

14

Disk Pool Migration—migdelay

Simon E: 125 B 3

Mark /apps 175 A 1

Claire D: 120 A 2

Ramon G: 160 A 2

Claire E: 140 B 1

Harley /home 100 A 4

high80%

low20%

node MBbackup or

archive

tapepool

# ofdays

Q: Whose data will be migrated first if migdelay=2?

Node __________ MB _____________

Mark /home 100 B 1

upd stg diskpool migdelay = 2

filespace


• • • ••



Using migdelay with migcontinue

There is another storage pool parameter that works with the migration delay parameter. This is the migration continue parameter. It tells Tivoli Storage Manager what should be done if, because of migration delay, there is not enough data available to meet the low migration parameter.

• Yes—Tells Tivoli Storage Manager that, if it is not able to reach the low migration value, it is to disregard the migration delay value.

• No—Tells Tivoli Storage Manager that it is OK if it is not able to reach the low migration value.

15

migdelay with migcontinue

Simon E: 125 B 3

Claire D: 120 A 2 Claire E: 140 B 1

Harley /home 100 A 4

high80%

low20%

node MBfile

space

upd stg diskpool migdelay = 2migcontinue=yes|no

days

Manfred D: 150 B 0

Manfred C: 200 B 0

Manfred C: 80 A 0

Mark /apps 175 A 1 Mark /home 100 B 1

Ramon G: 160 A 2

backup orarchive

tapepool


• • •••



Migrate STGpool Command

Use this command to migrate files from one storage pool to the next storage pool in the storage hierarchy. This command can only be used with primary storage pools. Only one migration or reclamation process for a given storage pool is allowed at any given time. If a migration or reclamation process is already running for the storage pool, you cannot start another migration process for the storage pool.

You should only use this command if you are not going to use automatic migration for the storage pool. To prevent automatic migration from running, set the HIGHMIG attribute of the storage pool definition to 100.

The migrate stgpool command will honor the values of the following parameters on the define stgpool and update stgpool commands:

– MIGPROCESS

– MIGDELAY

– MIGCONTINUE

– NEXTPOOL

– LOWMIG. However, you can override the value of this parameter on define stgpool and update stgpool by specifying a value for the LOWMIG parameter on the migrate stgpool command

16

Migrate STGpool Command•Used to control migration manually•Migrates regardless of the high migration threshold•Honors the following:

- Migprocess- Migdelay- Migcontinue- Nextpool- Lowmig – can override in the migrate stgpool command

•Can force reclamation before the migration•Only one process per storage pool at any time•Do not use if automated migration is to be used

• Set high migration threshold to 100 percent to keep migration from starting unless media is empty


• • • ••



The migrate stgpool command will ignore the value of the HIGHMIG parameter of the storage pool definition. Migration will occur regardless of the value of the HIGHMIG parameter. This command creates one or more migration processes that can be canceled with the cancel process command. The number of processes is limited by the MIGPROCESS attribute of the storage pool definition. To display information on background processes, use the query process command.

To issue this command, you must have system privilege, unrestricted storage privilege, or restricted storage privilege for both the storage pool from which the files are to be migrated and the next storage pool to which files are to be migrated.

MIGrate STGpool pool_name LOwmig = number DUration = minutes REClaim = No|Yes Wait = No|Yes


• • •••



Collocation

Tivoli Storage Manager provides the ability to collocate client data on tape volumes.

When files are moved to a collocated storage pool, Tivoli Storage Manager ensures that the files for a specific client are written to the same set of tapes. This can limit the number of tapes that must be mounted when restoring that client’s system.

Collocation can be done by the client or by the client’s file space.

There is a trade-off between collocation and non-collocation:

• Non-collocation increases performance on backups because Tivoli Storage Manager is not required to write the client’s data to the same set of tapes.

• Collocation increases performance on restores because of the possibility of having the client’s files on fewer tapes.

• Collocation is a parameter of any given storage pool. Each client whose data is placed in that storage pool will have its files collocated.

17

Collocation

Client A data Client B data Client C data

partiallyused

volume

partiallyused

volume

partiallyused

volumecollocated pool

scratchvolume

partiallyused

volume

fullvolume

non-collocated pool

• Tivoli Storage Manager provides the ability to collocate client data on tape sets.• Collocating versus not collocating involves performance trade-off:

– No collocation increases performance on writes.– Collocation can increase performance on restores.

• Collocation is enabled at the storage pool level.


• • • ••



Collocation by Group

Collocation by group is now supported. Groups of nodes can be defined, and the server can then collocate data based on these groups. Collocation by group can yield the following benefits:

• Reduce unused tape capacity by allowing more collocated data on individual tapes

• Minimize mounts of target volumes

• Minimize database scanning and reduce tape passes for sequential-to-sequential transfer

For newly defined storage pools, the default storage pool collocation setting is now GROUP.

Note: During collocation processing the message ANR1142I will be replaced with ANR1176I. See the Administrator’s Guide for more information.

18

Collocation by Group• Reduces unused tape capacity by allowing more collocated data on individual

tapes. • Minimizes mounts of target volumes during backups.• Minimizes database scanning and reduces tape passes for sequential-to-

sequential transfer.• Collocation by group is enabled at the storage pool level by default.

Client A data Client B data Client C data

Completelyused

volume

partiallyused

volume

Comlpetelyused

volume

Group Collocated pool

All other client data

partiallyused

volume

Collocation Group Collocation Group

Members

client A

client B

client C


• • •••



See the following new commands:

define collocgroupdefine collocmemberdelete collocgroupdelete collocmemberquery collocgroupquery nodedataupdate collocgroup

See the following changed commands:

define stgpoolmove nodedata query node query stgpool remove node update stgpool


• • • ••



Data Shredding

After client data has been deleted, it might still be possible to recover it. For sensitive data, this condition is a potential security exposure. Tivoli Storage Manager allows shredding (the destruction of deleted data) on data in random-access disk storage pools that have been configured as shreddable.

When shredding is not enforced, the server will delete expired data logically from the database. However, the storage used to contain the data will not be overwritten, and the data will still exist in storage until that storage is reused for other data. When shredding is enforced, the server will delete the data both logically and physically. The server will overwrite the storage used to contain the data.

19

Shredding of Data Stored on Disk

Database references to object a

a bab

a c

b

Shreddable Storage Pool

b b

a c

database

Database references deleted and object ais overwritten

object adeleted

or moved

• Allows disk storage pools to be designated as shreddable.• Random-access disk (not tape or other device types)• Primary pools only

• When a data object is moved or deleted from a shreddable pool, Tivoli Storage Manager server overwrites the object.

• Sensitive data objects are destroyed when deleted or moved, preventing undesirable data discovery.

a c

b

b b

a c

a b

b a

database


• • •••



Server Options for Data Shredding

Shredding can be done either automatically after the data is deleted or manually by command.

The advantage of automatic shredding is that it is performed without administrator intervention whenever deletion of data occurs.

The advantage of manual shredding is that it can be performed when it will not interfere with other server operations.

Shredding performance is affected by the amount of data to be shredded, the number of times that data is to be overwritten, and the speed of the disk and server hardware. You can specify that the data is to be overwritten up to 10 times. The default value is 0, which inhibits shredding.

20

Server Options for Data Shredding


• • • ••



Data Shredding Process

Steps for setting Tivoli Storage Manager for shredding on a random-access disk storage pool:

1. Specify that you want data to be shredded either automatically after it is deleted or manually by an administrator. Specify how the shredding is to be done by setting the SHREDDING server option. You can edit the server options file, or you can set the shredding option dynamically by using the setopt shredding automatic or setopt shredding manual command.

2. Define a new storage pool, or update an existing storage pool to enforce shredding, and specify how many times the data is to be overwritten after deletion with the shred=n parameter of the define stg or update stg command.

3. Define volumes to the storage pool and specify disks for which write caching can be disabled. If write caching is enabled, the overwrite operations are adversely affected.

4. Define and activate a policy for the sensitive data. For the copygroup destination in the management class, specify the shred storage pool.

5. Identify those client nodes whose data should be shredded after deletion, and assign them to the new domain.

21

Data Shredding ProcessHow to shred data:

• SHRED is an attribute on DEFINE STGPOOL and UPDATESTGPOOL commands

• SHREDDING server option • SHRED DATA command • QUERY SHREDSTATUS command

Example:

update stgp backuppool shred=2

ANR2017I Administrator SERVER_CONSOLE issued command: UPDATE STGPOOL backuppool shred=2

ANR1309W Shred value zero for storage pool DISKPOOL may render deleted data non shreddable.

ANR2202I Storage pool BACKUPPOOL updated.


• • •••



If you specified manual shredding, you can start the shredding process by issuing the shred data command. You can specify the maximum number of minutes the process runs before being automatically cancelled with the duration=n parameter. If you do not specify the duration, the server will stop the shredding process only after all deleted sensitive data has been shredded.

You can issue the query shredstatus command to see the status and amount of data waiting to be shredded. Issue the command query shredstatus format=detailed to display detailed information about data shredding on the server.

Note: Some other server operations involving moving or copying of data could compromise the intent and value of shredding. Consult the IBM Tivoli Storage Manager Administrator’s Guide for possible conflicts.


• • • ••



Reclamation

Based upon the retention period and version count defined in policy management objects, data on tapes eventually expires. Generally, not all the data on a tape expires at the same time, but the tape cannot be reused until all the data has expired.

When tapes reach a reclamation threshold, defined as a percentage of expired data, Tivoli Storage Manager moves the unexpired data to another tape. This consolidates the data and releases tape volumes to the scratch pool.

When collocation is enabled, data is only moved to volumes belonging to the same client.

22

Reclamation

reclaim threshold

50%

100%

TAPE_SP1

47% 38% 85%


• • •••



Reclamation—Collocated versus Non-collocated

In this illustration there is a collocated pool and a non-collocated pool. Both pools have volumes that are only partially used and are candidates for reclamation.

When reclamation occurs on the collocated pool, data is moved from partially full volumes to volumes that have the same client’s data on them. When reclamation occurs on a non-collocated pool, volumes are filled to capacity regardless of client.

23Reclamation—Collocated versus Non-collocated


• • • ••



Reclaim STGpool Command

Use this command to reclaim volumes in a sequential access storage pool. Reclamation makes the fragmented space on volumes usable again by moving any remaining files from one volume to another volume. This makes the original volume available for reuse. This command can only be used with sequential access storage pools.

Only one reclamation or migration process for a given storage pool is allowed at any given time. If a reclamation or a migration process is already running for the storage pool, you cannot start another reclamation process for the storage pool. You should only use this command if you are not going to use automatic reclamation for the storage pool. You can set the RECLAIM attribute of the storage pool definition to 100 to prevent automatic reclamation from running. This command will honor the values of the RECLAIMPROCESS, and RECLAIMSTGPOOL attributes of the storage pool definition.

To issue this command, you must have system privilege, unrestricted storage privilege, or restricted storage privilege for the storage pool being reclaimed and the reclaim storage pool, if applicable.

RECLaim STGpool pool_name THreshold=number DUration=minutes Wait=No|Yes

24

Reclaim STGpool Command•Used to control reclamation manually•Reclaims regardless of the reclamation threshold•Only one process per storage pool at any time•Do not use if automated migration is to be used

• Set reclamation threshold to 100 percent to keep automated process from starting


• • •••



Copy Storage Pools

Copy storage pools contain duplicates of data residing in primary storage pools. The tape volumes in a copy storage pool are typically sent off-site in the event of a disaster. Data is copied to storage pools incrementally. Only data that has not previously been copied to the copy storage pool is copied during a backup cycle.

The backing up of a primary storage pool to a copy storage pool should be done after the client’s data has been backed up or archived.

If possible, it is best to have two copy storage pools, one whose data is kept on-site and one whose data is taken off-site. The on-site copy storage pool is there for availability.

• You can have one or more copy storage pools.

• All the storage pools in a single hierarchy should be backed up to the same copy storage pool.

• More than one storage pool hierarchy can be backed up to the same copy storage pool.

25

Copy Storage Pools

pool hierarchy 1 pool hierarchy 2

backup stgpooloperations

diskpool

tapepool

diskpool

tapepool

copystorage

pool


• • • ••



Active-data Storage Pools

You can create a type of storage pool for storing active versions of client backup data. (Archive and space-managed data are not allowed.) These pools are called active-data pools. Whereas primary storage pools and copy storage pools contain both active and inactive data, active-data storage pools contain only active versions. As updated versions of backup data continue to be stored in active-data pools, older versions are deactivated and removed during reclamation processing.

You create an active-data storage pool with the define stgpool command:

>>-DEFine STGpool--pool_name--device_class_name----------------->

|>--POoltype--=--ACTIVEdata--+-----------------------------+----->

26

Active-data Storage Pool

Storage Pool Type: PrimaryStorage Pool Type: Copy

Storage Pool Type: Active-data


• • •••



Devices for Active-data Pools

Active-data pools can use any type of sequential-access storage. Depending on the device, use of active-data pools can offer specific benefits.

If the active-data pool is associated with a file device class (sequential-access disk), restoring client data is faster because volumes do not have to be physically mounted and because the server does not have to position past inactive files that do not have to be restored.

If the active-data pool is associated with removable media, such as tape, you can reduce the number of tapes being sent offsite. Also, client restores are faster because the server does not have to position past inactive files.

If the active-data pool is associated with a server device class for electronically vaulting data to a remote location, you save bandwidth by copying and restoring only active data.

27

Devices for Active-data Pools

Active-data pools on disk: Copies of active data for fast restore of client data

Active-data pools on tape: Copies of active data to reduce the number of tapes stored offsite

Active-data pools on remote server:Copies of active data to reduce bandwidth


• • • ••



Copy Activedata Command

You can copy active versions of backup data from a primary storage pool to an active-data storage pool by using the copy activedata command:

>>-COPY ACTIVEdata--primary_pool_name--active-data_pool_name---->

28

Copy Activedata Command


• • •••



Active-data Pool with Simultaneous WriteY

You can configure the Tivoli Storage Manager server to simultaneously write backup data to both a primary storage pool and to a specified active-data storage pool when the backups occur. Set the ACTIVEDATAPOOLS parameter of the define stgpool command for primary storage pools with random-access devices or sequential-access devices:

Syntax for random-access primary storage pools:

>>-DEFine STGpool--pool_name--DISK--+----------------------+----> -ACTIVEDATApools--=----active-data_pool_name-+-'

Syntax for sequential-access primary storage pools:

>>-DEFine STGpool--pool_name--device_class_name-----------------> -ACTIVEDATApools--=----active-data_pool_name-+-'

Regardless of whether you use the copy activedata command or simultaneous write, the Tivoli Storage Manager will only write data to an active-data pool if the data belongs to a node that is a member of a policy domain that specifies the active-data pool as the destination for active data. Use the following command to make that configuration:

>>-DEFine DOmain--domain_name-----------------------------------> -ACTIVEDESTination--=------active-data_pool_name---+-'

Then register the node to the domain.

29

Active-data Pool with Simultaneous Write


• • • ••



The function provided by the ACTIVEDATAPOOLS parameter is not intended to replace the copy activedata command. If you use the ACTIVEDATAPOOLS |parameter, use the copy activedata command to ensure that the active-data pool contains all active data of the primary storage pool.

If you want to change existing configurations to take advantage of active-data storage pools, use the update commands rather than the define commands.

Note: The Administration Center has been updated to encompass active-data configurations and manipulations.


• • •••



Basic Data Movement

In this illustration, data on Volume 1 is moved to one or more volumes in the same primary storage pool. If there is no space on any existing volumes, a scratch tape will be requested. This is assuming the MAXSCRATCH parameter is set to some value greater than zero and that the value has not been exceeded.

Moving files from one volume to other volumes in the same storage pool is useful in the following example situations:

• When you want to free up all space on a volume so that it can be deleted from the Tivoli Storage Manager server

• When you need to salvage readable files from a volume that has been damaged

• When you want to delete cached files from disk volumes

If you want to force the removal of cached files, you can delete them by moving data from one volume to another volume. During the move process, the server deletes cached files remaining on disk volumes.

30

Basic Move Data

backup_tape on-site primary pool

move data vol1

VOL 2 VOL 3 VOL 4VOL 1

No cache




Additional Data Movement

The move data command can also move all the data on a volume in one primary storage pool to another primary storage pool. When you specify a target storage pool that is different from the source storage pool, the server uses the storage hierarchy to move data if more space is required.

Note: Data cannot be moved from a primary storage pool to a copy storage pool. Data in a copy storage pool cannot be moved to any other storage pool.

You can move data from random access storage pools to sequential access storage pools. For example, if you have a damaged disk volume and you have a limited amount of disk storage space, you can move all files from the disk volume to a tape storage pool. Moving files from a disk volume to a sequential storage pool may require many volume mount operations if the target storage pool is collocated. Ensure that you have sufficient personnel and media to move files from disk to sequential storage.

Note: If the tape volume was scratch before it was used, it returns to scratch when moved. If it was not scratch before it was moved, you must delete it.

31

Additional Move Data

S01243 S01049

Primary

move data B01427 stg=backup_save

backup_tape backup_save

Primary

B01427 S01125




Moving Data in a Copy Pool

Data can be moved within a copy storage pool, in a process that is similar to the reclamation of off-site volumes. The data on the volume specified in the move data command is copied from the primary storage pool onto one or more volumes. The new volumes can then be taken off-site and marked as being off-site, and the volume that was freed can be returned on-site and marked as such.

32

Move Data in a Copy Pool

backup_tape backup_copy

Return

C01234 C01422 C01207

C01387 C01165 C01034off-site copyon-site primary

B01125 B01247 B01398

B01462 B01247 B01398

C01399

move data C01207




Disk Storage Pool Calculations

You want to have disk storage pool capacity that is equal to or greater than one night’s backup of your client data.

Some clients may be backing up large files, such as database files, that might better be sent directly to tape. These files should not be part of the disk storage pool calculations.

33

Disk Storage Pool Calculations

server

migrationbackups

network

How much disk storage is needed?•Have enough disk space available for at least one night’s backup•Consider the cost versus speed•Consider the portability of the media

backup pool




Dynamic Disk Requirements

Although you want enough disk storage pool space to contain a single night’s backup, there will be times when that capacity is exceeded. When this happens storage pool migration takes place. If the migration of files to another storage pool is not able to keep up with the client’s backups, then the backups stop until sufficient space has been created from migration.

To determine the minimum amount of disk space required to keep client backups running, you can graph the cumulative amount of data being backed up and the cumulative amount of data being migrated on an hourly bases. The greatest delta between the two for any one hour is the minimum amount of disk space required.

In the illustration 175 GB total data was backed up during the backup window. During the same time 95 GB total data was migrated. During the hour between 3 a.m. and 4 a.m. the total GBs of backup data was 95 GB greater than the total amount of data that had been migrated.

Because the migration process is slower than the backups in this illustration, there would need to be a minimum of 95 GB of free space in the disk storage pool before the backups and migrations began to prevent the backups from being interrupted.

34

Dynamic Disk Requirements

250

200

150

100

50

0

8 a.m. 10 12 2 4 6 a.m.

175170

150140125

105

857560

20 25 3545 55 65 75 85 95

S=95

I – 0 = S

cumulativestorage GB/hour

backup window

5 15Total data migrated in GB/hour

Total data backed up in GB/hour


Unit 3: Storage Pool ManipulationsStorage Pool Backup and Recovery


Lesson 2: Storage Pool Backup and Recovery

• Back up data without interruption of services

The design allows incremental backups to be taken while the server is operational and available to clients. Administrators can decide which storage pools should be backed up, then can also schedule these backups using the central scheduling feature of Tivoli Storage Manager. This means that storage pool backups can be run at a time convenient for the installation.

• Restore one volume in a storage pool

Restore individual volumes within a storage pool after a media failure.

• Restore a complete storage pool

Restore all damaged volumes within a storage pool after complete loss of a storage pool.

• Support disaster recovery

35

Storage Pool Backup and Recovery Objectives

• Back up data without interruption of service.

• Restore one volume in a storage pool.

• Restore a complete storage pool.

• Support disaster recovery.

• Automatically switch to copy if primary file is damaged.




Identify removable media volumes containing storage pool backup data to be sent to an off-site location for disaster recovery purposes. Provide for management of the volumes that are stored off-site for disaster recovery protection. Address the issues of file expiration, volume reclamation, and volume rotation. Identify which volumes to return to the on-site location if it becomes necessary to recover from a major disaster or from loss of a few volumes at the on-site location.

• Automatically switch to copy if primary file is damaged

If it becomes impossible to obtain the primary file, the copy should automatically be used if one is available.




Backup of Storage Pool Files

Use the backup stgpool command to create backup copies of files residing in a primary storage pool and store the copies in a copy storage pool. If a file is already duplicated in the specified copy storage pool, a new copy of the file is not made in that copy pool.

36

Backup of Storage Pool Files• Copy storage pool

• Sequential only• Multiple primary pools can use

one copy pool• Suggested back up of entire

primary storage pool hierarchy to same copy pool

• Back up to copy pool• Asynchronous from client

backup• File level granularity• Incremental backup of files

• Initiate backup• Manually with command or GUI• Automatically with administrative

command scheduling• Once per primary pool

backup stgpoolprimarypool copypool





• One primary pool can use multiple copy storage pools.

Storage pool backup enables files to be backed up to multiple copy storage pools. You can have many copy storage pools for each primary storage pool; you are not limited to two or three backup copies. Each copy pool, of course, requires additional database and storage pool space, a potentially important consideration.

• Used to support both media and disaster recovery.

By physically separating copies in multiple copy storage pools, Tivoli Storage Manager enables you to reduce the risk of losing data as a result of fire, flood, earthquake, or other disaster.

• Initiate for each copy using different copy pool targets.

You create multiple backup copies of a primary pool by initiating backup of the primary storage pool for each copy pool. The incremental backup is against each copy storage pool. The files are copied only if they are not already in that copy or if the copy in the copy pool has a different insertion date (that is, the file has changed since it was last copied into the copy pool).

37





• Volumes stored on-site.

You can use the on-site copy pool to reduce the potential for loss of data integrity due to media failure or other problems. If the primary file is not available or becomes corrupted, Tivoli Storage Manager accesses and uses a duplicate file from the on-site copy storage pool.

• Volumes stored off-site.

The purpose of off-site copy storage pools is disaster recovery.

Volumes in this disaster recovery copy storage pool are never automatically used by Tivoli Storage Manager. The volumes in this copy pool should be marked with the access mode of OFFSITE to make sure that Tivoli Storage Manager does not request a mount of the volume.

The volumes in the disaster recovery copy storage pool should be stored off-site together with the database backup, to make sure that you can recover the Tivoli Storage Manager environment if there is a disaster.

If you only use a copy storage pool with off-site volumes (that is, you do not maintain another copy storage pool with on-site volumes for the same data) and there is a problem with a file in the primary pool, the user gets an error message on trying to access the file. You can temporarily move volumes from the disaster recovery storage pool on-site and recover the user’s file.




Backup and Recovery Suggestions

Here are some suggestions for applying the new storage pool backup functions to your environment.

• Back up entire primary storage pool hierarchy to same copy pools.

If you set up your configuration so that each primary storage pool hierarchy is copied to the same copy pools, Tivoli Storage Manager is smart enough to recognize that when a file migrates down the storage hierarchy it does not need to be copied again into a copy pool. Or, if you move a file with the move data command, Tivoli Storage Manager again can recognize this situation.

• Consider using at least one copy pool for on-site recovery and one for disaster recovery.

It may be useful to have one backup copy of a primary pool that is kept on-site and can be used for media failures or damaged files. Recall that if users access a file that is damaged in the primary storage pool, Tivoli Storage Manager automatically lets the user access the copy in the copy pool, but only if the copy pool volumes are on-site.

It may be useful to have one backup copy of a primary pool that is used for off-site storing to be prepared for disaster situations. The primary pool is copied to on-site volumes in this special disaster recovery copy pool, and then the administrator

38

Backup and Recovery Suggestions

• Back up entire primary storage pool hierarchy to one or more copy pools.

• Consider using at least one copy pool for on-site recovery and one for disaster recovery.

• Use administrative command scheduling to automate storage pool backups.

• Back up the storage pools and then the database.

• Save the volume history and device configuration file (send off-site along with the server option file).

• Consider running audit volume periodically to mark damaged files.

• Run restore volume|stgpool periodically to replace damaged files.




changes the volume access mode to OFFSITE and physically moves the volumes to an off-site location.

• Use administrative command scheduling to automate storage pool backups.

The administrative command scheduling facility can be used to automate storage pool backups. The backup stgpool command can be scheduled to initiate at regular intervals.

• Back up storage pools and then the database.

To prepare for disaster recovery, the server storage pools should be backed up and then the database. In this way, the database and storage pools will correspond to a greater degree. This ensures that all storage pool files are recorded in the database. You may end up with extra records in the database, but at least there should be a record for each file in the storage pools. Also, since storage-pool backups may take longer than the database backup the storage pools should be backed up first, again to further reduce inconsistencies between the database and storage pools.

• Consider running the audit volume command periodically to mark damaged files.

Audit volume commands can be used to synchronize the database with the storage pools. Storage pool files with no record in the database are marked as damaged. Marking files as damaged aids in their restoration the next time a restore stgpool command is issued. If audit volume would take too long to run in your environment, issue it for certain critical volumes or use a rotating schedule.

• Run the restore stgpool command periodically to replace damaged files.

A damaged file is replaced when the restore stgpool command is issued for a primary storage pool. Then if a user tries to access the file, it may already have been replaced with the correct version.




Student Exercise


39

Student Exercise




Review Questions

1. True or False: During the migration process, data is written into the copy storage pool.

2. True or False: You can enforce data shredding on storage pools of device types that specify LTO tape media.

3. True or False: You can reduce the amount of time it takes to restore client data by using active-data storage pools.

4. True of False: If a tape in a primary storage pool has become partially damaged, the move data command is of no value.








Review Answers

1. True or False: During the migration process, data is written into the copy storage pool.

False. During migration process, data is written into the specified next storage pool.

2. True or False: You can enforce data shredding on storage pools of device types that specify LTO tape media.

False. Data shredding can only be enforced on random-access disks.

3. True or False: You can reduce the amount of time it takes to restore client data by using active-data storage pools.

True

4. True or False: If a tape in a primary storage pool has become partially damaged, the move data command is of no value.

False. You can salvage readable files from a volume that has become damaged with the move data command. Then, any remaining files that are not readable can be restored from the copy storage pool volumes. This could reduce the number of tapes that are needed to be returned from offsite, and reduce the amount of time it takes to recover the volume.


• • •••



Summary

40

Summary

You should now be able to:• Describe storage Tivoli Storage Manager storage pool concepts

• Perform Tivoli Storage Manager storage pool backup and recovery




• • • ••

4-1

Unit 4: Database and Storage Pool Problem Determinations





Introduction

This unit is an important part of your learning as an administrator. It will allow you to experience faults that can occur within the Tivoli Storage Manager server and how to recover. This unit yields a higher comfort level with the product and to show you the tolerance Tivoli Storage Manager has with failure as long as the proper steps have been taken to protect the server.

Objectives

2


• Recover from database mirror problems• Restore corrupt files on a storage pool volume• Recover a single database volume failure• Recover a primary storage pool• Recover the entire Tivoli Storage Manager database• Recover a database if these files are lost:

–Volume history file–Device configuration file


Unit 4: Database and Storage Pool Problem DeterminationsProblem Determination


Lesson 1: Problem Determination

This unit moves away from the lecture and lab scenario of class and merges you into exercises as a class. These exercises will give you real-life experience with recovering the Tivoli Storage Manager server environment. The exercises focus on manual recovery of database and log volumes rather than recovery with Disaster Recovery Manager.

3

Problem Determination•Learn the steps for problem determination when the Tivoli Storage Manager server has experienced a failure.

•Understand how the suggestions made for volume placement, mirroring and backup of Tivoli Storage Manager provide for Tivoli Storage Manager recoverability.

•Learn how to manually recover the Tivoli Storage Manager server in the event of a catastrophic failure (Disaster Recovery Manager can automate this process–an instructor-led class is available).

• Increase an administrator’s comfort with his or her ability to recover in the event of database failure, data corruption on a volume, or volume destruction.




Exercise Rules

Listed in the graphic above are some basic rules to follow during these exercises. These rules will help to ensure the exercises run smoothly. Several issues are created on the student systems during the exercises.It is very important that you adhere to the rules and consult your instructor when necessary.

4

Exercise Rules• The unit consists of exercises and a cursory overview of the

Tivoli Storage Manager Problem Determination Guide.• The exercises are conducted in an instructor-led

environment. It is very important to follow the steps with the instructor.

• The instructor will create problems on your systems that the class will repair together.

• Do not run commands that are not discussed, the systems need to be returned to the original class setup.

• If you encounter an issue, inform the instructor. Your issues can be a positive learning experience for all.

• Check your commands for syntax.• Remember this is a lab, this is the place to make mistakes.• A great deal can be gained from participation in the exercises.

4

Exercise Rules• The unit consists of exercises and a cursory overview of the

Tivoli Storage Manager Problem Determination Guide.• The exercises are conducted in an instructor-led

environment. It is very important to follow the steps with the instructor.

• The instructor will create problems on your systems that the class will repair together.

• Do not run commands that are not discussed, the systems need to be returned to the original class setup.

• If you encounter an issue, inform the instructor. Your issues can be a positive learning experience for all.

• Check your commands for syntax.• Remember this is a lab, this is the place to make mistakes.• A great deal can be gained from participation in the exercises.


• • •••



Student Exercise


5

Student Exercise


• • • ••



Review Questions

1. What four flat files will help you to substantially shorten the time it would take to restore the server in a disaster state?

2. True or False: If all the log and database volumes are lost, the DSMFMT command is all that is needed to allocate usable space for recovery.

3. True or False: A log in normal mode will allow you to restore a single database volume.

4. What database or log volume status would indicate a serious problem with a volume?

5. Based on the question above, what would you create on the server to avoid the server stopping when this status occurs?


• • •••





• • • ••



Review Answers

1. What four flat files will help you to substantially shorten the time it would take to restore the server in a disaster state?

devcnfg.out, volhist.out, dsmserv.opt, dsmserv.dsk

2. True or False: If all the log and database volumes are lost, the DSMFMT command is all that is needed to allocate usable space for recovery.

False, you need to use DSMSERV FORMAT command.

3. True or False: A log in normal mode will allow you to restore a single database volume.

False.

4. What database or log volume status would indicate a serious problem with a volume?

Offline.

5. Based on the question above, what would you create on the server to avoid the server stopping when this status occurs?

You would mirror the volumes. This keeps the server running when a database or log volume goes offline.


• • •••



Summary

6

Summary

You should now be able to:

• Recover from database mirror problems• Restore corrupt files on a storage pool volume• Recover a single database volume failure• Recover a primary storage pool• Recover the entire Tivoli Storage Manager database• Recover a database if these files are lost:

–Volume history file–Device configuration file


• • • ••



• • • ••

5-1

Unit 5: Macros and Scripts



• • • ••



Introduction

This unit covers the basics for writing macros and scripts to implement server automation. This unit also discusses creating a maintenance script with the new Administration Center interface.

Objectives

2

ObjectivesUpon completion of this unit, you will be able to:• Write and execute Tivoli Storage Manager macros • Define, manage, and use Tivoli Storage Manager server

scripts• Use macros and scripts to automate operations• Define a maintenance script using the Administration

Center


• • •••

Unit 5: Macros and ScriptsCoding Macros


Lesson 1: Coding Macros

Tivoli Storage Manager Macros

Use the macro command to call a file containing one or more Tivoli Storage Manager administrative commands to be performed. Writing commands in a macro can be especially helpful when you want to issue commands that are repetitive in nature and contain several parameters. You may also want to write macros to process related commands in a specific order.

3

Tivoli Storage Manager Macros

A macro is a file containing Tivoli Storage Manager administrative commands. Macro files are useful for:

• Eliminating errors and saving time by having commands preset.

• Repeating commands used many times a day.

• Reducing typing of commands that are long or complicated but are used regularly.

• Macros may contain other macro commands.

– This usage is referred to as a nested macro.– Can nest macros up to ten levels deep.

•Macros can be invoked from batch or interactive mode.•A macro cannot be invoked from the server console.

•The macro file must be available locally on thefile system to be run as opposed to scriptsthat are stored in the database.


• • • ••



Creating Tivoli Storage Manager Macros

Macros are text files that reside outside Tivoli Storage Manager. In order for other administrators to run a macro that you have created, they must have access to that text file.

You can only execute commands from a macro that are allowed by your administrative privileges.

4

Creating Tivoli Storage Manager Macros

Create your macro using your operating system editor.The macro name that you use must follow OS conventions.

Macros contain:• Comments for documentation• The commit statement• Variable substitution

Macro command uses and limitations:• Macros can issue the rollback command to back out commands.• Commands that start background processes cannot be rolled back.• Commands may be continued onto multiple lines.• A macro invoked from the tsm> prompt is a high-level macro.


• • •••



Documentation

Macros, by their very nature, are a source of documentation. But you can also add comments to document the macro within itself.

You cannot nest comments or span comments over more than one line.

5

Documentation

Macros can document important information:• They provide a record of the commands and sequence executed.• They can contain comments to identify the functions performed.

– Comments start with a slash and an asterisk (/*).– Comments end with an asterisk and a slash (*/).

• A comment can be on a line by itself or on a line that contains a command.


• • • ••



Committing a Command

When there is a dependency between commands, you may need to commit the first command before executing the second one. This can be accomplished through the use of the commit statement.

6

Committing a Command

• commit commandUsed when there is a dependency between commands.

• itemcommit option• Administrative client option.• Causes each command to be committed when processed.

• The end of the macro is an implied commit


• • •••



Macro Example with Line Continuation

Any administrative command, including SQL select statements, can be continued onto the next line through the use of the line continuation character in administrative client interactive mode. Continuation characters are useful when you want to process a command that is longer than your screen or window width. Without continuation characters you can enter up to 256 characters. With continuation characters you can enter up to 1500 characters. In the macro command, these maximums are for after any substitution variables have been applied.

With continuation characters, you can:

• Enter a dash at the end of the line you want to continue. For example:

register admin pease mypasswd - contact="mary, ext1234"

• Continue a list of values by entering a dash or a back slash, with no preceding blank spaces, after the last comma of the list that you enter on the first line. Enter the remaining items in the list on the next line with no preceding blank spaces. For example:

stgpools=stg1,stg2,stg3,- stg4,stg5,stg6

• Continue a string of values enclosed in quotation marks. Enclose the first line of the string in quotation marks, and follow it with a dash or a back slash at the end

7

Macro Example with Line Continuation

DEFINE SCHEDULE STANDARD DEMONSTRATION -DESCRIPTION="Selective backup schedule" -ACTION=SELECTIVE OBJECT="C:\CRITICAL\*.*" -STARTDATE=4/30/2002 STARTTIME=10:45 -DURATION=1 DURUNITS=HOURS PERIOD=1 -PERUNITS=DAYS DAYOFWEEK=WEEKDAY -EXPIRATION=NEVER OPTIONS="-VERBOSE"

/* */COMMIT/* */DEFINE ASSOCIATION STANDARD DEMONSTRATION CLIENT/* */COMMIT

/*******************************************************//* *//* *//* *//*******************************************************//* */

Name: SETDEMO.MACFunction: Set up a demonstration scheduleUpdated: 04/28/2002


• • • ••



of the line. Then, enclose the remainder of the string in the same type of quotation marks on the next line. For example:

contact=“mary pease, bldg. 100, room 1a, austin,”- “ext. 1234, alternate contact-norm pass,ext 2345”

Tivoli Storage Manager concatenates the two strings with no intervening blanks. You must use this method to continue a quoted string of values across more than one line.

Macro Examples—Task 1

You want to let users delete backed up or archived files from storage pools. From an administrative client, you can use the macro facility to register more than one client node at a time. For this example, you create a macro file named regeng.mac, that contains the following register node commands:

register node asinger choir contact='department 21'domain=engpoldom archdelete=yes backdelete=yesregister node jfrost skiing contact='department 21, second shift'domain=engpoldom archdelete=yes backdelete=yesregister node rob guitar contact='department 21, third shift'

8


Task 1Create a macro file named regng.macUse the macro to register and grant authority to a new administrator.Write the macro as follows:/* Register and grant authority to a new administrator */

REGister Admin robert passwd CONtact=“x1235”CommitGRant AUTHority robert CLasses=Policy

Command:macro regng.mac


• • •••



domain=engpoldom archdelete=yes backdelete=yes

Next, issue the macro command:

macro regeng.mac

Symbol Substitution in Macros

Variables used for symbol substitution are defined as a percent sign (%) followed by a number. This means that %1 is replaced by the first parameter specified on the macro execution statement, %2 by the second, and so on.

9

Symbol Substitution in Macros

define domain labteam%1

define policyset labteam%1 team%1

define mgmtclass labteam%1 team%1 %2

define copygroup labteam%1 team%1 %2 -standard type=backup destination=backuppool -frequency=0 verexist=5 verdeleted=2 -retextra=30 retonly=60 mode=modified -serialization=shrstatic


• • • ••



Executing a Tivoli Storage Manager Macro

To execute a macro, you must enter the macro command followed by the name of the file containing the administrative commands to be executed. You may also specify parameters that will be used for variable substitution.

10

Running a Macro

•A macro command can be run in batch or interactive mode.

•To run, enter the command macro followed by the name of the macro.

macro addnode.mac

•May pass parameters when run. In this case william is passed in a variable %1 as previously mentioned.

macro addnode.mac william


• • •••




11


Task 2Create a macro file named authrg, containing substitution variables, to register and grant authority to a new administrator. Write the macro as follows:

/* Register and grant authority to a new administrator */ REGister Admin %1 %2 - /* Enter userid and password %1 and %2 */ CONtact=%3 /* Enter contact info (in quotes if nec.) */ GRant AUTHority %1 - /* Server uses variable already defined by you */ CLasses=%4 /* Enter the privilege class */

Enter the values you want to pass to the server to process the command when you run the macro. Command:

macro authrg.mac jones passwd x1235 Policy


Unit 5: Macros and ScriptsCoding Server Scripts


Lesson 2: Coding Server Scripts

Tivoli Storage Manager Server Scripts

Server scripts are defined and stored in the Tivoli Storage Manager database. For this reason, any administrator with the authority to execute the administrative commands can run a server script.

Server scripts also give you the ability to code some rudimentary logic within the script.

12

Tivoli Storage Manager Server Scripts

• Provide for automation of common administrative tasks

• Scripts can be run from: • Administrative client

• Administration Center• Administrative command schedule

• Scripts can be defined, managed, and run• Scripts can no longer be run from the

server console• Stored in the server database versus

macros which are files and must be available




Script Flexibility and Usability

Scripts can be documented through the use of comments. They allow for the continuation of administrative commands onto additional lines. Substitution variables are defined using a dollar sign ($) followed by a number. The if statement and the goto statement can be used for conditional logic flow.

13

Script Flexibility and Usability

Scripts can include the following:

• Comments with /* */• Line continuation with -• Variable substitution with $• SQL select statements• Conditional logic flow


• • • ••



Administration Center Automation

Server scripts can be defined, maintained, and executed through both the command-line and the Administration Center interface.

When using the Administration Center, select the Server Maintenance item in the navigation tree.

When using the command-line interface, you issue the define script command. You may include the first line to be executed with this command. Any additional lines need to be added using the update script command. To issue this command, you must have system privilege.

Syntax

>>-define script--scriptname------------------------------------> >-----+------------------------------------------+--------------> .-line--=--1----------. -commandline---+-----------------------+ '-line--=--linevalue--' '-file--=--filename------------------------' > '-description--=--description text--'

14

Administration Center Automation


• • •••



Parameters

• scriptname–The name of the script, limited to 30 characters. This parameter is required.

• commandline–The command line that will be defined for the stored command (for example, query status). Command lines may contain substitution variables and may be continued across multiple command lines if a continuation character (-) is used at the end of the line. A command line is limited to 1200 characters. If the command contains blanks it must be enclosed in quotation marks.

• line=linevalue–The line number for the defined command line. Because stored commands can consist of multiple command lines, line numbers are used to order the lines for execution when the script is actually executed with the run command. The first line, or line 1, is the default.

• file=filename–The name of the file containing the commands for the script being defined.

If the file=filename parameter is specified, commandline and line=linevalue parameters cannot be specified.

• description=descriptiontext–A short description of the script. Description text is limited to 255 characters, and surrounding single or double quotation marks are required when the description contains blank characters. This parameter is optional.


• • • ••



Using the Administration Center to Define a Maintenance Script

This Maintenance Script wizard allows you to create basic scripts for the most important daily maintenance tasks.

Tip: To create more specialized scripts, use the command line or the manual scripting tool in the Health Monitor module of the Administration Center.

15Using the Administration Center to Define a Maintenance Script

Create maintenance script…


• • •••



Administration Center Maintenance Script Wizard

The Administration Center Maintenance Script wizard will create a script that will accomplish these tasks. Available tasks are preselected. Deselect any tasks you do not want to be part of your maintenance plan. After you have made your selections, click next. You will then be given the option of backing up the database.

Note: Tivoli Storage Manager’s default expiration interval is automated. The command q option will display the current ExpInterval value, expressed in hours. Automatic expiration can cause server degradation. If you want to control expiration processing either by command, or by the maintenance plan, then change the expiration interval to the value of 0. Make this change either in the dsmserv.opt file, or by the administrative command setopt expinterval 0.

16

Administration Center Maintenance Script Wizard

Creates a script to accomplish these

tasks




Selecting Available Tasks

The Tivoli Storage Manager Administration Center does not give you the option of selecting tasks where the resources to accomplish that task are not available.

Place your courser over to view information for performing that operation.

17

Select Available Tasks




Create a Database Backup Script Using the Wizard

Creating a database backup is one of the tasks accomplished through the Administration Center Maintenance Script wizard. This part of the maintenance script definition creates the portion of the script to perform the function.

18

Create a Database Backup Script Using the Wizard




Schedule Maintenance with Script Wizard

The Maintenance Script wizard allows you to schedule the script that was created. It is advised that if you use the maintenance script to not modify it manually. If you want more control, writing the scripts manually is a better option.

19

Schedule Maintenance with Script Wizard




Define a Custom Script in the Administration Center

Although the Maintenance Script wizard creates scripts to accomplish basic daily tasks, custom scripts may be created to accomplish other administrative tasks.

To find the custom scripting facility in the Administration Center, select Health Monitor from the main menu, select the link for the server that needs a script, select Database and Recovery Log Information, and select the link Work with Database and Recovery Log. You arrive at the screen above.

20Define a Custom Script in the Administration Center




Creating a Script in the Administration Center

Write the script in the Content box and preview the script. See the results in the Script Results box.

21

Creating a Script in the Administration Center




Conditional Logic Flow Statements

There are 25 symbolic return code values that can be checked for conditional logic flow. The return code numeric designator can also be checked. These return codes can be found in the appendices of the Administrator Guide.

22

Conditional Logic Flow

• Scripts have the following statements:– if (. . .)– exit– goto with label:

• The return code can be examined from the previous command in the if statement:– rc_ok (OK)– rc_notfound (Warning)– rc_unknown– rc_syntax


• • • ••



The if Statement

The if statement can check the return code of the previous administrative command.

23

The if Statement

Performs the action if the condition is true

Select * from sessionswhere upper(session_type) = ‘NODE’

if(rc_notfound) ba stg backup backup_tape w=y

condition action


• • •••



The exit Statement

The exit statement allows for the immediate termination of the script in response to a return code.

24

The exit Statement

Ends script processing

Select * from sessionswhere upper(session_type) = ‘NODE’

if (rc_ok) exitba stg backup backup_tape w=y


• • • ••



The goto-label Statement

A label is a character string followed by a colon (:). The goto statement allows sections of server script code to be bypassed.

25

The goto-label Statement

Move processing to the label specifiedSelect * from sessions

where upper(sessions_type) = ‘NODE’if (rc_ok) goto done

ba stg backup backup_copydone: exit


• • •••



A Sample Script

We can use this script to check whether any clients are still in session. If none are in session, we begin to back up the storage pool. If that is successful, back up the database.

If there still are clients in session with the server, we delete our schedule (which runs this script) and recreate it to start in twenty minutes.

26

Sample Script

Select * from sessions where upper(session_type) = ‘NODE’if(rc_ok) goto reschedule

backup stg backuppool copypool wait=yesif(error) exit

backup db dev=dlt3 type=full wait=yesdelete sched runbastgdefine sched runbastg t=a cmd=“run bastg” startt=03:30exit

reschedule:delete sched runbastgdefine sched runbastg t=a cmd="run bastg" startt=now+0:20


• • • ••



Server Processing Control in Scripts

Administrators can automate routine server operations. In server scripts, commands can now be processed in serial and parallel. This enables multiple processes to run concurrently, which reduces data transfer time.

See the following new commands for scripts:

PARALLELSERIAL

Control the running of multiple commands in a script with these new commands.

Use the PARALLEL command to allow any further commands in the script to run at the same time until a SERIAL command is reached. Use SERIAL to specify that any further commands must wait for the previous commands to finish before executing. When a script starts, all commands will run serially until a PARALLEL command is encountered.

Note: Parallel commands accessing common resources, such as tape drives, can run serially.

The script return code following a PARALLEL command will be the same as the script return code prior to the PARALLEL command. When a SERIAL command is encountered,

27

Sample Script in Parallel/*run multiple commands in parallel and wait for them to complete before proceeding*/PARALLEL/*back up four storage pools simultaneously*/BACKUP STGPOOL PRIMPOOL1 COPYPOOL1 WAIT=YESBACKUP STGPOOL PRIMPOOL2 COPYPOOL2 WAIT=YESBACKUP STGPOOL PRIMPOOL3 COPYPOOL3 WAIT=YESBACKUP STGPOOL PRIMPOOL4 COPYPOOL4 WAIT=YESSERIAL /*wait for all previous commands to finish*/PARALLEL /*after the backups complete, migrate stgpools simultaneously*/MIGRATE STGPOOL PRIMPOOL1 DURATION=90 WAIT=YESMIGRATE STGPOOL PRIMPOOL2 DURATION=90 WAIT=YESMIGRATE STGPOOL PRIMPOOL3 DURATION=90 WAIT=YESMIGRATE STGPOOL PRIMPOOL4 DURATION=90 WAIT=YESSERIAL /*wait for all previous commands to finish*/PARALLEL /*after migration completes, relcaim storage pools simultaneously*/RECLAIM STGPOOL PRIMPOOL1 DURATION=120 WAIT=YESRECLAIM STGPOOL PRIMPOOL2 DURATION=120 WAIT=YESRECLAIM STGPOOL PRIMPOOL3 DURATION=120 WAIT=YESRECLAIM STGPOOL PRIMPOOL4 DURATION=120 WAIT=YES


• • •••



the script return code will be set to the maximum return code from any previous commands run in parallel.

When invoking server commands that support the WAIT parameter after a PARALLEL command, the behavior is as follows:

– If you specify (or use the default) WAIT=NO, your script will not wait for the completion of the command when a subsequent SERIAL command is encountered and the return code from that command will reflect processing only up to the point that the command starts a background process. The final return code from the command will not be available to your script.

– If you specify WAIT=YES, your script will wait for the completion of the command when a subsequent SERIAL command is encountered and the return code from that command will reflect processing for the entire command.

In most cases, you should use WAIT=YES on commands being executed in parallel.


• • • ••



Student Exercise


28

Student Exercise


• • •••



Review Questions

1. True or False: Custom scripts cannot be created with the Administration Center utility.

2. True or False: Macros allow you to pass parameters.

3. True or False: Macros are only useful for running long commands.

4. True or False: Server scripts are saved as files.

5. True or False: Server scripts have conditional flow control.


• • • ••





• • •••



1. True or False: Custom scripts cannot be created with the Administration Center utility.

False.

2. True or False: Macros allow you to pass parameters.

True.

3. True or False: Macros are only useful for running long commands.

False.

4. True or False: Server scripts are saved as files.

False.

5. True or False: Server scripts have conditional flow control.

True.


• • • ••



Summary

29

SummaryYou should now be able to:• Write and execute Tivoli Storage Manager macros • Define, manage, and use Tivoli Storage Manager server

scripts• Use macros and scripts to automate operations• Define a maintenance script using the Administration

Center

• • • ••

6-1


Unit 6: Network Configuration and Tivoli Storage Manager

Performance Issues


• • • ••



Introduction

This unit covers network configuration, capacity planning, and parameters for critical Tivoli Storage Manager tuning.

Objectives

2

Upon completion of this unit, you will be able to:

• Describe Tivoli Storage Manager network planning issues

• Identify network topologies possible with Tivoli Storage Manager

• Explain Tivoli Storage Manager tuning issues

• Identify Tivoli Storage Manager networking tuning parameters

• Describe Tivoli Storage Manager server tuning parameters

• Identify Tivoli Storage Manager client tuning parameters

Objectives


• • •••

Unit 6: Network Configuration and Tivoli Storage Manager Performance IssuesPlanning Issues


Lesson 1: Planning Issues

Network Capacity Planning Issues

Network planning does not stop after Tivoli Storage Manager is installed. As new nodes are registered and additional networks are integrated into your environment, you will continue to plan to meet service level objectives.

Consider service level agreements for storage management as well as routine business network traffic. Check with all parties on the network.

Plan for the restore. Most plans center on client data backup. In most cases, after the initial backup the ongoing incremental backups are relatively fast. Always be mindful of whether you will be able to meet your restore objectives.

Part of your planning should be to determine if a single Tivoli Storage Manager server or multiple local servers are necessary. If the clients are within close proximity to the Tivoli Storage Manager server, consider using a storage area network (SAN) for storage traffic.

3

Network Capacity Planning Issues

• Network planning is ongoing

• Must match service level agreements

• Consider backup versus restore

• What is the proximity of the nodes?

• Data transport on LAN or SAN?


• • • ••



For more information on planning your storage area network, visit the IBM Redbooks Web site. Redbooks, like the following, will help you decide if and how to set up your SAN:

Introduction to Storage Area Network, SAN

Designing an IBM Storage Area Network

Planning and Implementing an IBM SAN

Application Issues

Complex applications such as Lotus Notes and Exchange can affect the Tivoli Storage Manager solution design.

If the environment has enough downtime to permit backups, the need for the Tivoli Data Protection agent decreases. In this case, determine if the scheduled downtime is enough to complete the backup.

If downtime is allowed, simply install the backup-archive client for all backups. This may or may not be advantageous, however. If the database or application is very large, the data captured on a daily basis during backup may overwhelm the server. Agent solutions usually perform an incremental-style backup, which backs up less than the backup-archive client.

4

R/3Oracle

AIX

SAPDBA

TSM

AIX

TDP for SAPR/3

synchronize Mirror

break Mirror

perform Backup

reconnect Mirror

R/3DBserver

backupserver

disksubsystem

Application IssuesDatabases

• Agent support?• Downtime

allowed?• Size and layout?

TSMserver


• • •••



Disaster Recovery Issues

With Tivoli Storage Manager Disaster Recovery Manager, you can back up both the database and primary storage pools to another Tivoli Storage Manager server. The ability to do this implies a sufficient distance between the two servers for disaster recovery. The network speed between the two servers is critical. A fiber connection is ideal. Investigate and test the connection over a WAN to ensure there is sufficient bandwidth and time to complete a disaster recovery.

If you have a disaster recovery site, could your clients access the server at this different location in the event of a disaster?

When doing this type of planning, it is best to consider restoring in the worst case scenario.

In the case of a disaster, if you are forced into doing slower network restores, consider prioritizing the restore of applications. Make recovery order decisions in advance with a business continuance plan. In most situations, all applications do not need to be up and running immediately.

5

• Does all data need to have an off-site copy?• How much data has to be handled each day?• What is the distance between the primary and off-site

locations?• How much time is available to create an off-site copy?• Is there an SLA about disaster recovery times?

Disaster Recovery Issues

Storage ConsolidationData SharingDisaster Tolerance


• • • ••



Planning for Recovery

Remember, the purpose of the Tivoli Storage Manager is not primarily for backups, but to restore lost or corrupted data.

Clients will have different restoration requirements. Some will need the restore in a few days. For others, you will have a few hours.

After the initial backup of most clients, the amount of incremental data backed up daily will seem comparatively small, and backup times short. It is the restoration of a complete system that usually catches people by surprise.

Also consider other recovery solutions in addition to Tivoli Storage Manager, such as high availability and hot-standby systems.

Best Practice: Test client system restores.

6

• Hot standby (HACMP or Cluster)• Intelligent Enterprise Storage server

solution• Restore

Planning for Recovery

R/3Oracle

AIX

SAPDBA

TSM

AIX

TDP for SAPR/3

Synchronize Mirror

Break Mirror

Perform Backup

Reconnect Mirror

R/3DBServer

BackupServer

DiskSubsystem


• • •••

Unit 6: Network Configuration and Tivoli Storage Manager Performance IssuesNetwork Characteristics


Lesson 2: Network Characteristics

Is there enough capacity on your network?

The type of network will directly affect throughput capabilities. An FDDI network can handle about 35 GB an hour. A T1 line will only support 0.5 GB per hour. Realize that while a high-speed network may deliver data to the Tivoli Storage Manager server at a high rate, a slow tape drive yields a slower backup or restore process. The slowest element between the data and the destination is the fastest that the process will run. Remember that Tivoli Storage Manager may not be the only application using the network in a shared environment.

If your clients can support multiple network adapters, consider using a separate backbone just for backups.

Efficiency is the percentage of the maximum throughput rate achievable. For example, the theoretical speed of Token Ring is 2 MB per second or 7.2 GB per hour. But realistically, Token Ring can only run at 80% of its theoretical limit or 5.7 GB per hour.

Part of your network throughput concerns will be just how fast a client can put data onto the network. The backup-archive client can initiate multiple sessions when doing backups and archives.

7

Network CharacteristicsWhere is the bottleneck?

160 MB/Sec 280 GB/hr

SSA SSA

MB/Sec GB/hrread 10 35write 5 17

logTSM

db

Server

Clientread write

F/W 3 <2Ultra 6 <4

MB/Sec

ProcessorMemoryAdapterBus speed

Network

803521251000Gb Ethernet

800.540.21.54T1

8015.85.645T3

80432150SPSwitch

5034.119.4155ATM

8035.212.5100FDDI

401812.5100Ethernet

805.7216Token Ring

401.81.2510Ethernet

% EfficiencyGB/hr w →MB/secMbps

2981043592 Gen 2

112403592 Gen 1

39143590-H1A

39143590-E1A

22480LTO Gen 3

4215LTO Gen 1

9835LTO Gen 2

144803580 Ult 3

GB/hr@ 80%

Native SpeedMB/sec

Drive Type

403.2160Ultra 3/W

100.840Ultra/W

806.4320Ultra 320

201.680Ultra 2/W

100.840F/W

GB/hr@ 70%

Native SpeedMB/sec

SCSI BusType

Note: These figures are for the bus in total.

ProcessorMemoryAdapterBus speed


• • • ••



How fast the server can receive data also influences backup speeds. If clients can be backed up directly to disk, the number of backup sessions will not be limited to the number of mountpoints available in the library.

Large files sometimes back up faster directly to tape. Too many tape drives on a single adapter may also reduce the maximum performance of each drive.

In a situation involving a large robotic library and a fast server, and DASD, attached to a Token Ring or 10 Mbps ethernet, the bottleneck will be in the network.

Tape drives will not perform at capacity with only one adapter on the server for all the drives. Tape drives need one or two drives per adapter.

Note: Throughput depends on drive performance and number of active drives.


• • •••



Student Exercise

Open your Student Exercises book and perform Exercise 1 for this unit.

8

Student Exercise


• • • ••

Unit 6: Network Configuration and Tivoli Storage Manager Performance IssuesPerformance Tuning Overview


Lesson 3: Performance Tuning Overview

The major performance factors concerning Tivoli Storage Manager performance are:

• Server configuration

• Networking

• Client configuration

• Profile of data

10

UNIX WindowsOS/390

ApplicationServer

Backup Server

FC FC

FC FC FC

Storage Area Network (SAN)

100 MB/s

TCP/IP

LAN/WAN

Tivoli Storage Manager Tuning Overview


• • •••



Performance Tuning Basics

Ongoing performance tuning tackles backup or restore issues because of the differences in data movement.

Many things might influence the movement of data from a client to a Tivoli Storage Manager server, or from the Tivoli Storage Manager server to the client. The network is usually considered the base performance improvement area. Be sure to broaden the field of components and rationale for performance testing, and to focus on individual elements and their effect on processes.

Follow the proper tuning procedures and then experiment with small changes. Make changes one at a time so you will know which one worked.

11

Performance Tuning Basics

• Decide if tuning for backup or for restore• Some settings might help for backup but not for restore or vice versa

• There are no set answers for each tuning issue• There are too many possible combinations of issues possible• A tuning action does not always give the expected result• Expect to test several tuning parameters to solve a tuning issue

• Keep things simple to arrive at the best tuning setup• Limit the amount of concurrent activity• Limit the amount of background tasks on the server• Limit the number of tuning changes for each test


• • • ••



Performance Tuning Basics Continued

You may need a support specialist to help tune a Tivoli Storage Manager environment. On the other hand, it may take a group representing the different operating systems, networking and hardware to investigate a Tivoli Storage Manager installation problem.

We will provide parameters and option settings in this unit with the suggested values. The values are valid in many circumstances, but may need to be modified to achieve the highest performance. One thing is certain, never change a value for performance tuning until you learn the potential impact.

12

Performance Tuning Basics - Continued

• Tivoli Storage Manager performance tuning is complex because Tivoli Storage Manager supports multiple OS platforms, many different types of disk and tape hardware and several communication protocols

• Tuning suggestions for one operating system do not always work on all operating systems

• Disk and tape devices might behave differently on different OS platforms

• Tuning requires finding the bottleneck, relieving that bottleneck then testing the settings

• It is an iterative process – always look for the bottleneck• Repeat the process until the bottleneck is relieved or when you are

satisfied with the throughput


• • •••



Tivoli Storage Manager Performance Considerations

Note the basic order of performance significance in the graphic. This would be the normal order for handling performance tuning issues. The rest of the unit follows this pattern.

13

Performance Tuning Considerations

Server (CPU, memory, I/O)

Server local network parameters

Client (CPU, memory, I/O)

Client local network parameters

Global network parameters

Tuning Order


• • • ••

Unit 6: Network Configuration and Tivoli Storage Manager Performance IssuesServer Tuning


Lesson 4: Server Tuning

Tivoli Storage Manager Server Considerations

14

• How much data needs to be handled?

• How many clients are connected?

• How much memory is needed?

• Is the server connected to the SAN or LAN?

• Is the tape library shared?

• Are there LAN and SAN managed nodes?

• Is the Tivoli Storage Manager server running

stand-alone?

Tivoli Storage Manager Server Considerations


• • •••



Tivoli Storage Manager Server Setup Tips

Database I/O is random in nature, and recovery log I/O is sequential. Serve these two different I/O patterns best by placing the database and recovery log on different physical devices. This also lessens the contention for a single disk device.

Also, do not place recovery logs and database volumes on UNIX raw volumes. There is not enough locking capability to prevent the inadvertent corruption at the database or recovery log files.

Do spread the database over multiple volumes on multiple devices to reduce seek time.

Striping does not help because database page reads and writes are done in 4K blocks, and striping will typically not chunk small blocks over multiple devices. Moreover, recovery log I/O is sequential, so only one device is involved. Striping is valuable when very large single reads or writes are performed. When using striping, use hardware striping on the device instead of operating system striping.

15

• Use database and recovery log mirroring.

• Place database volumes on different disk devices.

• Do not place a database volume on the same disk device as a log volume.

• Disk striping for the database and recovery log does not always help.

• Use two network adapters.

• Configure a maximum of two tape drives on a SCSI adapter.

• Tivoli Storage Manager generally needs at least two tape drives.

• Consider OS or hardware mirroring for disk storage pools.

Tivoli Storage Manager Server Setup Tips


• • • ••



Performance Considerations

Back Up to Disk Storage Pool

• Avoids delays for mounting tapes during a backup.

• Subsequent migration can be to a collocated tape pool.

• However, very fast clients may perform better going directly to tape.

• Clients with many large files (database) might see better performance directly to tape.

16

• Back up to disk storage pools

• Backup sets to reduce disaster recovery times

• Collocation or group collocation for critical clients

• Migration – multiple processes

• Multi-thread clients

• Compression

• Back up storage pools from disk

• Image backup versus incremental backup for disaster

recovery times

Performance Considerations


• • •••



Backupsets

Increases performance for restore (faster restore time).

Collocation

• Significantly decreases restore time for large numbers of files that might normally be spread over many tape volumes.

• If only a few clients need data collocation, move them into their own policy domain.

• Has a cost in migration time and increased tape usage.

• In general, primary tape pool should be collocated if needed.

• In general, copy storage pool should not be collocated.

Migration

• Multi-thread the migration if multiple tape devices are using the pool migprocess parameter on the storage pool.

• Prevent migration during backups.

Multi-threading Clients

• If your client supports multi-threading, by default, it will use two threads.

– One for metadata.

– One for data.

• Realize that when a client’s multiple threads are in session with the server, each session is counted toward the maxsession value.

Compression

• Experiment with compression.


• • • ••



• Understand the effect of sending compressed data through your tape hardware, which typically will also do compression.

• Do not compress already compressed files such as zipped files or JPEGs.

Generally, do not compress on the client. However, in cases with large databases or clients with a lot of available resources, compression may improve your network performance.

On the server side, compression allows the server to use fewer CPU cycles to read the data.

Storage Pool Backups

If your environment has slow or a limited number of tape drives, consider backing up the data while it is still on disk storage and then migrate the data to tape.


• • •••



Tuning the Tivoli Storage Manager Server with Options Overview

Logpoolsize

• The recovery log cache is less critical than the database cache, but still important.

• Allocate memory to produce a recovery log wait time = 0.

• Requires increased memory resources on the server.

• The range allocated can be from 128 KB to the size of the virtual memory available. The default is 512 KB.

The BUFPOOLSIZE and LOGPOOLSIZE must be set large enough to ensure that the server almost never has to read a database page or allocate a log page (initially) from disk, but can reference the page in memory. If your server system is severely memory constrained, these may have to be set to low values, though performance will suffer.

Configure the Tivoli Storage Manager server to have enough memory to achieve the performance shown above. Even the difference between a 98% and 99% cache hit percentage for the database can significantly improve performance.

17

• Logpoolsize

• Selftunebufpoolsize

• expinterval

• maxsessions

Tivoli Storage Manager Server Tuning Options


• • • ••



The LOGPOOLSIZE cannot be changed while the server is running, unlike the BUFPOOLSIZE.

Selftunebufpoolsize

• Allows for automatic adjustment in the database buffer pool size.

• Expiration processing is the process that would cause the value to increase.

• Database cache hit statistics are reset prior to expiration processing.

• Statistics are examined after the process.

• The buffer pool is adjusted if the hit ratio is less than 98%.

• The default is no.

Expinterval

• Specifies the interval in hours between automatic inventory expiration.

• The default is every 24 hours.

• Removes client backup and archive file copies from the database.

• Probably the most CPU-intensive and database-intensive processes in Tivoli Storage Manager.

• Set the interval to zero and schedule the administrative command expire inventory to perform expiration when Tivoli Storage Manager experiences the least usage.

• If you leave this setting at the default, or any non-zero value, and you create a schedule for expiration, expiration will run twice. Once at startup and after the number of hours set in this option continuously, and then also when the schedule runs.

Maxsessions

• Specifies the maximum number of simultaneous client sessions that can connect with the server.

• Default is 25 sessions.


• • •••



• Minimum is two sessions.

• Limiting sessions can improve server performance.

• Consider multi-threading clients, since clients can have up to 10 sessions, the number of server sessions must be monitored and this setting adjusted.


• • • ••



Server Configuration – Memory

Memory should not be a bottleneck on a Tivoli Storage Manager server. If you need it, add it. Since database reads and updates occur continuously on the Tivoli Storage Manager server, these processes should operate efficiently. Tivoli Storage Manager caches database entries in memory, which speeds data retrieval because the database is read from memory.

The recovery log also uses memory to manage updates. The improper allocation of memory buffer area to both the database and recovery log can adversely affect server performance.

Know the maximum addressable memory for the OS platform. It makes no sense to install more memory in a system than it can handle.

18

Server Configuration – Memory

•Largest use of memory in the Tivoli Storage Manager server is for the database bufferpool

•Most base server configurations have enough memory already installed

•2 GB RAM is usually sufficient for a dedicated Tivoli Storage Manager server machine

•Be aware of the largest amount of memory that can be addressed by a single process on a 32bit OS

•Tivoli Storage Manager servers on 64bit UNIX systems can use larger buffer pools and more memory


• • •••



Server Configuration – BUFPOOLSIZE

• Specifies the size of the database buffer pool, an amount of server memory allocated to cache database pages. This parameter is important because the database access speed affects server performance.

• Allocate as much memory as possible to this cache to obtain a database hit percentage > 99%. Requires increased memory resources on the server. In UNIX, increasing buffer memory takes up pinned memory, so a loss of performance may be experienced in other applications.

• q db f=d to show size.

• Range allocated can be from 256 KB to the virtual memory available. The default is 32756 KB.

• Buffer pool size starting point can be reset by using the selftunebufpoolsize command.

19

BUFPOOLSIZE

• Specifies the size of the server database buffer pool in KB• A large buffer pool can reduce database read and write disk I/Os• Initially set BUFPOOLSIZE between one-eighth and one-half of real memory

• Use up to 3/4 of real memory for a dedicated Tivoli Storage Manager server (no other applications)

• Do not increase the buffer pool if system paging is significant• See the query db command for tuning under a workload• Use the reset buffpool command to control the test or set selftunebufpoolsize yes

Minimum 256Maximum limited by available virtual memoryDefault 32756 (kb) default setting

Tune using the amount of real memory as a guideSuggested 131072 for server with 1GB real memory


• • • ••



Server Configuration – USELARGEBUFFERS

• Increases the client-server communication buffers

• Increases the disk device I/O buffers

• Increased from 32 KB to 256 KB

• Default is yes

20

USELARGEBUFFERS

• Specifies whether large buffers are used for server data movement• A large buffer makes data movement operations more efficient• A large buffer requires more memory, however, these memory requirements are modest

• Always set uselargebuffers yes

Values available yes (256KB buffers)no (32KB buffers)

Default yesSuggested yes


• • •••



Server Configuration – I/O Devices

The graphic above gives suggestions for I/O device placement to increase Tivoli Storage Manager server performance. These items mainly deal with hardware issues and require knowledge of system architecture. Consult the proper personnel for help with these suggestions if you are not skilled in this area.

21

Server Configuration – I/O Devices

•Balance I/O adapters and devices across multiple buses•Adapter placement can impact performance significantly•Use 66 MHz, 64 bit PCI slots for the most critical performance adapters

•Put network adapters on different PCI buses than storage adapters

•Limit the number of tape devices per SCSI or fibre adapter•Always put disk and tape devices on different fibre adapters


• • • ••



Server Configuration – Network

Network performance is a main issue for Tivoli Storage Manager implementation. Some suggestions require updates to routers and other communication equipment. Never assume a configuration is perfect. Ports speeds can be inadvertently changed and, as mentioned, hard set configurations do not use auto-detect features.

In most cases, TCP/IP will be the protocol used for data transport. Above are settings for performance dealing with TCP. This unit will cover these in more detail for the server and client.

22

Server Configuration – Network

• Use dedicated networks for backup (LAN or SAN)• Set Ethernet speed and duplex settings• Do not rely on auto-detect• Gb Ethernet jumbo frames (9000 bytes)

• Only useful if supported on client, server, and switch• Not all Gb Ethernet hardware supports jumbo frames• Improved throughput with lower host CPU usage

• Use Tivoli Storage Manager options to optimize network throughput• tcpwindowsize 63• tcpbuffsize 32• tcpnodelay yes


• • •••



Server Configuration – TCPWINDOWSIZE

Specifies, in kilobytes, the amount of received data that can be buffered at one time on a TCP/IP connection. The sending host cannot send more data until it receives an acknowledgment and a TCP receive window update. Each TCP packet contains the advertised TCP receive window on the connection. A larger window lets the sender continue to send data, and may improve communication performance, especially on fast networks with high latency.

• To improve backup performance, increase the TCPWINDOWSIZE on the server.

• The TCP window acts as a buffer on the network.

• A window size larger than the buffer space on the network adapter might degrade throughput due to resending packets that were lost on the adapter.

23

TCPWINDOWSIZE

•Specifies the amount of data in KB that can be buffered at one time within a session

• If the limit is reached, the sending host cannot send more data until a window update is received

•Larger window allows sender to continue sending data•Large window is useful on reliable, long distance, or high latency networks

Minimum 1Maximum 2048Default platform dependentSuggested 63

128 AIX in SP2 environments and GbitEthernet with Jumbo frames


• • • ••



Server Configuration – TCPNODELAY

Specifies whether the server disables the delay of sending successive small packets on the network. Change the value from the default of YES only under one of these conditions:

• Your service representative tells you to change the option.

• You fully understand the effects of the TCP Nagle algorithm on network transmissions. Setting the option to NO enables the Nagle algorithm, which delays sending small successive packets.

24

TCPNODELAY

•Specifies whether or not TCP/IP will buffer successive small outgoing packets

•Buffering can improve network utilization•Buffering requires a delay that can impact session throughput greatly

Value available yesno

Default platform dependentSuggested yes


• • •••



Server Configuration – Networks

Native drive throughput is key for backup and restore of large file workloads.

Time spent mounting a tape is time not spent backing up or restoring data. Reduce this time with fewer mounts by using larger capacity media, or by providing faster mounts.

Locate time is also time not spent backing up or restoring data. The backup must find the current end-of-data quickly and append. It is also meaningful for restore to quickly locate data blocks that contain data to restore. This is more important for Tivoli Storage Manager than for other backup-restore products. A full file system restore with other products would locate to the beginning of the data and read-restore all blocks belonging to the backupset being restored. A full file system restore with Tivoli Storage Manager would locate to each block containing data to be restored (not necessarily faster).

Native drive throughput has no measurable effect for backup and restore of small file workloads. It also has no throughput bottlenecks such as slow CPU, limited memory, slow disk, using client compression, slow network.

Backhitch time has an important impact on the backup of small file workloads direct to tape because Tivoli Storage Manager forces a buffer flush after every 256 files (maximum). For this reason, customers should back up these workloads to disk and migrate the data to tape.

Locate time during Tivoli Storage Manager restore of small file workloads can easily dwarf the time spent actually moving data. This will have measurable effect for backup and restore of slow clients like clients that have other applications running.

25

Server Configuration – Libraries

XX=important, X=meaningful, blank=not significant

XXXXXXstorage pool migration, collocation

XXXXXstorage pool migration without collocation

XXXrestore small file, slow client

XXXXback up small file, slow client

XXXXXrestore large file, fast client

XXXXXback up large file, fast client

Backhitchtime

Locate time

Mount timeCartridge capacity

Nativethroughput


• • • ••



Tivoli Storage Manager storage pool migration from disk to tape can typically use the maximum bandwidth of the slower device (disk or tape). This makes the tape native device throughput an important attribute when the disk is not the bottleneck.

Using Tivoli Storage Manager storage pool migration with collocation and many nodes or file spaces of data will cause many tape mounts and locates to find the EOD point. This makes mount time and locate time more important than when not using Tivoli Storage Manager collocation.


• • •••



Server Configuration – Tape

Sequential storage on tape is a viable part of the Tivoli Storage Manager implementation architecture. Tape is the normal media used to store backup and archive data off-site and in most cases is still used as the media for on-site primary storage pools, even though the use of disk-to-disk backup is growing. Tape provides a manageable media form for the processes of Tivoli Storage Manager.

The tape subsystem can be one of the most limiting factors to a Tivoli Storage Manager implementation. The speed of the drives for write and read will directly impact performance, of course, but the number of drives available for processing data is also important. Two is the minimum number of drives, but more drives would allow more concurrent processes to occur on the server. Installing the correct number of drives will make meeting the requirements for day- to-day operations more attainable.

26

Server Configuration – Tape

•Configure enough tape drives for:• The peak number of Tivoli Storage Manager client sessions backing up

direct to tape during the peak backup window• Add additional tape drives for other functions that run during the backup

window–Storage pool migration–Storage pool backup–Reclamation

•Key tape device attributes• native write-read throughput• compaction-compression throughput• cartridge capacity• mount time• locate time• backhitch time• reliability


• • • ••



Server Configuration – Disk

Tivoli Storage Manager can use disk drives for database and recovery log volumes. They are useful for the volumes used by disk storage pools, and can be used as sequential storage volumes with a devclass type of file. On SANs, using SANergy, you can create a file library and store data in sequential format on the drives. If you use disk as sequential volumes, remember that you are reading the file in sequential access methods and not random.

It is important to understand that using Redundant Array of Independent Drives (RAID) or Just of Bunch of Disks (JBOD) can affect performance. It is much better to use a hardware RAID solution that offloads the RAID algorithms to firmware on the adapter rather than letting the operating system perform this function. RAID becomes important when using disk-to-disk storage which will lessen the need for tape storage media.

27

Server Configuration – Disk

•Tivoli Storage Manager server can effectively use a multitude of disks

•Physical disks (JBOD) or RAID arrays?• RAID will require more disks for equivalent performance• Write throughput is important during backup-archive

• Dedicate physical disk-array to one Tivoli Storage Manager volume (at most two)

• recovery log• database• disk storage pool

•Place Tivoli Storage Manager volumes at outside diameter of physical disk

• better sequential throughput• faster seek time


• • •••



Server Configuration – Server Disk Volumes

The graphic above shows why multiple volumes on the same drive can cause a performance degradation. Since there is only one actuator arm in the drive holding the read-write heads, the movement of the actuator arm back and forth on successive reads or writes will slow the access. This is because there are two files. It can be much faster to have the data written to a single file (volume).

28

Server Disk Volumes

Partition or logical volume

Space being written to in current transaction

Tivoli Storage Manager volume

Used space

Increased seek activity can degrade performance

1 TSM volume per physical partition

2 TSM volumes per physical partition

Used space shown assuming all data was previously migrated


• • • ••



Server Configuration – Disk Write Cache

Use the information in the graphic above to make decisions about write cache.

29

Server Configuration – Disk Write Cache

•Use disk subsystem or adapter write cache for:• all RAID 5 arrays (for parity calculation)• physical disks with Tivoli Storage Manager database volumes (random I/O)

•Do not use disk subsystem or adapter write cache for:• physical disks with Tivoli Storage Manager storage pool volumes (sequential I/O)

•Large sequential I/O operations typically can overrun the cache


• • •••




Database and recovery log mirroring protects the Tivoli Storage Manager server from stopping should a volume corrupt. If the backup and restore acts on many small files, then the process of writing the data to the primary volumes and mirror copies can degrade the server if set to a sequential write status. This sequential format is preferable. Its higher level of fault tolerance ensures that the first write is successful before performing the write to the mirror.

To reduce the effect, use multiple volumes on different physical drives and make sure the bufpoolsize is set to its optimal value.

30


• Increased server availability•Can impact performance for backup and restore of small files• Impact can be mitigated by

• Use mirrorwrite db parallel and dbpageshadow yes– Page shadow file can be placed in server installation

directory• Use multiple database volumes• Use separate physical disks and arrays for each log or database volume

• Use disk subsystem or adapter write cache• Use optimal database buffer pool size


• • • ••



Server Summary

31

Network

Server

TCPWINDOWSIZE 64 (63)

TCPNODELAY Yes

BUFPOOLSIZELOGPOOLSIZE

MAXSESSIONS 25EXPIREINVENTORY 0

USELARGEBUFFERS 256 KB Yes

CacheMigration

CompressionCollocation

MOVEBATCHSIZE 1000

MOVESIZETHRESH 2048 MB

Back

ups

Server-related Tuning Summary

TXNGROUPMAX 256




* (3 x MTU)

**Set selftunebufpoolsize to yes to have Tivoli Storage Manager dynamically adjust the buffer pool.

To adjust the buffer pool manually, refer to the IBM Tivoli Storage Manager Administrator’s Guide for the steps needed to monitor statistics to determine appropriate settings.

UNIX AIX SP MVS Windows

bufpoolsize 32768 32768 32768 depends on memory**

logpoolsize 2048 2048 2048 512

txngroupmax 256 256 256 256

movebatchsize 1000 1000 1000 1000

movesizethresh 2048 2048 2048 2048

tcpwindowsize 64* 640* 64* 63

tcpnodedelay yes yes n/a n/a

uselargebuffers yes yes yes yes


• • • ••

Unit 6: Network Configuration and Tivoli Storage Manager Performance IssuesClient Tuning


Lesson 5: Client Tuning

Tivoli Storage Manager Client General Tuning Considerations

Incremental only backs up files that have changed.

Incremental by date can miss files that were copied onto your client. Also, it will not notify the server that files have been deleted from the client. This could affect your point-in-time restore capability.

Selective backups will require much greater resources during the backup process than incrementals. The trade-off will be the increased recovery time.

Command line uses fewer resources than the GUI.

When many small files are being backed up, more time is spent in opening and closing the files.

32

• Backup versus restore

• Incremental

• Incremental by date

• Selective

• GUI or command line

• Large versus small files

• Image backup

General Tuning Considerations for Tivoli Storage Manager Client


• • •••



Client Options

dirmc

• Specifies which management class to use for directories.

• The default is the management class in the active policy set of the policy domain with the longest retention period.

• Does not affect archived directories.

• May also be defined on the server.

tapeprompt

• Specifies if the user wants to be prompted for tape mounts.

• Default is no.

• Used for backup, archive, restore or retrieve options.

33

• dirmc• tapeprompt• include/exclude

Client Options


• • • ••



include/exclude

• Includes and excludes in the client options file can affect performance.

• The bottom of the list should have the most hits for the files.

• Ten excludes are better than 100.

• More important on slower processors.

• Must balance CPU time affecting throughput versus just sending data.

• Use exclude.dir to eliminate entire subdirectory structures.

• Back up critical data.


• • •••



Client Tuning Settings – Memory

34

• Tivoli Storage Manager backup-archive client may require lots of memory for incremental backup

• approximately 300 bytes per file or directory in a file system• so ... 300 MB for file system with 1 million files

• Can reduce memory requirements by using:• include or exclude to back up only what is needed• incremental by date (some files might be missed)• memoryefficientbackup yes (slower)• use multiple file systems• virtualmountpoint (UNIX only)• journal-based backup (Windows only)

Client Tuning Settings – Memory


• • • ••



Client Tuning Settings – MEMORYEFFICIENTBACKUP

The memoryefficientbackup option is:

• Useful for memory constrained workstations

• Usually found on NetWare

• Usually used with a limited amount of RAM and large file systems

The trade-off is Tivoli Storage Manager client CPU utilization and overhead

Use memoryefficientbackup=diskcachemethod for any file space that has too many files for Tivoli Storage Manager to complete the incremental backup with either the default setting, memoryefficientbackup=no, or with memoryefficientbackup=yes. The disk cache file created by the initial disk cache incremental backup can require up to 5 gigabytes of disk space for each million files or directories being backed up.

The disk cache file created by subsequent disk cache incremental backups should require considerably less disk space. The actual amount of disk space required for the disk cache file created by subsequent disk cache incremental backups depends on the number of files and directories included in the backup and on the length of the longest file or directory name which is backed up. Since the disk cache file can become very large, ensure that large file support is enabled on the file system that is being used for the disk cache file.

35

•Specifies whether to process an incremental backup one directory at a time

•Decreases the amount of memory used on the client•Can also decrease session throughput

Available values Yes (uses less memory)No (uses more memory)DISKCACHEmethod (uses much less

memory, but more disk space Default no (except for Macintosh systems)Suggested no

Client Tuning Settings – MEMORYEFFICIENTBACKUP


• • •••



Client Tuning Settings – Networks

36

•Use dedicated networks for backup (LAN or SAN)•Set Ethernet speed and duplex settings

- Do not rely on auto-detect•Gb Ethernet jumbo frames (9000 bytes)

- Only useful if supported on client, server, and switch- Not all Gb Ethernet hardware supports jumbo frames- Improved throughput with lower host CPU usage

•Use Tivoli Storage Manager options to optimize network throughput as seen on the following pages.

Client Tuning Settings – Networks


• • • ••



Client Tuning Settings – TCP

TCPWINDOWSIZE

The TCPWINDOWSIZE specifies, in kilobytes, the amount of received data that can be buffered at one time on a TCP/IP connection. The sending host cannot send more data until it receives an acknowledgment and a TCP receive window update. Each TCP packet contains the advertised TCP receive window on the connection. A larger window lets the sender continue to send data, and may improve communication performance, especially on fast networks with high latency.

• To improve restore performance, increase the TCPWINDOWSIZE on the client.

• The TCP window acts as a buffer on the network.

A window size larger than the buffer space on the network adapter might degrade throughput due to resending packets that were lost on the adapter.

37

Client Tuning Settings – TCP


TCPWINDOWSIZE

TCPBUFFSIZE

TCPNODELAYAvailable values yes

noDefault platform dependentSuggested yes


128 AIX in SP2 env. or Gb with Jumbo Frames


• • •••



TCPBUFFSIZE

The TCPBUFFSIZE option specifies the size of the internal TCP/IP communication buffer. It uses more memory, but a larger buffer can improve communication performance.

TCPNODELAY

The TCPNODELAY specifies whether to send small transactions to the server without buffering them first. A small transaction is smaller than the byte limit set with the TXNBYTELIMIT option. Specifying TCPNODELAY yes might improve performance in higher-speed networks.

This setting is the same as that on the server. It specifies whether the server disables the delay of sending successive small packets on the network. Change the value from the default of YES only under one of these conditions:

• Your service representative tells you to.

• You fully understand the effects of the TCP Nagle algorithm on network transmissions. Setting the option to NO enables the Nagle algorithm, which delays sending small successive packets.


• • • ••



Client Tuning Settings – DISKBUFFSIZE

• The default is 32 except on AIX clients when enablelanfree is set to no.

• Affects backups, archives and HSM migrations but not restores.

• Different but complementary to server.

• Valid ranges are 16 Kb through 1023 Kb.

• Helps on large file reads.

• Larger values use more client memory.

• Optimal backup, archive, or HSM migration client performance may be achieved if the value for this option is equal to or smaller than the amount of file read ahead provided by the client file system.

38

•Overrides LARGECOMMBUFFERS parameter•Option specifies the maximum disk I/O buffer size (in kilobytes) that the client may use when reading files

•Range of values is 16 through 1023•Large buffers can be more efficient•Tune value to file system read ahead to get benefits

Available values 16 - 1023 Default 32 (all platforms except AIX)

256 for AIX with enablelanfree set to noSuggested less than or equal to read ahead value

Client Tuning Settings – DISKBUFFSIZE


• • •••



Client Tuning Settings – COMPRESSION

• The process of saving storage space by eliminating empty fields or unnecessary data to shorten the length of the file.

• Compression processing usually results in a decrease in throughput.

• Saves on network capacity.

• Saves on server cycles and storage space.

• Alternatives are tape drive compression and operating system built-in compression.

• Files must be at least 12 KB in size before they will be compressed.

39

•Specifies whether to compress file data before sending it to the Tivoli Storage Manager server

•Decreases the amount of storage required on the server•Can also decrease session throughput•For maximum performance with a single fast client, fast network, fast server, turn compression off

•Remember, most tape devices have hardware compaction

Available values yesno

Default noSuggested no single client, fast network, fast server

yes multiple clients, slow network, slow server

Client Tuning Settings – COMPRESSION


• • • ••



Client Tuning Settings – Multi-session Client

The Tivoli Storage Manager client can back up and restore using more than a single session with the server, speeding up the task. Because it is a performance improvement, this topic is covered both here and in the unit on backups and restores.

Several options and parameters must be set correctly for a multi-session backup and restore. The client’s RESOURCEUTILIZATION value option must be greater than 1, and the maximum number of mount points must be greater than 1 if utilizing sequential storage (set for the node at the server). The data must also reside on more than one volume.

Note: There is a lab showing this functionality in the unit on backup and restore methodology.

40


• Improves backup and restore throughput for large clients

• Multiple file systems• RAID disks• Fast networks

• Multiple sessions sending or receiving data• Controlled by client resourceutilization option• When using sequential storage, each data session

requires a mount point• May need to update node maximum mount points• update node ... maxnummp=2


• • •••



Client Tuning Settings – Multi-session Client Backup

The graphic illustrates the differences between a single session and a multi-session backup. Notice that the number of paths for the data to follow increases, which can also increase the performance of the operation.

41

Client Tuning Settings – Multi-session Client Backup

Single session backup:

Multiple session backup:


• • • ••




If there is a small amount of data to be backed up incrementally, there may not be a need to spawn more than one session for the process. In this case, there is no performance improvement because there is not a multi-session backup.

In a multi-session backup scenario, each session can be used independently to send data. This lack of affinity means the session is not tied to a specific file space. This could, and probably would, adversely affect a backup using the file space form of collocation. The objective is to place the backups for a file space on a node on the same set of tapes.

The lack of affinity could lead to file spaces being mixed between the tapes that mount in support of multi-session backup.

It is sometimes better not to configure multi-session if the node always backs up a single file space on a disk. The multiple sessions will cause contention with each other during disk access requests.

Also check the maxsessions setting. A low setting will inhibit the server’s ability to generate extra sessions for the client.

42


• Only one session comparing attributes for incremental backup

• Incremental backup throughput does not improve for a single file system with a small amount of changed data (<1% per day)

• Sessions do not have file system affinity, so each session can send data from multiple file systems

• This is good for workload balancing• This is bad if directly using a storage pool collocated by file space

– Do not use the multi-session client to back up directly to a storage pool collocated by file space - use multiple commands, one per file space

• Using multiple sessions for backing up a single file system may increase contention on the source disk and degrade performance due to increased seek activity

• Multiple sessions may not get started because of few entries on txn queue• May need to increase the maximum number of server

sessions


• • •••



Client Tuning Settings – Multi-session Client Restore

The graphic reiterates the settings needed for multi-sessions.

43

Client Tuning Settings – Multi-session Client Restore

• Number of sessions is limited by:• Number of sequential access volumes with data to be restored• Number of mount points• Client resourceutilization option

Client server

session 1

session 2

session 3


• • • ••



Client Tuning Settings – Multi-session Restore

Multi-session restores work a little differently than backups. A restore from a disk storage pool will always only use a single session.

Multi-session restores can greatly reduce restore time. With single session only capability, a client with data to restore on nine volumes, using the Tivoli Storage Manager server with three tape drives, would need nine consecutive tape mounts. By increasing the appropriate parameters and options, the client could open three sessions. If the drives were idle, the process could mount three volumes at one time for the restore. As long as the data transfer can support the flow, the restore time could be cut to a third.

44

Client Tuning Settings – Multi-session Restore

• Only one session for restoring files from disk storage pools• Even small clients can gain throughput for restores

requiring a lot of tape mounts or locates• Can only restore a single file space at a time with

command line• Multiple sessions may still be used for a single file space

• Watch out for tape volume contention when restoring multiple file spaces from a storage pool collocated by node, or non-collocated


• • •••



Client Tuning Settings – RESOURCEUTILIZATION

• Allows multiple session when sending data to the server

• Values are from 1 to 10–sliding scale

• Default is 2

• Number of sessions is self-determined

• Used by both command-line and GUI interface

• Client must support multi-threading

By increasing the value, you could add threads created during this process. However, the number is just a scale value, and does not imply the number of threads used.

One thread monitors the activity and decides if the thread count should increase or decrease.

Threads are divided so that some are reading from disk (providers) and some are in session with the server, sending data (consumers).

To take advantage of more than one consumer and one provider thread, back up files from more than one file space.

45

Minimum 1Maximum 10Default 1Suggested 1 workstation (1 data session)

5 small servers (2 data sessions)10 large servers (4 data sessions)

Client Tuning Settings – RESOURCEUTILIZATION

•Controls how many client sessions are established with the server

•More sessions require more client and server resources


• • • ••



Take into consideration:

• It can result in multiple accounting files (1 per session)

• It can run out of sessions. Check maxsessions and maxschedsessions

Query node may not summerize client activity correctly.


• • •••



Client Tuning Settings – QUIET

• Keeps messages from being written to the screen during backups.

• The default is verbose.

• Especially helpful on the client’s initial backup.

• File information is not written to log files.

46

Client Tuning Settings – QUIET

• Code either QUIET or VERBOSE (default)• QUIET reduces client message output• May get a slight performance improvement by reducing I/O

to the dsmsched.log file or reducing I/O to the display device

• You may lose some important diagnostic information if you code QUIET

Default VERBOSESuggested QUIET


• • • ••



Client-related Summary

UNIX AIX SP Windows

txnbytelimit 2097162 2097162 2097162

tcpwindowsize 64 (3 x MTU) 256 (backup)640 (restore)

63

tcpbuffsize (multiples of 4) 1/3–1/2 tcpwindowsize

1/3–1/2 tcpwindowsize 32

tcpnodedelay yes yes n/a

tapeprompt no no no

quiet yes yes yes

largecommbuffers yes yes no

47

Network

ClientMEMORYEFFICIENTBACKUP

TAPEPROMPT No

INCLUDE/EXCLUDE


TCPNODELAY Yes

TXNBYTELIMIT 25600 KB

TCPBUFFSIZE 32 KB

LARGECOMMBUFFERS256 KB

dsmsched.log

QUIET

Client-related Summary

COMPRESSIONDIRMc


• • •••

Unit 6: Network Configuration and Tivoli Storage Manager Performance IssuesTransaction Tuning


Lesson 6: Transaction Tuning

Errors and Interruptions

Transactions update the database during backups and archiving, and performance depends on uninterrupted streams of data flowing from the client to the server. Interruptions can occur when a file grows during compression and then compression is canceled. It is sometimes better to turn on COMPRESSALWAYS so that the compression continues. Files in use will cause retries according to the serialization setting in the policy. Try to mitigate, as much as possible, any issue that can cause retries.

Retries can be seen in the detail information on the screen or in the log files. Review and address these regularly to keep performance high. If you use the quiet option for the client, you will lose the visibility of these retries. On the other hand, you may use alternate option files on a scheduled basis that use the verbose option to accumulate the information and parse the results looking for retries.

48

Errors and Interruptions

•Errors or interruptions during backup and archive cause retries

• Client file changed• Client compression results in larger file• Prompting for offline media

•Observe in-session output, or dsmsched.log file•Not seen if using the quiet option•Retries can significantly degrade performance

• Transaction data must be sent again


• • • ••



Suggested Tivoli Storage Manager Options

The graphic above indicates the suggested values to be set on the server and the client for transaction performance tuning. These options will be covered in more depth on the following pages.

49

Suggested Tivoli Storage Manager Options

•Server• TXNGROUPMAX 256 (can set larger, by node, up to 4096)• MOVEBATCHSIZE 1000• MOVESIZETHRESH 2048

•Client• TXNBYTELIMIT 25600• TXNBYTELIMIT 2097152 (with LTO or DLT)


• • •••



TXNGROUPMAX

• Specifies the number of files in a transaction transferred as a group between commit points. Works in conjunction with the TXNBYTELIMIT client option.

• Affects the file aggregation feature during backup and archive operations.

• Client retries causing entire transaction to be resent.

• Set to at least the suggested value of 256 on all servers.

TXNGROUPMAX sets the maximum number of files a client can send to a server in a single transaction. If the number of objects sent from the client reaches this value before the sum of the bytes reaches the value of the client parameter TXNBYTELIMIT, the transaction will be committed with the number of objects in the transaction equal to the TXNGROUPMAX value. If the sum of the bytes sent exceeds the TXNBYTELIMIT, a transaction is written that contains the number of objects up to and including the object that caused the value to be exceeded. Therefore, the number of files sent in a given transaction can be quite dynamic. If all files are very small and TXNBYTELIMIT is very large, then most transactions will contain the TXNGROUPMAX value number of objects. Occasional large objects will cause fewer objects to be sent in a transaction because the TXNBYTELIMIT having been reached will cause a transaction to complete with less than the TXNGROUPMAX number of objects in them.

50

TXNGROUPMAX – Server Setting

•Specifies the maximum number of objects (files and directories) included in a client session transaction

• New maximum value – 65000• Lab tests show no value beyond 4096• Can now be set at the node level – txngroupmax=4096 on register node

Minimum 4Maximum 65000Default 256Suggested 256 as long as file retries are rare,

and recovery log space is available


• • • ••



TXNBYTELIMIT

• Affects the Version 3 file aggregation feature during backup and archive operations.

• Works in conjunction with the server parameter txngroupmax by reducing the number of server transactions.

• Requires increased recovery log space usage on the server.

• Suggestion:

– Set to the maximum of 2 GB (2097162) on all clients.

– Maximum aggregate size is still 25600 KB.

– On very slow networks, such as modem dial-up lines, set the default to 2048.

51

TXNBYTELIMIT – Client Setting

•Specifies the maximum number of KB included in a client session transaction

•A single file exceeding this size is always processed as a single transaction

•Backup-archive client only• HSM and TDP clients may use different rules

Minimum 300Maximum 2097152Default 2048Suggested 25600 as long as file retries are rareSuggested 2097152 for backup direct to LTO or DLT


• • •••



If you have serialization in your copy group set to shrstatic or shrdynamic and you want to back up open files, a large txnbytelimit value can have an adverse effect. In this case, if you try to back up an open file, Tivoli Storage Manager will cancel or end the transaction in order to retry accessing the file.


• • • ••



MOVEBATCHSIZE

• Specifies number of files in a server internal transaction.

• Affects backup stgpool, move data, migration, reclamation.

• Requires increased log space usage on the server.

• Set to the maximum for your platform server unless the server is heavily overloaded or log space is severely constrained.

• movebatchsize=1000.

• Use the define spacetrigger command to allow for the expansion of the recovery log.

52

MOVEBATCHSIZE – Server Setting

Specifies the maximum number of objects (server physical bitfiles) included in a server data movement transaction

• Storage pool migration• Storage pool backup and restore• Reclamation• Move data

Minimum 1Maximum 1000Default 1000Suggested 1000


• • •••



MOVESIZETHRESH

• Specifies amount of data in a server internal transaction.

• Affects backup stgpool, move data, migration, reclamation.

• Set to the maximum for your server platform unless the server is heavily overloaded or log space is severely constrained.

• movesizethresh=2048

• selftunetxnsize option is not functional.

Beginning in Tivoli Storage Manager 5.3 the default parameter is no and is ignored if set to yes.

• Requires increased log space usage on the server.

• Use the define spacetrigger command to allow for the expansion of the recovery log.

53

MOVESIZETHRESH – Server Setting

Specifies the maximum number of MBs included in a server data movement transaction

• storage pool migration• storage pool backup and restore• reclamation• move data

Minimum 1Maximum 2048Default 2048Suggested 2048


• • • ••



Overview of Tuning

54

Network

server clientMEMORYEFFICIENTBACKUP

TAPEPROMPT NO

INCLUDE/EXCLUDE

BUFPOOLSIZELOGPOOLSIZE

MAXSESSIONS 25EXPIREINVENTORY 0

TXNGROUPMAX 256

USELARGEBUFFERS 256 KB Yes

CacheMigration

compressioncollocation

TXNBYTELIMIT 2 GB

TCPBUFFSIZE 32 KB

COMPRESSIONDIRMC

LARGECOMMBUFFERS

256 KB

dsmsched.log

QUIET

MOVEBATCHSIZE 1000

MOVESIZETHRESH 2048 MB

Bac

kups

Overview of Tuning

TCPBUFFSIZE 32 KB


TCPNODELAY YES


• • •••



Student Exercise


55

Student Exercise


• • • ••



Review Questions

1. True or False: Once your Tivoli Storage Manager network is designed, there is no need for reevaluation.

2. True or False: The types of applications that are being backed up have a direct effect on your network.

3. Which of the following data types can be sent across the network by Tivoli Storage Manager?

a. Database backups

b. Client file restores

c. Storage pool copies

d. Client data backups

e. All of the above

4. True or False: Clients and servers have no direct effect on your network throughput.

5. TCPWINDOWSIZE goes on which Tivoli Storage Manager component.

a. Client

b. Server

c. Both

d. Neither

6. True or False: Quiet mode eliminates client detailed messages.

7. True or False: Increasing the movebatchsize and movesizethresh values has an impact on the server log size.

8. True or False: During a backup operation, a client can have multiple threads in session with the server.


• • •••




• • • ••



Review Answers

1. True or False: Once your Tivoli Storage Manager network is designed, there is no need for reevaluation.

False

2. True or False: The types of applications that are being backed up have a direct effect on your network.

True

3. Which of the following data types can be sent across the network by Tivoli Storage Manager?

e. All of the above

4. True or False: Clients and servers have no direct effect on your network throughput.

False

5. TCPWINDOWSIZE goes on which Tivoli Storage Manager component.

c. Both

6. True or False: Quiet mode eliminates client detailed messages.

True

7. True or False: Increasing the movebatchsize and movesizethresh values has an impact on the server log size.

True

8. True or False: During a backup operation, a client can have multiple threads in session with the server.

True


• • •••



Summary

56

You should now be able to:• Describe Tivoli Storage Manager network planning issues

• Identify network topologies possible with Tivoli Storage Manager

• Explain Tivoli Storage Manager tuning issues

• Identify Tivoli Storage Manager networking tuning parameters

• Describe Tivoli Storage Manager server tuning parameters

• Identify Tivoli Storage Manager client tuning parameters

Summary


• • • ••



• • • ••

7-1

Unit 7: Reporting

Unit 7: Reporting


• • • ••

Unit 7: Reporting


Introduction

This unit discusses the SQL interface that can be used to query the Tivoli Storage Manager database, configure the ODBC interface, install and configure Tivoli Storage Manager Operational Reporting. This unit also discusses the Tivoli Data Warehouse.

Objectives

2


Describe relational database access techniques

Access Tivoli Storage Manager database information

using the SQL administrative interface

Implement Tivoli Storage Manager Operational Reporting

Describe Tivoli Data Warehouse


• • •••

Unit 7: ReportingTivoli Storage Manager SQL Interface


Lesson 1: Tivoli Storage Manager SQL Interface

Accessing Information in the Tivoli Storage Manager Database

An administrator is able to use the SQL SELECT statement to access data stored in the Tivoli Storage Manager database. Much of this information can also be accessed using queries at the Tivoli Storage Manager command line. However, select statements give the administrator flexibility to choose exactly what information he would like to see and in what format he would like to see it.

3Accessing Information in the Tivoli Storage Manager Database

QUERYor

Selectselect * from LIBVOLUMES

TSM databaseLIBVOLUMES

BACKUPSfilter

serverevents

LIBRARY_NAME VOLUME--------------------------------------DLTLIB DLT01DLTLIB DLT02DLTLIBDLTLIB

TSMServer


• • • ••



Requesting Information from RDBMS

Access to the Tivoli Storage Manager database can be through the predefined Tivoli Storage Manager query commands or through a SQL SELECT interface. The benefit of the SQL interface is the ability to do your own queries.

4Requesting Information from Relational Database Management System (RDBMS)


• • •••



SQL Select Statement Syntax

Use help select (to access the manual page entry for this command).

The syntax of the Tivoli Storage Manager select statement is compliant to ANSI SQL syntax but limited from the full specification provided by the SQL92/93 standard. These limitations reduce query complexity and the corresponding load on the Tivoli Storage Manager server database.

An SQL select statement can be divided into a number of clauses.

SELECT and FROM

The basic clause is SELECT, used to select columns FROM tables. The columns selected can optionally be aggregated and displayed with a different output column name:

column_name-Column name of the table or tables specified in the FROM clause

aggregate_function-Functions such as SUM, COUNT, MAX, or AVG that extract a single value from a group of columns

output_column_name-Column title displayed by the query output for the specified column name (default is to display the original column name)

5

SELECT column_name | aggregate_function [ AS output_column_name ], ...

FROM table_name, ...

[ WHERE predicate ]

[ GROUP BY [table_name.] column_name ]

[ HAVING predicate ]

[ ORDER BY output_column_name [ ASC | DESC ], ...

SQL Select Statement Syntax


• • • ••



table_name-Table name from which query rows and columns are extracted

WHERE

As part of a SELECT statement the WHERE clause can be used to predicate, or limit, the rows returned from the query.

Predicate-Expression to limit rows returned from the query

– node_name='BLACK'

– logical_mb>=1000

– platform_name IS NULL

– bytes_sent BETWEEN 100 AND 500

– nodes.node_name=occupancy.node_name (join criteria for selecting one value from two tables)

GROUP BY

The GROUP BY clause is used to summarize the output from a query by column name.

column_name-Rows to be grouped for resolving aggregate functions. This clause is generally used in conjunction with an aggregate function.

HAVING

The HAVING clause is used to apply a condition to an aggregated column to limit the returned output.

predicate-Expression to filter output of aggregated values before displaying. The HAVING clause is used to include a condition for an aggregate function.


• • •••



ORDER BY

The ORDER BY clause is used to format the output returned from a query based on a sorted column:.

output_column_name-Specifies output sorting, either in ascending (ASC) or descending (DESC) order. If more than one column_name is specified, first column has primary sort order, then second column, and so on.


• • • ••



Server SQL Interface

The Tivoli Storage Manager server has an SQL interface that can be used to access data from the Tivoli Storage Manager database. Anyone with the proper administrative authority can issue SQL commands.

6

Server SQL Interface

Administrative select command


• • •••



Administrative Select Command

SELECT Command

The select interface is consistent with relational database products and presents server information in the form of relational tables containing rows and columns. Select statements can be issued from the Tivoli Storage Manager administrative command-line interface.

System Catalog Tables

Three system catalog tables are implemented to assist the administrator in determining the information available.

SYSCAT.TABLES–Contains information about all tables that are available for querying with the select statement.

SYSCAT.COLUMNS–Describes the columns that reside in each of the tables.

SYSCAT.ENUM–For columns that have an enumerated data type this table defines the legal values for each enumerated data type and the ordering of the

7

• SQL select command–Consistent with other relational database products

– Invoked only from Tivoli Storage Manager administrative command line

–Read Only access to database

• System catalog tablesImplemented in Tivoli Storage Manager database to assist administrator

• Confirmation messageResource intensive queries can be canceled before executing

Administrative Select Command


• • • ••



different values for the type; an enumerated data type is a value that is assigned a numerical number rather than text.

Confirmation Message

For SQL queries that operate on large tables or require significant time and resources from the Tivoli Storage Manager server, a confirmation message is displayed indicating that the query is resource intensive and may take a lot of time to generate results. This gives the opportunity to cancel the query before resources are consumed or to run the query anyway. Since queries may tie up sessions for some time, they cannot be executed from the server console. This keeps the console session available for managing other administrative functions.


• • •••



Server SQL Processing

Virtual SQL Tables

Tivoli Storage Manager server information is mapped into virtual tables to make it available for SQL select statements. This is also true for dynamic information. For example, the SESSIONS table can be queried with a select statement to display client sessions that are currently logged on to the server.

Many forms of SQL select statements generate intermediate data before the final select output report is displayed. Functions causing this are, for example, nested select statements, order conditions, or matching criteria. This intermediate data is stored in temporary tables. Temporary tables reside in the Tivoli Storage Manager database in order to save memory resources. They are deleted upon completion of the SELECT statement.

Due to generating temporary tables, SQL activity requires a minimum of a 4 MB partition in the Tivoli Storage Manager database. For larger queries, substantially more free space is needed.

Since the SQL processing uses the database buffer pool and I/O resources, SQL select statements affect Tivoli Storage Manager server performance. Very complicated or long-running select statements can slow down server performance significantly during processing.

8

• Virtual SQL tables• Virtualize Tivoli Storage Manager server information in form of SQL tables

•The SESSIONS table (containing current client sessions) is an example

• Temporary tables• Used for buffering intermediate information, such as sort results or matching columns

• Stored in the Tivoli Storage Manager database

• Impact on TSM server• SQL activity requires minimum of 4 MB in Tivoli Storage Manager database (for large queries substantially more space is needed)

• Buffer pool and I/O usage affect server performance

Server SQL Processing


• • • ••



SQL Parameters and Commands

SQLDISPLAYMODE

The SQLDISPLAYMODE parameter controls the manner in which SQL data types are displayed. It can be set to one of the following values:

• Narrow–Column display width is 18; any wider string is displayed on multiple lines

• Wide–Column display width is 250

9

• Set SQLDISPlaymode• Specifies the output display mode for SQL query results

• SET SQLDISP NARROW | WIDE

• Set SQLDATETIMEformat• Controls the format of displaying SQL date-time types

• SET SQLDATETIME ISO | USA | EUR | JIS | LOCAL

• Set SQLMATHmode• Sets the mode in which SQL arithmetic is carried out

• SET SQLMATH TRUNCATE | ROUND

• Query SQLsession• Displays the current values of SQL parameters

SQL Parameters and Commands


• • •••



SQLDATETIMEFORMAT

This parameter controls the format in which SQL date and time data are displayed. One of the following formats can be chosen.

ISO–International Standard Organization

USA–IBM USA standard

EUR–IBM European standard

JIS–Japanese Industrial standard

Local–Site-defined (Currently JIS and Local are the same as ISO)

SQLMATHMODE

The SQLMATHMODE parameter defines the mode in which SQL arithmetic is carried out. There are two options:

Truncate–Decimal numbers are truncated

Round–Decimal numbers are rounded

QUERY SQLSESSION

The command QUERY SQLSESSION displays the current SQL settings:

query sqlsession


• • • ••



SQL Example in a Script

In this example an SQL statement is used in a script. This example is part of a script used to start reclamation. When scheduled, this script makes Tivoli Storage Manager reclaim the volumes with a lower percent utilization first rather than letting Tivoli Storage Manager randomly pick any volume for reclamation. A complete script using this SQL statement allows more volumes to be reclaimed in a shorter amount of time.

10

SQL Example in a Script

select VOLUME_NAME, pct_utilized, stgpool_name from volumes wherepct_utilized>0 and pct_utilized<65 and devclass_name!='DISK' andSTGPOOL_NAME = 'OFFSITEPOOL'

if (rc_notfound) goto reclaim30upd stg offsitepool rec=35

exit

reclaim30:select VOLUME_NAME, pct_utilized, stgpool_name from volumes where

pct_utilized>0 and pct_utilized<70 and devclass_name!='DISK' andSTGPOOL_NAME = 'OFFSITEPOOL'

if (rc_notfound) goto exitnowupd stg offsitepool rec=30

Exit

exitnow: exit


• • •••



ODBC Implementation

Using the ODBC Driver

Tivoli Storage Manager provides an ODBC driver for Windows. The driver supports the ODBC version 2.5 application programming interface (API). Because Tivoli Storage Manager supports only the SQL select statement (or query), the driver does not conform to any ODBC API or SQL grammar conformance level. After you install this driver, you can use a spreadsheet or database application that complies with ODBC to access the database for information.

On Windows platforms, an ODBC driver is available that allows you to use a relational database product, such as Lotus Approach, to query the database and display the results. Please refer to the readme.odbc.txt file for installation instructions. This file is available in the Tivoli Storage Manager ODBC driver installation package. Also, see the Tivoli Storage Manager for Windows Administrator’s Guide (GC32-0782).

11

ODBC Implementation


• • • ••



ODBC Configuration

Defining Data Sources

The ODBC data source administrator in the control panel can be used to add, remove, and configure data source entries. ODBC applications can then be used to connect to the Tivoli Storage Manager server.

Note: For 64-bit Windows XP and 2003 users, the Tivoli Storage Manager ODBC driver is only available in a 32-bit version. Therefore, you must use the 32-bit version of the ODBC data source administrator (odbcad32.exe) to configure Tivoli Storage Manager ODBC driver DSNs. This executable file is located in directory X:\WINDOWS\SysWOW64, where X: is the drive letter where Windows is installed (usually C:).

The Tivoli Storage Manager ODBC setup program will create a default user data source name (DSN) named Tivoli Storage Manager ODBC Sample User DSN. You can either update this entry or remove it and create one or more new DSNs to connect to your Tivoli Storage Manager servers. When creating new DSNs, use the Tivoli Storage Manager ODBC Driver.

12

ODBC Configuration

Setup utility icon in control panel configures:

•Communication•Server address•Port number

ODBC application:

•Opens server as ODBC datasource

•Links or imports server tables•Queries server using selectstatements

•Requires administrator user IDand password


• • •••



The ODBC data source setup panel for Tivoli Storage Manager has the following controls:

• Data source name (required)–Specify the name you want to use for the DSN. This is the name that will appear in the Name column when you run the ODBC data source administrator. The maximum allowed length is 255 characters.

• Description (optional)–This is a free-form text field where you may enter a comment or description about this DSN. The maximum allowed length is 255 characters.

• Language (required)–Select the language in which ANSxxxxx messages should be displayed. The available languages are those that were selected at the time the ODBC driver was installed.

Note: Only the ANSxxxxx messages are available in other languages. The ODBC driver dialogs and other non-ANSxxxxx text strings are currently displayed in English only.

• Administrator name (optional)–Specify the Tivoli Storage Manager administrator ID to use with this DSN.

• Fetch size (required)–Specify the number of records that the ODBC driver requests from the Tivoli Storage Manager server per fetch request. Larger values may improve performance, especially when fetching a large number of records. This value must be between 1 and 1000000 (inclusive). The default value is 50.

• TCP/IP address (required)–Specify the TCP/IP address of the Tivoli Storage Manager server you are connecting to. If you do not know which address to use, see your administrator.

• TCP/IP port (required)–Specify the TCP/IP port number of the Tivoli Storage Manager server you are connecting to. If you do not know which port number to use, see your network administrator.

• Enable trace (optional)–Selecting this box activates Tivoli Storage Manager ODBC driver tracing. Normally, this box should remain cleared. Tracing should only be performed at the request of Tivoli Storage Manager service or development.

• Trace file name (optional)–This field is available only when Enable trace has been selected. Here, you can specify a fully-qualified file name to which trace data is written. Tracing should only be performed at the request of Tivoli Storage Manager service or development.


• • • ••

Unit 7: ReportingTivoli Storage Manager Operational Reporting


Lesson 2: Tivoli Storage Manager Operational Reporting

Introduction

Tivoli Storage Manager Operational Reporting automates some of the monitoring tasks you typically perform manually. By generating reports and monitors, Tivoli Storage Manager Operational Reporting notifies you if a server requires attention.

Operational reports can be scheduled to run daily and are generated even if there are no problems. Operational monitors are special types of reports, and can be scheduled to run hourly. The monitors will send you a notification only if there are issues.

Operational Reporting does not maintain a separate database of information and is not a trending tool. Operational Reporting is included as part of the Tivoli Storage Manager for Windows server and is also available as a stand-alone package for a Windows server.

13

Tivoli Storage Manager Operational Reporting

•Set up•Post-installation•Configuration


• • •••



For information on installing the stand-alone package, see Installing the IBM Tivoli Storage Manager Operational Reporting Stand-alone Package in the IBM Tivoli Storage Manager for Windows Administrator’s Guide.


• • • ••



Tivoli Storage Manager Operational Reporting Overview

Minimizes the amount of time needed to administer Tivoli Storage Manager

– A daily report can be sent to you using e-mail.

– The report provides you with a summary of operational activity.

– The subject indicates if the server is running smoothly or if it needs attention.

– If server needs attention issues are identified and recommendations are provided.

– Reports and monitors are color coded to quickly identify warnings and errors.

– A monitor can notify you of issues by e-mail or with a message to your desktop.

Easy to use

– Runs on Windows and supports all Tivoli Storage Manager server platforms and versions.

14Tivoli Storage Manager Operational Reporting Overview

• Helps minimize the amount of time needed to administer Tivoli Storage Manager

• Easy to use • Reports and monitors can be viewed interactively, from a Web site, or

in an e-mail• Reports and monitors can be customized and extended• Custom report and monitor templates can be shared with others• Can automatically notify clients of missed or failed schedules by e-mail


• • •••



– Integrated with the Tivoli Storage Manager management console.

– Small number of settings with simple interface to access them.

– Defaults are provided.

– Reports and monitors can be viewed interactively, from a Web site, or in e-mail.

– Reports and monitors can be customized and extended.

– Custom report and monitor templates can be shared with others.

– Can automatically notify clients of missed or failed schedules using e-mail.


• • • ••



Operational Reporting Installation

15

Operational Reporting Installation• Installs automatically with Tivoli Storage Manager server in a new installation

•For upgrade uninstall the previous version

Operational Reporting objects installed in the

TSM MMC


• • •••



Tivoli Storage Manager Management Console

This slide shows how to add a new Tivoli Storage Manager server to your reporting tools.

16

Tivoli Storage Manager Management Console


• • • ••



Add Tivoli Storage Manager Network Resources

This screen capture shows the interface for adding a new Tivoli Storage Manager computer. You can have the Tivoli Storage Manager Operational Reporting tool find other Tivoli Storage Manager servers on the network if you are not sure of the IP or computer name.

17

Add Tivoli Storage Manager Network Resources

Click New to add a new Tivoli Storage Manager computer


• • •••



Add a Tivoli Storage Manager Computer

18

Add a Tivoli Storage Manager Computer

•Specify the computer’s name•Select the operating system•Select Include Report Features•Select TSM Web Administrator•Specify TSM server information

•Server name•TCP/IP Address•TCP/IP Port

•One TSM server instance per machine is typical more instances can be added as required.




Operational Report List

From this screen it is easy to add an operational report to any Tivoli Storage Manager server managed in the MMC console.

19

Operational Report List

•Select a computer and server instance•Click Add to add a new Operational Report




Add Report and Report Details

This screen allows you to choose which of the out-of-box reports you would like to use and when you want them to be run.

20

Add Report and Report Details

• Type in the report name• Select check boxes of the sections

you want to include in the report• Specify timing information• Web versions refer to the number of

copies of this report you want to appearin a summary Web page

• Select a report section and clickSettings… to customize it

• Select Add to extend the set of sectionsusing a custom SQL select statement




Add Report Recipients

If you are sending multiple reports to the same recipients, export your recipient list from one report and import it into another.

• Format can be text, HTML, or URL. URL will send an e-mail link to the Web summary page.

21

Add Report Recipients

Exporting your recipient list will make it easy to use the same list in other reports.




Add Report Missed Schedule Notifications

Since you as the Tivoli Storage Manager administrator may not be able to access all clients, use this tool to send out e-mail reports of missed or failed schedules to the machine’s administrator. That administrator can then begin troubleshooting a missed or failed schedule from the client side.

22

Add Report Missed Schedule NotificationsOperational reporting can automatically send clients an e-mail message when their schedule has been missed or failed.

• Click Query• Select Failed/Missed check boxes• Click Node Update• Click Nodes Update to send all

changes back to the TSM server• Information is stored in the node

contact field• Click Failed or Missed to customize

the message to be sent

Tip: The Scheduled nodes only check box will cause the Query button to return only nodes that are associated with a schedule.

Tip: To activate automatic notification you need to select the corresponding report sections in the Report Detailstab.




View After Reports Have Been Defined

23

View After Reports Have Been Defined

Each server can have multiple reports




Operational Monitor List

Operational monitors can be created to notify the Tivoli Storage Manager administrator of events and errors close to the time when they occur. The operational report is generally used for daily reporting whereas the operational monitor is used to keep abreast of current issues that might arise on your Tivoli Storage Manager server. For instance, if there are more tapes in your on-site storage pool than there is space for in the library, tape mounts will be necessary. You could have an operational monitor check every 15 or 30 minutes for mount requests and send e-mail to the desk of the operator.

24

Operational Monitors List

•Reports and monitors are configured in the same way•Click Add to add a new operational monitor




Add Monitor Details

This screen works the same way that the screen in the reporting tool works with the exception that there are not nearly as many built-in reports. The monitors are intended to be customized for your environment and issues.

25

Add Monitor Details

• Type in the monitor name• Specify timing information• Specify the number of versions of

the monitor results to display if using the Web summary page

• Select Settings… to customize themonitor

• Timing information can help optimizequeries

• Use extensions – an extension willonly be executed if a notification ruleis triggered in the custom summary




Add Monitor Recipients

The recipient list is managed the same way as an operational report.

26

Add Monitor Recipients

• Click Import… to retrieve e-mailinformation exported earlier.

•Format can be Text, HTML, orURL. URL will send an e-maillink to the Web summary page.




Add Monitor Desktop Notify

The operational monitor can send a message to the desk of an operator or administrator to make them aware of issues or conditions on the Tivoli Storage Manager server.

27

Add Monitor Desktop Notify

•Operational monitoring can send a message to your desktop to alert you of any issues.




Configure Reporting Service

• You can start or stop the reporting service.

• You can update the service startup type.

• You can enable and review log and trace information.

28

Configure Reporting Service

• Set the Windowsservice details




Configure Operational Reporting E-mail Account

Set up the e-mail address and SMTP server information from this tab. You may want the e-mail administrator to set up an account, such as BACKUP ADMIN, so that all recipients will know that the e-mail contains important information about their client backups.

29Configure Operational Reporting E-mail Account

• The reporting servicewill use this mail account to send e-mail.

• The SMTP server TCP/IP address should be the fully qualified TCP/IP address.

• You can send a test message to verify that the SMTP server settings are correct.




Configure Web Summary

If you choose to send a URL to report recipients, this tab has the default location from which the reports will need to be published.

30

Configure Web Summary

• When enabled, operationalreporting will produce a Webpage with links to reports

• The Web versions parameter on the report’s property sheetdetermines how many versionsto display

• WebSphere (IBM Web HTTPserver) and IIS use different default pages

• You can create a log file for useby a Tivoli TEC log file adapter which can send events to TEC




Custom Summary Section

To configure the custom summary or any of the configurable report extensions, simply select the menu that you would like to change and select Settings.

31

Custom Summary Section

• The custom summary can betailored to your environment

• The custom summary isavailable for reports andmonitors

• Select Custom Summaryand click Settings…

• Rules can be defined in thecustom summary sectionthat will alert you to variousconditions

Note: A monitor differs from a report in that it only includes custom summary, timing, and extension sections. It only runs extensions and notifies you if there are issues. Monitors typically run quickly and frequently.


• • •••



Report Templates

32

Report Templates

•Settings are saved in a custom XML template file.•Share template files between reports and with others.


• • • ••



Report Statements

33

Report Statements•This dialog box will allow you to edit a line in the custom summary section•Specify a unique SQL variable, report label, and an SQL select statement•Select statements can be tested•Click Edit Rule… if you want to define a rule that will trigger a notification when a value matches the criteria set


• • •••



Report Rules

34

Report Rules

• Create a notification rule• When the rule is triggered in a report you will get an entry in the

Issues and Recommendations section• When the rule is triggered in a monitor you will be notified of the

situation through e-mail and with a message to your desktop ifenabled


• • • ••



Notification Options

35

Notification Options

Phone services make it easy to get notified any time, any place


• • •••



Report Extensions Using SQL Statements

Customized extensions can be added to the standard report extensions. You can use command-line queries or select statements.

36

Report Extensions Using SQL Statements

•Click Add to extend the set of sections•Click Settings to edit an extension•Click Advanced to build an SQL select statement•Create once, use in any report•Share extensions with others (*.ext files)•A section is dynamically formatted based on the results of the SQL select•Extensions can be deleted•Select check box to include new section in a report


• • • ••



Test Extension Select Statements

37

Test Extension Select Statements


• • •••



Select Statement Builder

Use the illustration above to help identify the dialog box for these tasks.

38

• Advanced featurehelps you to build SQLselect statements

• Select table columns• Type over Alias to

update column header• Cut,Copy,Paste,Delete,

Select, multi-levelUndo,Redo, Dragbetween workspace andactual select statement

• Per server platform,version, language,table caching for fastinteraction

• Test before saving

Select Statement Builder


• • • ••



Creates Interactive Reports

• Expand the tree

• Right-click report

• Choose Refresh

• Report can be sent as an e-mail

• Click server name to start Web Administration

Note: On manual refresh e-mail is not sent to recipients but missed and failed notifications are sent.

39

Interactive Reports


• • •••



Interactive E-mail Reports

40

•You can interactively send e-mail using the e-mail icon on the toolbar

•You can include reports and other attachments•You can click Query Nodeto look up the addresses used to notify users of missed schedules•You can click Import to look up addresses used to send reports to adminstrators•Click Feedback to send feedback to TSM development

Interactive E-mail


• • • ••

Unit 7: ReportingTivoli Storage Manager Reporting with Tivoli Data Warehouse


Lesson 3: Tivoli Storage Manager Reporting with Tivoli Data Warehouse

Tivoli Data Warehouse:

TDW replaces Tivoli Decision Support and builds a central data warehouse (CDW) containing data from the entire IT infrastructure (including Tivoli Storage Manager environment).

A subset of the CDW, called a data mart, can be used to analyze IT costs, infrastructure, performance and do historical trending.

Vendors can write their own packages to interface with TDW for their software applications, and customers can also develop their own packages as well. These are referred to as Enablement Packages.

TDW contains several software components and can run on several platforms.

41Tivoli Storage Manager Reporting with Tivoli Data WarehouseObjectives:•Discuss Tivoli Data Warehouse architecture•Describe components of TDW•View examples of Tivoli Data Warehouse reports for Tivoli Storage Manager


• • •••



Tivoli Data Warehouse Architecture

42

Tivoli Data Warehouse Architecture• Environmental data is extracted from various data sources,

Tivoli Storage Manager DB, for example, using an ETL (extract, transform, and load) program.

• All data is then loaded into the CDW (Central Data Warehouse).

• Another ETL program then loads a subset of this data into a data mart database. For example, Tivoli Storage Manager has its own data mart database.

• Reports are then created from the data mart, or other tools can be used for analysis of the data mart.


• • • ••



Tivoli Data Warehouse Components

43

Tivoli Data Warehouse Components•Control Server

•Central Data Warehouse

•Data marts

•DB2 warehouse agents

•Crystal Enterprise (for reporting)

•Warehouse enablement packs


• • •••



Tivoli Data Warehouse Software Components

44

Tivoli Data Warehouse Software Components

•IBM DB2 UDB Enterprise Edition•DB2 Warehouse Manager (for remote agents only)•Crystal Enterprise Reporting•Tivoli Data Warehouse•For Tivoli Data Warehouse with Tivoli Storage Manager reports

• The Tivoli Storage Manager Web Enablement Pack• The ODBC component of the Tivoli Storage Manager client


• • • ••



Tivoli Data Warehouse Report Examples

Note: These report examples are included to show the type of reports available using TivoliData Warehouse.

45

TDW TSM Report Example 1


• • •••




46


For TSMServer1


• • • ••




47



• • •••




48



• • • ••




49



• • •••



Tivoli Data Warehouse Reference Materials

50

Tivoli Data Warehouse Reference MaterialsReference Information

Web Site:

http://www-306.ibm.com/software/tivoli/products/data-warehouse/

http://publib.boulder.ibm.com/tividd/td/TivoliDataWarehouse1.2.htmlfor TDW

http://www-1.ibm.com/support/entdocview.wss?uid=swg27004654(SG24-7100)

http://publibb.boulder.ibm.com/Redbooks.nsf/RedbookAbstracts/sg247100.html?Open

These URLs are current as of this document. You can also search www.ibm.comfor Tivoli Data Warehouse and find more information.


• • • ••



Student Exercise


51

Student Exercise


• • •••



Review Questions

1. True or False? The Tivoli Storage Manager SQL interface uses the industry standard interface.

2. True or False? The select statement allows you to access data from only a single table at a time.

3. True or False? You have no control over the order in which the data is returned from an SQL select.

4. True or False? The server SQL interface allows you to read and update the Tivoli Storage Manager database.

5. True or False? Even the most complex SQL query has minimal impact on the Tivoli Storage Manager database performance.

6. True or False? The ODBC driver must be installed before any SQL select commands can access the Tivoli Storage Manager database.


• • • ••





• • •••



Review Answers

1. True or False? The Tivoli Storage Manager SQL interface uses the industry standard interface.

False.

2. True or False? The select statement allows you to access data from only a single table at a time.

True.

3. True or False? You have no control over the order in which the data is returned from an SQL select.

False.

4. True or False? The server SQL interface allows you to read and update the Tivoli Storage Manager database.

False.

5. True or False? Even the most complex SQL query has minimal impact on the Tivoli Storage Manager database performance.

False

6. True or False? The ODBC driver must be installed before any SQL select commands can access the Tivoli Storage Manager database.

False.


• • • ••



Summary

52


Describe relational database access techniques

Access Tivoli Storage Manager database information

using the SQL administrative interface

Implement the Tivoli Storage Manager Operational Reporting

Manager

Describe Tivoli Data Warehouse

• • • ••

8-1

Unit 8: Problem Determination Concepts



• • • ••



Introduction

Problem determination can be complicated. As a Tivoli Storage Manager environment gets more complex, getting to the root of a problem becomes more difficult. Be careful not to mistake the symptoms of a problem for its cause. This unit will cover troubleshooting and client and server tracing.

Objectives

2


• Troubleshoot Tivoli Storage Manager.• Implement the server trace function for problem determination.

• Implement the client trace function for problem determination.

• Assist problem determination for a SAN by implementing cyclic redundancy checking (CRC) in Tivoli Storage Manager.

• Explain instrumentation data and its application to problem determination.


Unit 8: Problem Determination ConceptsProblem Determination Methodology


Lesson 1: Problem Determination Methodology

Steps for Problem Determination

The graphic on this page lists the steps to diagnose a problem with the Tivoli Storage Manager environment. Many of the steps require use of the available query commands to analyze the situation. This lesson will provide tips on dealing with specific problems.

3

Steps for Problem Determination

Consider these points while diagnosing a problem:

• What is the problem?• Where did it occur?• When did it begin?• What action was being performed?• Were any messages issued?• How frequently does the error occur?• Verify that devices are still accessible to the system and to Tivoli Storage Manager.

• Search the online Knowledge Base at IBM.com to match error messages or problem descriptions.

• Test other operations to better determine the scope and impact of the problem. This may also help determine if a specific sequence of events causes the problem.




Locate the Problem

Various problems can crop up in a variety of places. From the moment the client requests a backup and begins to read the data from disk, to the point that the Tivoli Storage Manager server stores the data in the storage pools, there are many possible points for error.

On the client, those points are numerous. For example:

• The disk and the organization of the disk (RAID, striping, and so forth)

• The connection from the disk to the client (fibre channel, SSA, SCSI, and so forth)

• The adapter in the client server

• The memory in the client server

• The network adapter (100 Mbps, 1 GB, ATM, and so forth)

• The processor capacity

• The operating system

• The communication protocol (TCP/IP)

4

Locate the Problem

•Many areas to look at when a problem occurs•Movement of data starts at the disk on the client and moves to either a disk or tape media on the server

•Data moves through many components, base (in the end systems) and transfer components

•Bottlenecks can occur in any of these•Points of consideration

–Drivers–Adapters–Network issues–Memory–Routers–Other applications




Many of the same possibilities exist for the Tivoli Storage Manager server.

Another factor is the drivers or microcode needed to run these components.

Between the Tivoli Storage Manager server and the Tivoli Storage Manager client, two network configurations are possible. The connection can be done using LAN or SAN. Refer to the LAN-free backup topic discussed earlier.

Both of the networks have their own configuration, microcode, hardware cabling, and many other possible differences. So whenever there is a problem, start with a plan.




Hints and Tips for Libraries, Drives and Device Drivers

Device driver problems have many causes. The trouble may be with the operating system, the application using the device, the device firmware, or the device hardware itself.

When a problem arises, first ask: Have any changes been made recently to your backup environment? What changes were made since the hardware worked? Any changes between the machine trying to use the device and the device itself could be suspect.

5Hints and Tips for Libraries, Drives and Device Drivers

• Has the operating system changed?• Has the host bus adapter (HBA) or SCSI adapter connecting to the device been changed, updated, or replaced?

• Has the adapter firmware changed?• Has the cabling between the computer and device changed?

• Are any of the cable connections loose?• Has the device firmware changed?• Are there error messages in the system error log for this device?

• Has a device driver changed?




Diagnosing a Client Problem

Use the suggestions in the graphic above to narrow down the possibilities of where the problem exists.

6

Diagnosing a Client Problem•Examine any error messages that were issued.•Examine the server activity log messages for this session.

• Is this an error connecting (communicating) to the server?

•Were client options changed?•Were policy settings changed on the server?• Is the client being run with the QUIET option?•Verify INCLUDE/EXCLUDE syntax and ordering.•Was this the correct Tivoli Storage Manager server?• Identify when and where the problem occurs.• If the problem can be reproduced, try to minimize the circumstances under which it can occur.

•Collect documentation.




Tivoli Storage Manager Client Error Collection

One way to gather information about the system is to use the available query commands.

Development has collected the most important information required by most customers, and has created a new command–query systeminfo.

It is a client command, and the output will give the following information:

• dsmoptfile – The contents of dsm.opt file.

• env – Environment variables.

• errorlog – The Tivoli Storage Manager error log file.

• file – Attributes for the file name that you specify.

• inclexcl – Compiles a list of include-exclude in the order in which they are processed during backup and archive operations.

• options – Compiled options.

• osinfo – Name and version of the client operating system.

• policy – Policy set dump.

• msinfo – Windows system information (output from msinfo32.exe).

7

Tivoli Storage Manager Client Error Collection

It is important to know what data can be found in the client.

storagepool

query systeminfo

sched.log

dsierror.log

dsmerror.log

(nt)event.log

dsmwebcl.log


• • •••



• registry – Windows Tivoli Storage Manager-related registry entries.

• systemobject – Windows system object information.

• cluster – Windows cluster information.

This command is intended primarily as a diagnostic aid. With the console option, the output is not formatted to accommodate screen height or width. Therefore, the console output may be difficult to read due to length and line-wrapping. In this case, use the query systeminfo command with the option for output written to a file so that the results can later be submitted to technical support. Also, if the query systeminfo command is run from the dsmc command line, the output will generate the file in the baclient directory.


• • • ••



Diagnosing a Server Problem

For a server problem, use the steps in the graphic above to try to isolate or resolve the problem.

8

Diagnosing a Server Problem

•Check the server activity log.•Check HELP for Tivoli Storage Manager messages issued for the problem.

•Can the problem be duplicated?•Is the problem an error reading or writing to a device?

•Have any server options or settings been changed?•Did a scheduled client operation fail?•Did the server run out of space?•Are connections by clients or administrators failing?

•Collect documentation.


• • •••



Tivoli Storage Manager Server Error Collection

The query system command is the Tivoli Storage Manager server command for data collection from its large storehouse of information and file locations. This command should provide most of the necessary information to start a proper determination. The output of query system is a compilation of these server queries:

• query association

• query copygroup

• query db

• query dbvolume

• query devclass

• query domain

• query log

• query logvolume

9

Tivoli Storage Manager Server Error Collection

Q ACTLOG

dsmserv.err

OS Eventlog

dsmaccnt.log

data validation (CRC)

Set TapeAlertMSG

It is important to know what data can be found in the client.

storagepool


• • • ••



• query mgmtclass

• query options

• query process

• query schedule

• query session

• query status

• query stgpool

• query volume

Two other commands give very helpful information: q nodes and q filespace format=detailed.

These commands can show how recently each node and file space was backed up. The q filespace command can locate unnecessary data in the database, which can cause problems.


• • •••



Problems with Data Flow in a Storage Area Network

The storage area network (SAN) connection introduced a whole new realm of potential problems into the Tivoli Storage Manager environment. When all disks are locally attached, however, and the data travels over the LAN and SCSI bus, the data transfer is well protected.

In TCP/IP there is a type of handshaking that occurs for every package sent from client to server. On the SCSI bus, all bits are sent in parallel to the device under the control of the SCSI protocol.

In a SAN, however, the data transfers serially. In the original design of the fibre channel protocol (FCP), no built-in system verifies that the data is received correctly at destination. For this reason, Tivoli Storage Manager created the cyclic redundancy checking (CRC) feature. The function is exactly the same as the handshaking in TCP/IP. It creates a hash total of the data, and then that checksum is checked upon arrival at the destination. This feature is also incorporated into SAN hardware in some environments.

There are several locations in Tivoli Storage Manager to enable and disable CRC checking. The checking feature is not always on because it degrades performance.

The data transfer between client and server, client to storage pool, and storage pool to storage pool can be checked.

10Problems with Data Flow in a Storage Area Network

Enable data validation (CRC)

storagepool


• • • ••



Validating a Node's Data

After Tivoli Storage Manager completes the data validation, the CRC values generated between the client and server in the current session are discarded.


• • •••

Unit 8: Problem Determination ConceptsUsing Client and Server Tracing


Lesson 2: Using Client and Server Tracing

Running a Trace

Always contact a service representative before running a trace to solve a problem. After you have followed the proper diagnostic procedures and contacted Tivoli Storage Manager Support assistance, and user specifications have been checked for accuracy, you should have the information described in the following section.

Reporting a Problem

Please have the following information ready when you report a problem:

• The Tivoli Storage Manager server version, release, modification, and service level number. To get this information, enter the query status command at the Tivoli Storage Manager command line.

11

Running a Trace

On advice from supportRun a TRACE

storagepool


• • • ••



• The Tivoli Storage Manager client version, release, modification, and service level number. To get this information, enter dsmc at the command line.

• The communication protocol (for example, TCP/IP), version, and release number you are using.

• The activity you were attempting when the problem struck, listing the steps you followed before the problem occurred.

• The exact text of any error messages.

Running Trace Commands

Your service representative may ask you to run trace commands to perform diagnostic functions. The output from the trace commands is classified as diagnosis, modification and tuning information.

There is also a trace facility guide available to help you with the correct statements or options. The basic steps are provided in this overview.


• • •••



Server Trace Facility

Diagnosing Server Problems with Trace Commands

Administrators with analyst privileges can diagnose server problems by enabling, disabling, and controlling the server trace activity. Any administrator with general privilege can display information about trace activities. Authorized administrators can trace Tivoli Storage Manager events by initiating trace functions from the server screen, or trace functions as the server is started by placing the trace commands in the server options file.

Trace output is displayed on the server screen or sent to a file specified with the trace begin command. Tracing is achieved by enabling and disabling trace classes. Each trace class produces diagnostic information for a specific functional area of the server. Some trace classes generate extremely large amounts of data, while others produce relatively small amounts. Depending on which trace classes are enabled, expect moderate to severe performance degradation in the server. For this reason, perform tracing on the server in a controlled environment with general user access limited to those client or administrator nodes necessary to recreate the problem, and only at the request of an IBM service representative.

12

Server Trace Facility

Example (to trace SCSI library):TRACE ENABLE PVR MMSTRACE BEGIN TRACEFILE.OUTTRACE DISABLETRACE END

1. query2. list3. enable4. begin5. flush6. end7. disable storage

pool


• • • ••



Trace Enable

Use the trace enable command to activate server trace classes. Tracing does not start until a trace begin command is issued. There are aggregate trace classes defined to enable multiple trace classes, grouped together for logical function tracing.

Trace Begin

Use the trace begin command to start server tracing. Tracing can adversely affect the performance of the Tivoli Storage Manager system. The trace begin command should only be used at the request of an IBM service representative. Trace output is displayed on the server screen or sent to a file that you specify with the trace begin command.

Trace Flush

Use the trace flush command to write any trace records in the trace buffer to an output file. The trace buffer writes its contents to a file when it is full. Unless you use this command, the contents in the buffer are not displayed except when the trace end command is issued. If you use the trace flush command, be sure to issue it before the trace end command to prevent loss of trace data in the last trace buffer.

Trace Disable

This command deactivates trace classes specified in the trace class list. Trace classes are deactivated immediately. However, tracing continues for other active trace classes until the trace is ended with the trace end command.

Trace End

This command ends all tracing but will not deactivate the trace classes. See trace disable above.

Trace List

This command lists all common aggregate trace classes.


• • •••



Query Trace

This command queries the status of enabled trace classes and trace activity.

There are some additional options available. For instance, the tracesegsize option is used to split trace files, and the tracemax unit now takes units in MB and not KB. The valid range of values for the tracemax function is 1 (MB) to 2047 (MB). There is no default value for tracemax.


• • • ••

Unit 8: Problem Determination ConceptsClient Trace Facility


Lesson 3: Client Trace Facility

The client trace facility helps you diagnose client problems by tracing specific events. The trace facility uses trace functions from the options file or the command line. Trace output is displayed on the client screen or sent to a specified file. You can use tracing while in interactive mode or while using the graphical user interface.

Each trace flag enables tracing for a specified functional area of the client. Some trace flags generate large amounts of data, while others produce relatively small amounts. Depending on which trace flags are enabled, you can expect small to moderate performance degradation in the client.

From a performance standpoint, you should not keep your trace settings activated if you are in a stable environment and are not experiencing problems. Trace routines require processing time and could slow down the response times of your workstation applications. If a problem occurs, you can always activate the trace routines to gather trace information for diagnostic purposes. Call your IBM service representative for assistance in diagnosing Tivoli Storage Manager problems.

13

Client Trace Facility


• • •••



The client trace facility provides the following options:

• tracemax

• tracefile

• traceflags

The TRACEFLAGS option lets you set specific trace flags. For example, if you are having a problem with the incremental command, you might want to turn on the ERROR trace flag when you execute the command. The ERROR trace flag captures all severe error messages. You can turn it on by entering:

dsmc incremental –traceflags=error

You can turn on all of the trace flags by using the ALL parameter. You can also use the ALL parameter with exceptions. The ALL parameter enables all trace flags. Be careful using the all trace flag because of the amount of data created and the negative performance issues you will encounter.

To turn on particular trace flags while you are using the graphical user interface, specify the TRACEFLAGS option when you start the GUI session:

dsm -traceflags=general (Windows)dsmj -traceflags=general (UNIX)

You can include TRACEMAX, TRACEFILE, and TRACEFLAGS in your client option file to enable these options whenever you start a Tivoli Storage Manager session. The options can be overridden during a session by using the same options as in a Tivoli Storage Manager command.


• • • ••




14

• Specific client activities can be traced to diagnose problems.

• Trace functions can be initiated from the command line, GUI, or by putting commands into the client options file.

• Output can go to the client console or to a specified file.

• Expect light to heavy performance degradation.

• Tracing is achieved by enabling trace flags.

• 63 tracing flags are available.

• Trace functions can be initiated or flags changed while the client is running with the DSMTRACE function.



• • •••



Client Trace Options

The first two options in the graphic can be understood from the information in the graphic. The last option is explained in more detail below.

TRACESEGSIZE

You can also choose that the trace be automatically split up by the Tivoli Storage Manager backup-archive client into smaller segments. The segments can range from 1 - 1000 MB by specifying the following in the client options file: TRACESEGSIZE <trace segment size in MB>.

This can be useful if you need to collect a large amount of trace information and want to easily manage sending the segments to IBM for analysis. For example, to specify a trace segment size of 10 MB:

tracesegsize 10

If you use the TRACESEGSIZE option, the trace file segments will be named using the tracefile option with an additional extension using the segment number. For example, trace.out.1.

15

• tracemax=sizethe maximum size of the trace file

– range is 0–10000 KB– default is 0 which disables the maximum

• tracefile=filespec• the name of the trace output file • default is standard output

• tracesegsize=size in mbdetermines the maximum size of a file segment

Tracing Options


• • • ••



Trace Options - Continued

Dynamic client tracing

A new command-line utility, dsmtrace, is available to enable tracing, disable tracing, or change trace flags while the client is running. See the Problem Determination Guide for information about using this utility.

Note: Tracing is an advanced diagnostic feature intended for use only at the recommendation of IBM support and development, or as outlined in the Problem Determination Guide.

16

traceflags = flag1, flag2, flag3,– Set trace filtering.– Commas separate flags.– 63 flags available.– Flags can be separate or grouped.

DSMTRACE utility is used to enable or disable tracing while the client is in session.

Tracing Options (Continued)


• • •••

Unit 8: Problem Determination ConceptsClient and Server Instrumentation


Lesson 4: Client and Server Instrumentation

Instrumentation data is only valuable if you understand what it means. This lesson will give you a basic understanding of the data produced from client and server instrumentation enablement. Another reason to introduce you to this information, just as with tracing, is to produce the result that Tivoli Storage Manager support requires.

testflag for Client Instrumentation

Beginning with v5.1 the testflags PERFORM and INSTR_CLIENT_DETAIL are no longer available. From that version through this version, those tags were replaced with a testflag for use with collecting instrumentation information. This information can be valuable in diagnosing the location of a problem and can even help identify a component setting that needs to be increased or tuned.

The testflag can only be used from the command line or scheduler and is not usable in the GUI interface, the API or the TDPs. If you start a process with this testflag set on and want to stop the accumulation, access the server and cancel the session running with the option turned on. It is also best to use this with a batch type command line so the instrumentation will not be in effect after the process has completed.

17

• Provides client performance instrumentation statistics per thread• Only shows threads with instrumented activities• Also includes command, options and the statistics information• Enable using:

-testflag=instrument:detail (at the command line)testflag instrument:detail (in the option file)

• Example of using this option on the command line:dsmc s –testflag=instrument:detail filename

• Output is appended to the file – dsminstr.report• Cancel the session from the server to see result without command

completing• Command line or scheduler only. Not available in GUI, Web Client, API or the

TDPs

testflag for Client Instrumentation


• • • ••



Example Output for Client Instrumentation

The information in the graphic above is from a selective backup while the testflag for client instrumentation detail is set on. The actual information in the file, or seen on the console, would show the individual commands, values of the client options, and panels of instrumentation data.

18

Example Output from Instrumentation------------------------------------------------------------------TSM Client final instrumentation statistics: Sun Jan 16 20:07:51 2005

Instrumentation class: Client detailCompletion status: Success------------------------------------------------------------------Detailed Instrumentation statistics forThread: 276 Elapsed time 2.023 secSection Actual (sec) Average(msec) Frequency used------------------------------------------------------------------Process Dirs 0.000 0.0 0Solve Tree 0.000 0.0 0Compute 0.000 0.0 2BeginTxn Verb 0.000 0.0 1Transaction 0.010 10.0 1File I/O 0.000 0.0 4Compression 0.000 0.0 0Encryption 0.000 0.0 0CRC 0.000 0.0 0Delta 0.000 0.0 0Data Verb 0.000 0.0 2Confirm Verb 0.000 0.0 0EndTxn Verb 0.360 360.0 1Sleep 0.000 0.0 0Thread Wait 1.582 395.5 4Other 0.071 0.0 0

------------------------------------------------------------------


• • •••



Client Instrumentation Detail Sections

The client setup line in the detail information deals with session setup and there is not much that can be tuned or changed in this area. The setup occurs once for the session.

Process dirs is a client-to-server call for file lists and backup status information. A high utilization of this process during the session will normally indicate a client issue. A large amount of time for process dirs is often tied to clients with many small files.

The solve tree process performs comparisons to find differences in the directory structure. Incremental backups always keep the directory structure current. For selective backups, the Solve Tree would indicate directories that need to be backed up.

Compute is used for transfer and throughput sizes and is used by estimation.

19

Client Instrumentation Detail Sections•Client Setup

Initial processing including log in, authorization and query for policy set and file system information.

•Process DirsProcessing file and directory information before backing up or restoring any files. For incremental backups, it includes querying the server for backup status information. For classic restore, it includes retrieving the file list. For no-query restore this is not used.

•Solve TreeFor selective backup, determining if there are any new directories in the path that need to be backed up. This involves querying the server for backup status information on directories. This can belarge if there is a large number of directories.

•ComputeComputing transfer and throughput sizes.


• • • ••



Client Instrumentation Detail Sections, continued

Transaction is used to count and store the time used for extraneous processes not included in any of the other instrumentation categories. As an example, the time used to display information when in verbose mode is collected in the Transaction detail.

BeginTxn Verb starts a transaction to the server. This, as mentioned in the graphic, will be directly related to the TXNGROUPMAX and TXNBYTELIMIT. If you are backing up 512 two kilobyte files and the TXNGROUPMAX equals 256 with a TXNBYTELIMIT of 25600, then there should be a value of two for the BeginTxn Verb. This comes from the TXNBYTELIMIT not being exceeded and dividing the number of files by the TXNGROUPMAX (512/256=2).

File I/O is a request to the OS to open, read, write or close a file. Many times the value of File I/O on a backup of small files will be twice the number of files, which is an indication of the file open and then the file read.

Compression should be negligible for the instrumentation detail because it is advised to accumulate instrumentation data with compression turned off.

Encryption is the process of enforcing privacy by using a key to hash the data.

20

Client Instrumentation Detail, continued• Transaction

A general category to capture all time not accounted for in the other sections. Includes file open and close time, which can be large, especially during restore of many small files. Also includes message display time which can be large if running in verbose mode or with detailed tracing options enabled. The average time for this section is not meaningful.

• BeginTxn VerbFrequency used indicates the number of transactions that were used during the session. The txngroupmax and txnbytelimit will directly affect this value.

• File I/ORequesting data to be read or written on the client file system. Includes the opening and closing of files.

• CompressionCompressing or uncompressing data.

• EncryptionEncrypting or decrypting data.




Client Instrumentation Detail Sections, continued

Use CRC to compute a value based on part or all of the data in a message. Use CRC to include or store the data so the computation can be applied to the data later to uncover whether the data has changed. This is used to provide a higher level of data integrity.

Delta is only used during a subfile backup. It holds values for the time and number of times used to determine which data has changed between the base reference file and the current file.

The data verb section through the endtxn verb section are on the server side. High values in the data verb will point toward a server issue.

The EndTxn Verb will, in most cases, match the BeginTxn Verb value. For each begin, there should be an end.

21

Client Instrumentation Detail, continued• CRC

Computing or comparing CRC data.• Delta

Adaptive subfile backup processing, including determining the changed file bytes or blocks.

• Data VerbSending or receiving data to and from the communication layer. High data verb time prompts an investigation of server performance and thecommunication layer (includes elements of both the client and server).

• Confirm VerbDuring backup, sending a confirm verb and waiting for a response to confirm that the data is being received by the server.

• EndTxn VerbDuring backup, waiting for the server to commit the transaction. This includes updating the database pages in memory, and writing the recovery log data. For backup directly to tape this includes flushing the tape buffer.




Server Instrumentation

Server instrumentation detail is analogous to the client side information. The two are performed based on threads. In many cases, looking at the threads for both sides, client and server, while looking for high thresholds, will help in diagnosing problems.

The server instrumentation is turned on using the administrative command-line client or at a foreground console (not a production scenario). Once turned on, the server will monitor and produce information on the threads created and store the information in memory. This information can be sent to a file when the instrumentation ends using the File parameter. This is the easiest method to collect the data.

Producing the information can degrade the performance of the server and should not be enabled for extended time periods. Usually the information will be collected for less than 24 hours.

22

Server Instrumentation•Provides server or storage agent performance instrumentation by thread

•Shows threads with disk, tape or network I/O operations•Data stored in memory and lost when the server or storage agent is halted

•To initiate: INSTrumentation Begin [Maxthread=#]• Default maximum number of 1024 threads instrumented

•To end: INSTrumentation End [File=filename]• Default output is sent to console or administrative display

•Look for threads where the majority of time is not in wait time

•Designed to be used for less than 24 hours•Use as a problem-solving diagnostic tool




Server Instrumentation

This is the output from a selective backup to a disk storage pool. You can see the different facts of the processes where data was collected. Not shown are the individual threads and for Logical Volume Manager (LVM), disk process, network and others. This is the summary information at the end of the accumulated data after the end of collection and movement of the data to a file.

23

Example Output from Server Instrumentation

TOTAL SERVER SUMMARYOperation Count Tottime Avgtime Maxtime InstTput RealTput Total KB----------------------------------------------------------------------------Disk Read 17 0.191 0.011 0.030 356.0 1.2 68Disk Write 81 0.650 0.008 0.150 1046.2 12.4 680Network Recv 385 34.360 0.089 31.025 1.3 0.8 43Network Send 21 0.040 0.002 0.010 25.0 0.0 1Acquire Latch 1884 0.340 0.000 0.340Acquire XLatch 2606 0.000 0.000 0.000Thread Wait 77 164.918 2.142 31.235

Instrumentation output complete.


• • • ••



Diagnostics with Instrumentation Data

Use the collected information to determine whether the client, server, or network is presenting the problem with throughput or transfer. Many times, you must also concurrently accumulate performance data from the OS to see if another process or lack of adequate resources was the problem.

24

Diagnostics with Instrumentation Data• Find the bottleneck

• Collect instrumentation data to isolate the cause–Server (high DataVerb and EndTxn Verb are indicators)–Client (high ProcessDirs and File I/O are indicators)–Network (differences in network and aggregate time, watch

network values in server instrumentation results)• Collect resource usage using OS tools to find high use component

–Service Management Framework (SMF), Resource Measurement Facility (RMF), vmstat, iostat, netstat, Performance Monitor

• Change how the component is used or increase the number, size and amount of the component according to what you find

• Example:–Increasing the amount of memory, change processor affinity,

adding a second processor or adapter–Watch out for virus scanners that over utilize the CPU during

backup or restore• Retest and repeat as needed to ensure proper performance




Reporting a Problem

Use the information above to contact Tivoli for help with a problem. It is a good idea to document the steps used in your own problem resolution so when you need to call for support, the information will be readily available.

Using the Problem Determination Guide and the Knowledge Base will often lead to a solution. You are encouraged to check these areas before contacting support.

25

Reporting a ProblemFor support for you can contact IBM Customer Support in one of the following ways:

• Visit the IBM Tivoli Storage Manager technical support Web site.• Submit a problem management record (PMR) electronically at IBMLink.• Submit a problem management record (PMR) electronically at IBM

Software Support.• Customers in the United States can also call 1-800-Tivoli8 (1-800-848-6548)

or 1-800-IBMSERV (1-800-426-7378).• International customers should consult the Web site for customer support

telephone numbers.• Hearing-impaired customers should visit the TDD/TTY Voice Relay

Services and Accessiblity Center Web site.• You can also review the IBM Software Support Guide.

When you contact IBM Software Support, be prepared to provide identification information for your company so that support personnel can readily assist you. Company identification information is needed to register for online support available on the Web site.




Providing Information to Support

The graphic above lists the information to have available when you call Tivoli support. This will help to speed up the problem-solving process.

26

Information for SupportHave the following information ready when you report a

problem:• The Tivoli Storage Manager server version, release, modification, and

service level number. You can get this information by entering the QUERY STATUS command at the Tivoli Storage Manager command line.

• The Tivoli Storage Manager client version, release, modification, and service level number. You can get this information by entering dsmc at the command line.

• The communication protocol (for example, TCP/IP), version, and release number that you are using.

• The activity that you were doing when the problem occurred, listing the steps that you followed before the problem occurred.

• A description of the symptom or error encountered.• The exact text of any error messages relating to the symptom or error

encountered.• Be prepared to provide any error logs or other related documentation

for the problem.• Use the QUERY ACTLOG server command and collect the server

activity log message starting 30 minutes prior to the problem until 30 minutes after the problem.




Student Exercise

Open your Student Exercises book and perform the lab exercises for this unit.

27

Student Exercise


• • • ••



Review Questions

1. True or False? Querying the node on the server gives information on the last session.

2. True or False? Using server tracing has minimal impact on server performance.

3. True or False? Running an FTP of the same data that you are backing up is a good way to check the network transfer rate.

4. True or False? The trace file can be split automatically to prevent having files that are too big to send.

5. True or False? Only one client trace flag can be turned on at the same time.

6. True or False? Aggregate trace flags are a way of specifying multiple related trace flags.


• • •••





• • • ••



Review Answers

1. True or False? Querying the node on the server gives information on the last session.

True.

2. True or False? Using server tracing has minimal impact on server performance.

False.

3. True or False? Running an FTP of the same data that you are backing up is a good way to check the network transfer rate.

True.

4. True or False? The trace file can be split automatically to prevent having files that are too big to send.

True.

5. True or False? Only one client trace flag can be turned on at the same time.

False.

6. True or False? Aggregate trace flags are a way of specifying multiple related trace flags.

True.


• • •••



Summary

28


• Troubleshoot Tivoli Storage Manager.• Implement the server trace function for problem determination.

• Implement the client trace function for problem determination.

• Assist problem determination for a SAN by implementing cyclic redundancy checking (CRC) in Tivoli Storage Manager.

• Explain instrumentation data and its application to problem determination.





Unit 9: Backup and Restore Methodology







Introduction

This unit is a review of some Tivoli Storage Manager client functions, and an introduction to client features and functions that can be used in addition to normal operations of the backup-archive client.

Objectives

2


• Design and create backup-archive client solutions.• Configure and use multi-session backup and restore.• Set up and use backup sets.• Configure and use subfile backup and restore.• Configure and use online image backup and restore.• Enable and use open file support.• Configure and use file grouping.• Set up and use the move node data for faster restores.• Configure and use firewall support.• Configure and use space management on AIX (or HSM).


Unit 9: Backup and Restore MethodologyBackup and Restore Requirements


Lesson 1: Backup and Restore Requirements

Before Tivoli Storage Manager is used as a data protection tool or space management tool, the environment must be analyzed. It is not as simple as installing the backup-archive client on a server and then all data protection challenges are solved. Research, analysis, planning, and implementation should always be the roadmap to building a good solution.

The reason this approach is taken is that the panacea solution does not exist. There is no one single solution to fit all, but there is always a way for Tivoli Storage Manager to be set up and managed to fit an environment’s SLA requirements.

The challenge could be a huge amount of data, a low bandwidth network connection, billions of files, or a limited window in which the data must be backed up or space managed. For disaster recovery testing and machine or media failures, the primary focus should be on the restore. Time is allocated each day for backup processing, but when data needs to be restored, it generally needs to be done immediately.

3

Backup and Restore Requirements




Client Options

One of the benefits of Tivoli Storage Manager is the enormous flexibility available to the administrator when configuring clients.

However, that flexibility can be a source for misunderstanding and improper configuration issues. Although this slide does not include all the options, you can see in the slide there are ten categories listed that are available to optimize the client.

This unit will focus on those shown above.

Some of the client options are only changeable in the dsm.opt (or dsm.sys) file, using a text editor. The include and the exclude statements are good examples. You can do some configuration, but not all, in the GUI.

There are many ways to improve backup and restore performance. For instance, the include and the exclude statements can be very powerful.

Once server data has been analyzed, you should design policy rules, including management classes, to tailor the management of the data to the best solution possible. There is always a trade-off when you do this. When the service level agreement requires that the backup covers everything and keeps many versions, the implication is there will be a lot of data in the storage pools. This implies that the workload on the database will be heavy when there is a request to restore even single files.

4

Client Options

•authorization•error processing•commands processing• format and language•scheduling• restore-retrieve processing•Web client• transaction processing•diagnostic•backup-archive processing


• • •••



Data Movement

This slide gives a basic overview of the data paths that are available to back up or restore data. The data resides on the disk, which is attached to the server. That means that whatever action is done, one of the limits will be the read or write speed from the locally attached disks. This also implies that the organization of the data could have a significant influence on the performance.

For example, when you want to use multiple sessions to read the data from the client but there is only one disk, there will probably not be much improvement on the overall throughput. Do you have to send the data over the LAN, or is it possible to read and write over the SAN? Is the focus on restore? Should you create the possibility for a local restore, or perhaps a local restore over the SAN?

5

Data Movement


• • • ••

Unit 9: Backup and Restore MethodologyLAN-free Data Movement


Lesson 2: LAN-free Data Movement

6

LAN-free Data Movement

In this lesson, we will cover:• LAN-free data movement.• Installation of storage agent:

–In combination with shared tape library.–In combination with shared disk pool.




LAN-free Data Functionality

The LAN-free data transfer function effectively exploits SAN environments by moving back-end office and IT data transfers from the communications network to a data network or SAN. LAN communications bandwidth then can be used to enhance and improve service levels for the environment.

LAN-free data transfer is an extension of the LAN-free data transfer capabilities available with Tivoli Storage Manager. SANs provide an alternative path for data movement between the Tivoli Storage Manager client and server. LAN-free data transfer exploits this SAN path by enabling the Tivoli Storage Manager client to back up and restore data directly to and from SAN-attached storage which is shared between the Tivoli Storage Manager server and client and managed by the server.

The existing LAN connection is used to exchange control information such as policy information and metadata about the objects being backed up, but the data movement utilizes the SAN to write directly to the storage media. Data movement is off-loaded from the LAN and from the Tivoli Storage Manager server processor providing greater scalability for the server. The server can support a greater number of simultaneous client connections.

The LAN-free client will communicate with the Tivoli Storage Manager server to obtain and store database information, and coordinate device and volume access of the storage managed by the server. The server will determine if the client is making a request to access storage for which the client has a SAN path defined. If there is a SAN path defined, then the data will be transferred on that path. If there is a failure when using the SAN path, the

7

LAN-free Data Functionality




LAN-free client will recover from these errors by failing over to a LAN connection to the Tivoli Storage Manager server and proceed to move data over the LAN network.

The LAN-free data transfer will use the SAN tape library sharing and a Tivoli Storage Manager server defined as the library manager. This library manager will manage the storage hierarchy and perform the server functions including storage pool migrations, reclamation, collocations, and backup-restore of the storage pools and the database.

Setup of the Storage Agent

Note: The storage agent must be at the same level as the Tivoli Storage Manager server.

Following is an example command:

dsmsta setstorageserver myname=myservername mypassword=myserverpw myhladdress=myserverhladdress servername=storageservername serverpassword=storageserverpassword hladdress=tcpaddress lladdress=tcpport

This command must be issued after the storage agent has been installed to set up the server-to-server communication between the storage agent and the Tivoli Storage Manager server.

8

Setup of the Storage Agent


• • •••



Up to this moment the storage agent is still separate. The activation for this storage agent must be done on the client by enabling LANFREE in the dsm.opt file.

The backup-archive client uses named pipes, TCP/IP, or shared memory to communicate from the client code to the storage agent code.

If a storage agent is defined to a Tivoli Storage Manager server that acts as a library client, the storage agent must also be able to contact the library manager directly when making mount requests. The Tivoli Storage Manager server that is the library client does not proxy the storage agent's request to the library manager for a mount point.

The storage agent also needs to be defined to both the library client and the library manager. If the storage agent is only defined to the library client, it will use the information from the library client to cross-define itself to the library manager.

The backup-archive client in a LAN-free environment contacts the storage agent and requests the communication information (TCP/IP) for the server. After this initial request is made, the backup-archive client contacts the Tivoli Storage Manager server directly. This establishes the two connections needed to perform operations; one to the storage agent, and one to the Tivoli Storage Manager server.

In this case, since the storage agent is performing LAN-free services for the library client, the backup-archive client will only need to contact the Tivoli Storage Manager server that is the library client, and will not contact the library manager directly. The storage agent, however, will need routes to both the library client server and the library manager server.

You can check what IP address is being used by the server definitions in the devconfig files.

dsmsta setstorageserver myname=myservername mypassword=myserverpw myhladdress=myserverhladdress servername=storageservername serverpassword=storageserverpassword hladdress=tcpaddress lladdress=tcpport


• • • ••



Components of the Shared Disk Pool Setup

SANergy is a product which allows file systems to be shared across a storage area network. SANergy provides a function, known as a metadata controller (MDC), which owns the file system. Any SANergy client may gain access to that storage across the SAN by first requesting the MDC for permission and the location of the data. After the MDC completes the request, the client accesses the volumes by normal NFS or CIFS methods. However, the data blocks are read and written using the SAN, rather than the LAN communication path.

SANergy can be exploited by Tivoli Storage Manager to provide LAN-free backups to storage pool volumes defined as files on a SANergy managed disk. This works in a similar way to the base product when it provides LAN-free backups to tape volumes, in that metadata about the client data is sent across the LAN to the Tivoli Storage Manager server but the data itself travels across the SAN.

The slide shows the flow of data and metadata when using SANergy to share a volume for LAN-free backups. Using this function not only gives the advantages of LAN-free backup, reducing the load on the LAN, but also helps reduce the limitation of the number of available tape drives on LAN-free capability for backup. Data could also be restored using LAN-free either directly from tape or by first using commands within the Tivoli Storage Manager server to move the data back to SANergy managed disk.

9

Components of the Shared Disk Pool Setup

SANergyclient

TSM client SANergyMDC

storage agent

SAN managed node

SANTSM server

shared storage


• • •••



The Shared File Library

Step-by-step Setup of the Shared Disk Library

Start with the Tivoli Storage Manager server to define a device class TYPE=FILE and the option SHARED. Tivoli Storage Manager has a built-in utility which will automatically generate the following items from this command:

• A tape library will be defined with the name of the device class.

• There will be as many tape drives defined as you have specified in the MOUNTLIMIT parameter.

• There is a device class that is defined. This device class is needed because that is a mandatory part of the storage pools.

10

The Shared File Library

A FILE device class is defined SHARED=YESThe TSM server has access to the disks

A storage pool is created with the device class FILE pointing to the shared disk

TSM server

SANergyMDC

client


• • • ••



Total Overview of the Shared Pool Setup Commands

Tivoli Storage Manager Administrator Commands

The following command:

DEFine DEVClass SAN DEVType=FILE MOUNTLimit=2 MAXCAPacity=200m DIRectory=D:\SA1\Storage SHARED=Yes

creates the following actions:

– Library SAN defined.

– Drive SAN1 defined in library SAN.

– Drive SAN2 defined in library SAN.

– Device class SAN defined.

11

Overview of the Shared Pool Setup Commands

1. DEFine DEVC .. DEVType=FILE .. SHARED=Yes

2. DEFine STGpool SANPOOL

3. DEFine PATH

4. OS administrator allows resource sharing

5. Windows client maps shared resource: NET USE


• • •••



The following command:

DEFine STGpool SANPOOL SAN ACCess=READWrite NEXTstgpool=BACKUP_TAPE HIGHmig=70 LOWig=30 COLLocate=NO MAXSCRatch=10 REUSEdelay=7 DATAFormat=NATIVE

creates the following actions:

– Storage pool name: SANPOOL

– Storage pool type: Primary

– Device class name: SAN

– Estimated capacity (MB): 0.0

– Next storage pool: BACKUP_TAPEDEFINE path SA1 SAN1 SRCType=server DESTType=drive DEVIce=FILE LIBRary=san DIRectory=D:\SA1Storage ONLine=Yes

Administrator of TSM1 system allows resource sharing: D:\SA1\Storage as Storage.

The Windows client maps the shared resource:

net use V: \\TSM1\Storage /persistent:yes


• • • ••



Overview of the Customized Files

Windows Step 1

The dsmserv.opt file of the Tivoli Storage Manager server:

COMMMETHOD TCPIPTCPPORT 1500 # port used by clients / cmd admins tcpport# and server/stagents lladdress parameter

TCPWINDOWSIZE 63TCPNODELAY NO

COMMMETHOD HTTPHTTPPORT 1580 # port used by web-administrator sessions only

* allow communication via namedpipe from local client (WIN)COMMMETHOD NAMEDPIPENAMEDPIPENAME \\.\PIPE\TSM_server_name# using \\.\PIPE\LANFREE was successful

* allow communication via sharedmem from local client (UNIX)COMMMETHOD SHAREDMEMSHMPORT 1510 # default server_sharedmem_port

12

Overview of the Customized Files

client

storage agent

server

defineregister


• • •••



VOLUMEHistory VOLHIST.TXTDEVCONFIG DEVCONFIG.TXT

Windows Step 2

Tivoli Storage Manager definitions:

/* Verify that your server has defined itself */Set SERVername central_server_nameSet SERVERPAssword central_server_passwordSet SERVERHladdress central_server_tcpip_addrSet SERVERLladdress central_server_tcpip_portSet CROSSDefine ON

/* Install storageagent and setup server-to-server connection */DEFine SERver stagent_name SERVERPAssword=stagent_passwordCOMMmethod=TCPIP HLAddress=stagent_ipaddress LLAddress=stagent_portDESCription='stagents description'

/* Define the virtual SAN file devices */DEFine DEVClass san_devc_name DEVType=FILE SHARED=YESMAXCAPacity=400M MOUNTLimit=4 DIRectory=server_local_dir_name

/* Define a storagepool to use the SAN file deviceclass */DEFine STGpool san_stgpool_name POOLType=PRIMARY DEVClass=san_devc_name

/* Define an own policy structure to use SAN file devices */DEFine DOMain san_domain_name DESCription='SAN Domain'DEFine POlicyset san_domain_name san_policyset_nameDESCription='SAN Policyset'DEFine MGmtclass san_domain_name san_policyset_name san_mgmtcl_nameDESCription='SAN Default Mgmtclass'DEFine COpygroup san_domain_name san_policyset_name san_mgmtcl_nameSTANDARD Type=Backup DESTination=san_stgpool_nameDEFine COpygroup san_domain_name san_policyset_name san_mgmtcl_nameSTANDARD Type=Archive DESTination=san_stgpool_nameASsign DEFMGclass san_domain_name san_policyset_name san_mgmtcl_nameACTivate POlicyset san_domain_name san_policyset_name

/* Register the client and it connect to the SAN domain */REGister Node client_name pw san_domain_name BACKDELete=YesKEEPMP=Yes MAXNUMMP=4


• • • ••



/* Define the path to storageagents drive*/DEFine PATH stagent_name san_library_name san_drive_name1SRCType=server DESTType=drive DEVIce=FILE ONLine=YesLIBRary=san_library_name DIRectory=storageagents_dir_name

Windows Step 3

The devconfig.txt file of the Tivoli Storage Manager storage agent is created as follows:

dsmsta setstorageserver myname=myservernamemypassword=myserverpwmyhladdress=myserverhladdressservername=storageservernameserverpassword=storageserverpasswordhladdress=tcpaddress lladdress=tcpport

SET STANAME stagent_nameSET STAPASSWORD stagent_passwordSET STAHLADDRESS stagent_tcpip_addrSET STALLADDRESS stagent_tcpip_portDEFINE SERVER servername HLADDRESS=server_tcpip_addrLLADDRESS=server_tcpip_port SERVERPASSWORD=server_password

Windows Step 4

The dsmsta.opt file of the Tivoli Storage Manager storage agent:

SERVERNAME stagent_nameDEVCONFIG DEVCONFIG.TXT

* allow connections through TCPIPCOMMMETHOD TCPIPSTAMAXPOOLEDSESSIONS number of sessions # 0, no sessions, default 25RESOURCETIMEOUT minutes # default 10 min.

* allow connections through named pipes (WIN)COMMMETHOD NAMEDPIPENAMEDPIPENAME \\.\PIPE\pipename # must correlate to dsm.opt

* allow connections through sharedmem (UNIX)COMMMETHOD SHAREDMEMSHMPORT stagent_sharedmem_port # must correlate to dsm.sys


• • •••



Windows Step 5

Make the share in the explorer GUI (NSF mount for UNIX).

Windows Step 6

Issue the following command:

net use x: \\server_name\resource_name /persistent:yes

Example:

net use x: \\TSM1\STORAGE /persistant:yes

AIX Steps 1 and 2

Repeat steps 1 and 2 from the Windows section. They are done on the Tivoli Storage Manager server, so they are the same for both platforms.

AIX Step 3

• dsm.opt

• SErvername Windows

• DOMAIN ALL-LOCAL

• dsm.sys

SErvername WindowsNODENAME client_namePASSWORDACCESS GENERATE

* establish LAN session to serverCOMMMETHOD TCPIPTCPSERVERADDRESS server_tcpip_addrTCPPORT 1500 # server_tcpip_port - must correspond to dsmserv.optTCPWINDOWSIZE 63TCPNODELAY YES


• • • ••



* establish TCP/IP session to storage agentENABLELANFREE YESLANFREECOMMMETHOD TCPIPLANFREETCPPORT 1600 # stagent_tcpip_port - must correspond to dsmsta.opt

* or

* establish Unix shared memory internal session to storage agentENABLELANFREE YESLANFREECOMMMETHOD SHAREDMEMLANFREESHMPORT 1610 # stagent_sharedmem_port - must correspond to dsmsta.opt

AIX Step 4

The devconfig.txt file of Tivoli Storage Manager storage agent is created as follows:

dsmsta setstorageserver myname=myservername mypassword=myserverpwmyhladdress=myserverhladdressservername=storageservernameserverpassword=storageserverpasswordhladdress=tcpaddress lladdress=tcpport

SET STANAME KOPFSASET STAPASSWORD 1877581dSET STAHLADDRESS 10.12.154.29DEFINE SERVER HG2204X HLADDRESS=10.12.154.30 LLADDRESS=1500 SERVERPA=183299eb

AIX Step 5

* ========================================================* Tivoli Tivoli Storage Manager * Storage Agent* Version 5, Release 2* 5698-ITSM (C) Copyright IBM Corporation, 2001, All Rights * Reserved.* ========================================================* Storage Agent Options File (dsmsta.opt.smp)* Platform: AIXTCPPORT 1600SHMPort 1610DEVCONFIG devconfig.txt


• • •••



AIX Step 6

mount windows:/M/

AIX Step 7

# SANergyconfig.txt - input file for SANergyconfig# rules - no tabs and no leading spaces and # are comments## The owner command should be first in the list.## nfs volumes to accelerated by SANergy# Example:# fuse|/ddd where /ddd is the nfs mount point# fuse|/eee##ownerfuse|/M

AIX Step 8

# SANergy sh and ksh variable setup# The following line(s) can be added to your .profile file# or can be cut and pasted to the command line.## /usr/SANergy/SANergyshsetupaix#LIBPATH=/usr/SANergyexport LIBPATHSANPATH=/usr/SANergyexport SANPATHSANTISDIR=/usr/SANergy/nls/codesetexport SANTISDIR

AIX Step 9

#!/bin/sh## $Revision: 1.1 $#


• • • ••



#### modified:## "mkdev -l aio0" inserted to put the aio device into Available state.## SANergy will not work without this.#### correct way to get it working:## 1. mount NFS volumes## 2. /etc/rc.sanergy## 3. for each SANergy-accelerated application do (in ksh syntax)## . /usr/SANergy/SANergyshsetup#### BEWARE:## nothing (not even ps, ls, ...) will work, if the variables from## /usr/SANergy/SANergyshsetup are set and sanergyd is not working.## do not set these variables in /etc/environment or any other## global config file.##

killproc() { # kill the named process(es) pid=`/usr/bin/ps -e |/usr/bin/grep -w $1 |/usr/bin/sed -e 's/^ *//' -e 's/ .*//'`[ "$pid" != "" ] && kill $pid}

umask 002

ECHO=echoMKDEV=mkdevKEYFILE=/usr/SANergy/SANergykey.txtCFFILE=/usr/SANergy/SANergyconfig.txtCONFIG=/usr/SANergy/SANergyconfigFIXUP=/usr/SANergy/preparevols

case "$1" in'stop')

killproc sanergyd;;*)$ECHO "enabling asynchronous I/O"$MKDEV -l aio0

$ECHO "sanergy - start"

#$FIXUP

#$ECHO "sanergy - setting env vars"LIBPATH=/usr/SANergyexport LIBPATH


• • •••



$ECHO "sanergy - running SANergyconfig"if test -r $KEYFILE; then$ECHO "sanergy - setting key"#$CONFIG -d < $KEYFILE >> /dev/null 2>&1&$CONFIG -d < $KEYFILEif test -r $CFFILE; then$ECHO "sanergy - set fused"#$CONFIG -d < $CFFILE >> /dev/null 2>&1&$CONFIG -d < $CFFILE$ECHO "sanergy - complete"fifi

killproc sanergyd/usr/SANergy/sanergyd

;;esac

Now the AIX storage agent is ready to use the shared disk pool.


• • • ••

Unit 9: Backup and Restore MethodologyMulti-session Backup and Restore


Lesson 3: Multi-session Backup and Restore

Resource Utilization

Use the RESOURCEUTILIZATION option in your client options file (dsm.opt) to regulate the level of resources the Tivoli Storage Manager server and client can use during processing.

Regulating Backup and Archive Sessions

When you request a backup or an archive, the client can use more than one session to communicate with the server. The default is to use a maximum of two sessions: one to query the server and one to send file data. The client can use only one server session if you specify a RESOURCEUTILIZATION setting of 1.

A client can use more than the default number of sessions when connecting to a server that is version 3.7 or higher. For example, RESOURCEUTILIZATION=10 permits up to ten

13

Multi-session Backup and Restore

REGISTER NODE:MAXNUMPKEEPMPQ Status:Maximum sessionsDeviceclassMOUNTLIMIT

session 1

session 2

session 3

dsm.optresourceutilization


• • •••



sessions with the server. Multiple sessions may be used for querying the server and sending file data.

Multiple query sessions are used when you specify multiple file specifications with a backup or archive command. For example, if you enter:

include filespaceA filespaceB

and you specify RESOURCEUTILIZATION=5, the client may start a second session to query files on file space B. Whether or not the second session starts depends on how long it takes to query the server about files backed up on file space A. The client may also try to read data from the file system and send it to the server on multiple sessions.

Regulating Restore Sessions

When you request a restore, the default is to use a maximum of one session, based on how many tapes the requested data is stored on, how many tape drives are available, and the maximum number of mount points allowed for the node.

Multi-session Restore

If all of the files are on disk, only one session is used. There is no multi-session for a pure disk storage pool restore. However, if you are performing a restore in which the files reside on four tapes and on disk, you could use up to five sessions during the restore.

The Tivoli Storage Manager server can set the maximum number of mount points a node can use on the server using the MAXNUMMP parameter. If the RESOURCEUTILIZATION option value exceeds the value of the MAXNUMMP on the server for a node, the backup can fail with an unknown system error message.

For example, if the data you want to restore is on five different tape volumes, the maximum number of mount points is five for your node, and RESOURCEUTILIZATION is set to three, then three sessions will be used for the restore. If you increase the RESOURCEUTILIZATION setting to five, then five sessions will be used for the restore. Multiple restore sessions are only allowed for no query restore operations.


• • • ••



Multi-session Backup and Restore Considerations

The following factors can affect the throughput of multiple sessions:

• The server's ability to handle multiple client sessions. Is there sufficient memory, storage volumes, and CPU cycles to increase backup throughput?

• The client's ability to drive multiple sessions (sufficient CPU, memory, and so forth).

• The configuration of the client storage subsystem. File systems that are striped across multiple disks, using either software striping or RAID-5 can better handle an increase in random read requests than a single drive file system. Additionally, a single drive file system may not see performance improvement if it attempts to handle many random concurrent read requests.

• Sufficient bandwidth in the network to support the increased traffic.

14

Multi-session Backup and Restore Considerations

Data organizationin storage pool

session 1

session 2

session 3

No queryrestore protocol


• • •••



Other things to consider when running multiple sessions include:

• The client could produce multiple accounting records.

• The server may not be able to start enough concurrent sessions. To avoid this, the server MAXSESSIONS parameter must be reviewed and possibly changed.

• A query node command may not summarize all client activity for a session.


• • • ••



Student Exercise


15

Student Exercise


• • •••

Unit 9: Backup and Restore MethodologyBackup Sets


Lesson 4: Backup Sets

Backup sets provide an instant archive and rapid recovery capability.

Instant Archive

This capability allows an administrator to create an archive collection from backup versions already stored on the server.

Rapid Recovery

When an administrator is away from his office without a network connection and loses data, it can be restored from the backup set.

16

Backup Sets

LAN

SANclientserver

generate backupset local input


• • • ••



Creating and Using Client Backup Sets

A backup set is a collection of backed up data from one client, stored and managed as a single object, on specific media, in server storage. WebSphere Application Server (WAS) and group backups can also be added to a backup set. The server creates copies of active versions of a client's backed up objects that are within the one or more file spaces specified with the generate backupset command, and consolidates them on sequential media. Currently, the backup object types supported for backup sets include directories and files only.

The media should be directly readable by a device such as a CD-ROM, JAZZ, or ZIP drive attached to a client's machine.

Administrators can generate multiple copies of backup sets that correspond to some point in time. The backup sets can be retained for various time periods. This is an efficient way to create long-term storage of periodic backups, without requiring the data to be sent over the network again.

While an administrator can generate a backup set from any client's backed up files, backup sets can only be used by a backup-archive client. You cannot generate a backup set for a network attached storage (NAS) node.

Generating Client Backup Sets on the Server

The administrator has to have system or restricted policy over the domain to which the node is assigned. You can generate backup sets on the server for client nodes. The client node for which a backup set is generated must be registered to the server. An incremental backup must be completed for a client node before the server can generate a backup set for the client node.

Choosing Media for Generating the Backup Set

To generate a backup set, you must specify a device class that is associated with the media to which the backup set will be written.When you select a device class for writing the backup set, the backup set must be written on any sequential access device whose device types are supported on both the client and server machines. If you do not have access to compatible devices, you will need to define a device class for a device type that is supported on both the client and server.


• • •••



You can write backup sets to sequential media: sequential tape and device class FILE. For device class FILE, the server creates each backup set with a file extension of host. You can copy FILE device class volumes to removable media that is associated with CD-ROM, JAZZ, or ZIP devices, by using the REMOVABLEFILE device type.

The tape volumes containing the backup set are not associated with storage pools, and therefore, are not migrated through the storage pool hierarchy.

Using Scratch Media

You can determine whether to use scratch volumes when you generate a backup set. If you do not use specific volumes, the server uses scratch volumes for the backup set. You can use specific volumes for the backup set. If there is not enough space to store the backup set on the volumes, the server uses scratch volumes to store the remainder of the backup set.


• • • ••



Managing the Backup Set

The backup set is managed by the administrator in the Tivoli Storage Manager database.

All backup sets have an entry in the volume history file. As long as the entry is in the volume history file, the volume cannot be reused by the Tivoli Storage Manager server.

On the command to generate the backup set, there is a parameter which defines how long the backup set volume will be registered in the volume history file. The default value is retention=365. The value is in days. There is also the option to define the retention parameter as nolimit.

It is possible to move a backup set from one Tivoli Storage Manager server to another. That can be done by using the define command. The same rules you have between the Tivoli Storage Manager server and the backup-archive client apply here. The media must be the same.

The update is specially designed for the possibility to update the location and the retention parameter.

The query backupsetcontents command displays information about the files and directories contained in a client node's backup set. Be aware that processing this command can use considerable network resources and mount points.

17

Managing the Backup Set

define backupsetupdate backupsetquery backupsetdelete backupsetq backupsetcontents

TSM databasebackupsetname.12214

volumehistory

generate backupset

server


• • •••



Restoring the Backup Set

Localbackupset

The LOCALBACKUPSET option specifies whether the Tivoli Storage Manager GUI bypasses initial logon with the Tivoli Storage Manager server to restore a local backup set on a stand-alone workstation. You can use this option on the command line or place it in your client options file (dsm.opt).

If you set the LOCALBACKUPSET option to YES, the GUI does not attempt initial logon with the server. In this case, the GUI only enables the restore functionality.

If you set the LOCALBACKUPSET option to NO (the default), the GUI attempts initial logon with the server and enables all GUI functions.

To start the GUI and bypass the initial logon with the server to restore a local backup set on a stand-alone workstation, enter:

dsm -localbackupset=yes

18

Restoring the Backup Set

commandlineoptions:localbackupset ifnewerdirsonly locationfilesonly preservepath

quietreplace dsm.optsubdir

TSM>Restore BackupsetLAN

SAN

client

local input


• • • ••



Restore Backupset

The restore backupset command restores a backup set from the server, a local file, or a local tape device.

If you are restoring a file space from a backup set to a system that did not perform the original backup, you may need to specify a destination because the directory information probably will not be the same as from the node that generated the backupset.

The syntax should be, for example:

dsmc restore backupset backupsetname {\\machinename\c$}\* c:\destdir -subdir=yes

You can restore one or all system objects from a backup set using the SYSTEMOBJECTNAME parameter. On the GUI, you can only restore the entire backup set, which may contain system objects.

For Windows Server 2003, you cannot restore individual system state components. You can use backup sets to restore the entire system state.

You can restore a group from a backup set with the following considerations:

• You must set the SUBDIR option to YES.

• The SOURCEFILESPEC must be the virtual file space name enclosed in brackets ({}), followed with a terminating directory delimiter.

The entire group, or groups, in the virtual file space will be restored. You cannot restore a partial group by specifying a qualified source file space.

A backup set can also be restored from a tape device on all Windows clients.

If you use the restore backupset command on the initial command line, and you set the location option to tape or file, no attempt is made to contact the server.

If you are unable to restore a backup set from portable media, check with your Tivoli Storage Manager administrator to ensure that the portable media was created on a device using a compatible format.

Be careful when you change node names. If the object you want to restore is part of a backup set generated on a node and the node name is changed on the server, any backup set objects that were generated prior to the name change will not match the new node name. Ensure that the node name is the same as the node for which the backup set was generated.


• • •••



Restoring Backup Sets In a SAN Environment

You can restore backup sets in a storage area network (SAN) in the following ways:

• If the backup set is on a SAN-attached storage device, specify the device using the FILENAME parameter and use the LOCATION=TAPE option. Tivoli Storage Manager restores the backup set directly from the SAN-attached storage device, gaining high-speed restore performance.

You as administrator or user must ensure that the correct tape is mounted in the SAN-attached tape drive prior to issuing the restore command. The backup-archive client will not initiate a SCSI autochanger to mount the tape automatically.

• If the backup set is not on local media or a SAN-attached storage device, you can specify the backup set using the BACKUPSETNAME parameter. Use the LOCATION=SERVER option to restore the backup set directly from the server using the LAN.


• • • ••



Student Exercise


19

Student Exercise


• • •••

Unit 9: Backup and Restore MethodologyAdaptive Subfile Backup


Lesson 5: Adaptive Subfile Backup

Adaptive subfile backup is a backup method that backs up only the parts of a file that have changed since the last backup. The server stores the base file (the complete initial backup of the file) and subsequent subfile (the changed parts) that depend on the base file.

The process works with either the standard progressive incremental backup or with selective backup.

Applicable to clients on Windows systems, the process maintains backups of data while minimizing connect time and data transmission for mobile and remote users. The base file plus a maximum of one subfile is restored to the client.

Adaptive subfile backup has the following features:

• Backup of changed file data on byte or block level

• Differential backup of changed file data

• Dynamic selection of backup level based on initial file size

• Minimizes network traffic; well suited for backup of mobile systems

• Fully integrated into Tivoli Storage Manager

20

Adaptive Subfile Backup

Differential backup of changed file data Dynamic selection of backup level based on initial file size

LAN

TSMserver

subfile1

subfile2

subfile3

basefile mobile client

Backup of changed file data on byte or block level


• • • ••



• Transparent to the user

• Policy management applies as before

• Fault-tolerant backup and restore

• Multithreading-aware

The Subfile Architectural Components

Setting Subfile Backup for Client Nodes

Use the set subfile command to set up the server to allow clients to back up subfile. On the client's machine, the SUBFILEBACKUP, SUBFILECACHEPATH, and SUBFILECACHESIZE options must be specified in the client option file (dsm.opt).

With subfile backups, when a client's file has been previously backed up, any subsequent backups are typically made to the portion (a subfile) of the client's file that has changed, rather than the entire file.

21

The Subfile Architectural Components


• • •••



Use the query status command to determine whether subfiles can be backed up to the server running this command.

Performing a Backup with Limited Bandwidth

If you plan to perform a backup over a network device that has limited bandwidth, such as a modem, you can help reduce network traffic by using adaptive subfile backup. Adaptive subfile backup requires a version 4 or later server. Reducing traffic can also increase the speed of your backup. An adaptive subfile backup sends only changed portions of a file to the server during successive backup operations.

Perform the following steps to use adaptive subfile backup:

1. Ensure that the server allows this type of backup. Check with your system administrator.

2. Set the SUBFILEBACKUP, SUBFILECACHEPATH and SUBFILECACHESIZE options in your client options file (dsm.opt). For more information on adaptive subfile backup options, see:

– Subfilebackup

– Subfilecachepath

– Subfilecachesize

3. Add an EXCLUDE.DIR option to your client option file, which excludes the directory containing the subfile cache information.

4. When you are ready to back up changes, adaptive subfile backup will occur as part of incremental and selective backups.

You can restore an adaptive subfile backup using the restore command. The server will restore the base file along with the updates from the changed (delta) subfile, so that what you receive is the latest backed up version of your file.


• • • ••



Adaptive Subfile Backup and Restore

When a user tries to back up a file with adaptive subfile backup enabled, two different kinds of backups may result. The user has no control over this. The backup-archive client itself decides, following built-in rules, whether to back up the file as a base file component or as a delta file component.

The first step is always to take a backup of the entire file. This is a base file backup. Subsequent backups use delta file backups, which take a copy only of what has changed relative to the reference file. Delta file backup may occur up to 20 times before another base file backup occurs. However, if the changes to the file are too numerous, and at the last delta file backup a compression of at least 40 percent could not be achieved, then a new base file backup is initiated.

The base file and delta file represent different versions of the same file. To restore a file that has been backed up using adaptive subfile technology, Tivoli Storage Manager requires only the base file and the delta file from the last day and recreates the latest version of the file.

22

Adaptive Subfile Backup and Restore


• • •••



Base File Backup Process

The first step of the subfile backup process is a base file backup. As soon as subfile backup is enabled and an eligible file is backed up with the incremental or the selective backup function, the backup-archive client starts the following process:

1. The file is copied to the client reference cache to create a reference file there. Depending on the size of the file, this is either a byte by byte copy or a sequence of block signatures. An entry is added in the cache control database that hashes the real file name to the unique name of the reference file and contains additional control information.

2. The digital signature of the file is taken and also stored in the client reference cache.

3. The original file is backed up to server storage along with the newly created digital signature, which is stored in the server database.

23

Base File Backup Process


• • • ••



Delta File Backup Process

When the base file changes after being backed up once, and the user initiates another incremental or selective backup on that file, the backup-archive client performs the following process:

1. The client checks for the existence of a reference file in the client reference cache. Then, the client ensures that all the other conditions for a subfile backup are fulfilled.

2. The client creates the delta file in the client cache, using its patented differencing algorithm. Depending on whether byte-level or block-level differencing is used, the delta file contains information about the changed bytes or changed blocks in the current version of the file relative to the reference file (or base file). The differencing is not performed on non-file data such as access control list (ACL) data. Thus, the delta file also contains all the non-file data (ACLs) of the current version of the file.

3. The delta file and the digital signature stored in the client cache are sent to the server. The server compares this digital signature with the stored digital signature from the base file backup, and if they match, the server accepts the delta file. If the signatures do not match, a transaction failure occurs and a new base file backup is triggered.

4. The delta file is removed from the client reference cache.

24

Delta File Backup Process


• • •••



Adaptive Subfile Restore Process

The server versions of the file are represented either through the base file or the delta file component.

To restore a file version, the user initiates a restore of one of those components. Behind the scenes, however, the client and server are performing the following steps:

1. The corresponding base and delta file components of the file version are restored from the server to a temporary reconstruction directory under \~TSMtemp on the logical drive to which the file is being restored.

The file components are restored under a temporary name to allow their reconstruction before moving the version file to the original location, and also to keep track of the restore status and the relationship between the components.

2. Delta file data is copied to a temporary rebuild file. This enables reconstruction of the file through overwriting of the restored delta file data so it can utilize the ACL data of that delta file version.

3. The base file and the delta file component stored by the temporary reconstruction file are reconstructed into the version file by overwriting the data part of the restored delta file. The ACLs of this component (representing the ACL of the restored version of the file) remain intact.

25

Adaptive Subfile Restore Process


• • • ••



4. The base file and the temporary reconstruction file are deleted from the reconstruction directory.

5. The version file is renamed to its original name. File components can be restored in any order. The components need not be restored sequentially.


• • •••



Student Exercise


26

Student Exercise


• • • ••

Unit 9: Backup and Restore MethodologyImage Backup


Lesson 6: Image Backup

Logical Volume Backup (Image Backup)

The Tivoli Storage Manager Windows 2000/XP/2003 clients are enhanced to support a logical volume image backup of file systems and raw volumes. The image backup and restore functionality has been available in Tivoli Storage Manager on Windows 2000, XP and 2003 since version 5.2, and has been part of some of the Tivoli Storage Manager UNIX clients since Version 3.7 (AIX, Solaris, HP).

New same-drive image back-up feature in TSM 5.3 for Windows:

You can now perform open file support (OFS) backups or online image backups on machines with a single NTFS-based C: drive. You can also easily verify if the LVSA is functioning properly or find the root cause of any failure by checking the Windows event log, without turning on any trace facilities.

Note: Windows 2003 LVSA for OFS has delayed availability.

27

Image Backup

Client VolumesRAW NTFS FAT32FAT

OFFLINEorONLINE

dsmc backup image

LAN

SANServer Client


• • •••



Overview

The traditional offline image backup prevents access to the volume by other system applications during the operation. If the logical volume snapshot agent (LVSA) is installed and configured, the Tivoli Storage Manager client performs an online image backup, during which the volume is available to other system applications. An image backup provides the following benefits:

• Improves the speed with which Tivoli Storage Manager backs up and restores file systems containing many small files.

• Conserves resources on the server during backups since only one entry is required for the image.

• Provides a point-in-time picture of your file system, which may be useful if your enterprise needs to recall that information.

• Restores a corrupt file system or raw logical volume. Data is restored to the same state it was when the last logical volume backup was performed.

Supported volumes are:

• Formatted (NTFS, FAT32 and FAT)

• Unformatted (RAW)

The Tivoli Storage Manager server does not track individual files in the file system image. File system images are tracked as individual objects and management class policy will be applied to the file system image as a whole. There are three different types of image backup available, as described in the following sections.

Image Backup (Offline)

The volume being backed up is made unavailable to all other applications such as Windows Explorer, CMD shells, database software, and so forth. It is not even available for reading operations on the volume like in its UNIX offline (static) counterpart. The backup-archive client is the only one with access to the volume. Once the image backup is complete, the backup-archive client makes the volume available to system and other applications again.

Image Snapshot Backup (Online)

The volume being backed up remains available to all applications during backup. The backup consists of a point-in-time snapshot of the entire volume. The default used by the backup-archive client is Online unless otherwise specified.


• • • ••



Incremental Image Backup (Date-only)

Incremental image backup backs up only those files whose modification dates and times are newer than the date and time of the last image backup. You can perform incremental image (date-only) of last image backup regardless of whether the full image was backed up offline or online. You can use the native GUI, the command-line interface, and the Web client GUI to create an image of file systems and raw logical volumes on the Tivoli Storage Manager server.

Note: On Windows 2000 NTFS formatted volumes, Tivoli Storage Manager will optionally transfer just the used blocks of the volume.

Setting Up Online Image Backup

Following are some related options and values.

DOMAIN.IMAGE

This is not a new option, but it now applies to Windows 2000 clients. This option allows users to enter a set of volumes, which are backed up as part of the client machine's image domain. It is used when the command is entered without any volume specified (dsmc backup image) and from the GUI using the Backup Domain Image action menu.

28

Setting Up Online Image Backup


• • •••



DOMAIN.IMAGE d: e:INCLUDE.IMAGE -imagetype=SNAPSHOT/STATIC

This is also not a new option, but it now applies to Windows 2000 clients. It is enhanced to allow users to set different image options per volume.

• IMAGETYPE=SNAPSHOT – Specifies that you want to perform an online image backup during which the volume is available to other system applications. This is the default if the logical volume snapshot agent (LVSA) is installed and configured.

• IMAGETYPE=STATIC – Specifies that you want to perform an offline image backup during which the volume is locked and unavailable to other system applications.

When you have specified IMAGETYPE=STATIC, then there is an additional option SNAPSHOTCACHESIZE. This option specifies the maximum size of the OBF file (old blocks file), in which old data blocks are stored during an online image backup. The value is a percentage of the total size of the volume being backed up. The range of values is 1 through 100 percent. The default is 100 percent.

The second option on that specification is SNAPSHOTCACHELOCATION. This option specifies the valid path to the location where the LVSA will place the OBF file (old blocks file) during an online image backup. The default location is C:\TSMLVSA.

The third additional option is SNAPSHOTFSIDLEWAIT. This option is used with the backup image command or the INCLUDE.IMAGE option when performing an online image backup to specify the amount of time that must pass in which there is no write activity on a volume before a snapshot is taken. The range of values is 0 through 999; the default is 1. You may specify ms (milliseconds) or s (seconds) qualifiers with the value.

This last option is the possibility to retry—SNAPSHOTFSIDLERETRIES. Use this option with the backup image command or the include.image option when performing an online image backup to specify the number of additional times the LVSA should try to achieve the snapshot FS idle wait time before the online image backup operation fails. The range of values is 0 through 99; the default is 10.

In case the company wants to quiesce activities, there is the PRESNAPSHOTCMD option. This allows you to quiesce an application before the LVSA starts the snapshot during an online image backup. For the same reason, the POSTSNAPSHOTCMD option allows you to start an application after the LVSA starts a snapshot during an online image backup.

There is also an important option named IMAGEGAPSIZE. This specifies the minimum size of empty regions in a formatted logical volume that should be skipped during an image backup. This option is only valid for NTFS formatted volumes. You may specify k (kilobytes), m (megabytes), or g (gigabytes) qualifiers with the value. If this option is set to the block size on the disk the image backup will be the size of the data and not the size of the disk. This could degrade performance during backup but takes up less room in storage.


• • • ••



Policy for Logical Volume Backups

Consider defining a management class specifically for logical volume backups. To enable clients to restore a logical volume and then reconcile the results of any file backup operations since the logical volume backup was made, you must set up management classes with the backup copy group set up differently from the STANDARD. The versions data exists, versions data deleted, and retain extra versions parameters work together to determine over what time period a client can restore a logical volume image and reconcile later file backups. Also, you may have server storage constraints that require you to control the number of backup versions allowed for logical volumes. Backups of logical volumes are intended to help speed the restoration of a machine. One way to use the capability is to have users periodically (for example, once a month) perform a logical volume backup, and schedule daily full incremental backups. If a user restores a logical volume, the program first restores the logical volume backup and then any files that were changed since the backup (incremental or other file backup processes). The user can also specify that the restore process reconcile any discrepancies that can result when files are deleted.

Note: For logical volume backups, the server ignores the frequency attribute in the backup copy group.


• • •••



Setting Up Third-party Snapshot Backup

© 2003 IBM Corporation

Provide Support for External Snapshot Providers

• Some of the snapshot providers are PSM (Windows), Open File Manager (Windows/NetWare), and AIX 5.2 JFS2. Windows 2000 and XP use the Tivoli Storage Manager LVSA, which is already integrated into the product.

• Snapshots are point-in-time views of a logical file system.

• Each of these snapshot providers can have multiple snapshots (for instance, PSM allows 255) of the same logical file system. For example, there are four snapshots of /home called /snapshot.1, /snapshot.2, /snapshot.3, and /snapshot.4.

29

Setting Up Third-party Snapshot Backup


• • • ••



Support for External Snapshot Providers

Snapshots have their own file spaces on the Tivoli Storage Manager server. However, you must remember that these snapshots are logically all the same view of the same file space. That is why Tivoli Storage Manager has introduced the SNAPSHOTROOT option.

The following is an example:

INCR /home snapshotroot=/snapshot.1

This command will make an incremental of my /home file system; the files selected come from the /snapshot.1 snapshot of the /home file system, but the files are going to be managed on the Tivoli Storage Manager server with the /home file space.

Another example is provided below. A selective backup is performed:

SEL C:\directory\* -snapshotroot=\\machine\snapshot.1

In this case, the request is to back up the contents of the directory on the C: drive; the administrator wants the files to come from the \\machine\snapshot.1 snapshot of the C: drive, but the data owner wants the files to be managed on the Tivoli Storage Manager server with the C: file space.

The snapshot provider needs to provide a virtualized, logical file system, for instance, a mount point on UNIX, a volume mount on Windows, and so forth. The products that were unit tested (PSM for Windows, JFS2 snapshot for AIX 5.2) seem to indicate that all products would behave in this manner.

Considerations

It is not recommended to use the SNAPSHOTROOT option in combination with journal-based backup. Journal-based backups record changes to real volumes, and backups will occur against the snapshot. That could create a problem if records change after the snapshot is taken.

The SNAPSHOTROOT option does not work with any of the federated backup syntax (VIRTUALFS, FILELIST, and so forth).

It also does not work with special file systems, for instance, Windows system objects, NetWare NDS, or server-specific information.

The option can only be used in combination with the command line; not on the GUI, in dsm.opt, dsm.sys, or server client option sets.


• • •••



It is advisable to use this primarily when scheduling through a script. Use the preschedule and postschedule commands to control the snapshot. The script example will be something similar to:

# take snapshot of %1 and put it on %2# dsmc INCR %1 –snapshotroot=%2

Restore Image Commands

Logical Volume Restore

When performing an image restore, the following things are to be considered.

You can restore NTFS to FAT32, FAT to NTFS, RAW to NTFS, NTFS to RAW and so on, but it is suggested that the destination volume be RAW (not formatted) or formatted with the same type of file system as the original one.

Restoring the image of a volume will restore the volume to the same state that it was in when you performed your last image backup. Be absolutely sure that you need to restore an image, for it will replace your entire current file system or raw volume with the image on the server.

30

Restore Image Commands


• • • ••



Ensure that the file system or volume to which you are restoring the image is at least the same size as the image that is being restored.

When restoring an image to a file system, ensure that the file system exists on your machine before you begin the restore. Otherwise, the restore operation will fail.

Image restores are always offline. Ensure that the file system is not in use. The client will lock the file system before starting the restore. If the file system is in use when the client attempts to lock the file system, the restore will fail. The client will unlock the file system after the restore completes.

You cannot restore an image to the same location that the Tivoli Storage Manager client program is installed.

If you created an image of the system drive, you cannot restore the image to the same location because the client cannot have an exclusive lock of the system drive. Also, because of different system component configurations, the system image may not be consistent across components (such as active directory). Some of these components can be configured to use different volumes where parts are installed on the system drive and others to non-system volumes.

If you have run progressive incremental backups and image backups on your file system, you can perform an incremental restore of the file system. This process updates the original image with individual files backed up after the last image backup. Optionally, if files were deleted after the original backup, the incremental restore can delete those files from the base image. Incremental backups and restores can be performed only on mounted file systems, not on raw logical volumes.

To ensure that you can perform point-in-time restores of your file systems, including deleting original files which no longer exist on the logical volume, use a combination of full image backups and the incremental command as described in these steps:

1. Perform a full incremental backup of the logical volume, for example:

dsmc incremental g:

2. Perform an image backup of the same logical volume, for example:

dsmc backup image g:

3. Periodically, perform incremental backups, for example:

dsmc incremental g:

You must follow these steps in the order shown to ensure that the server records additions and deletions accurately. The following command restores the file system to its exact state as of the last incremental backup:

dsmc restore image h: -incr -del


• • •••



If you do not follow the steps exactly, two things can occur:

1. After the original image is restored, all files backed up with the incremental command are restored individually.

2. If you perform a backup image before performing an incremental, files deleted from the original image are not deleted from the final restored file system.

In the GUI, you can do the same restore image. However, you will need to choose all the options from the given windows.


• • • ••



Student Exercise


31

Student Exercise


• • •••

Unit 9: Backup and Restore MethodologySetting Up Open File Support


Lesson 7: Setting Up Open File Support

Open File Support for Backup Operations (Windows 2000, XP)

If the logical volume snapshot agent (LVSA) is installed and configured for open file support, by default Tivoli Storage Manager performs a snapshot backup or archive of files that are open (or locked) by other applications. The snapshot allows the backup to be taken from a point-in-time copy that matches the file system at the time the snapshot is taken. Subsequent changes to the file system are not included in the backup. You can set the FILELEVELTYPE parameter of the INCLUDE.FS option to DYNAMIC to specify which drives do not use open file support.

To control an open file support operation, you can specify these additional options in your dsm.opt file or as values of the INCLUDE.FS option: SNAPSHOTCACHELOCATION, SNAPSHOTCACHESIZE, SNAPSHOTFSIDLERETRIES, SNAPSHOTFSIDLEWAIT, PRESNAPSHOTCMD, and POSTSNAPSHOTCMD.

32

Setting Up Open File Support


• • • ••



Note: You can only set the FILELEVELTYPE option as a value for the INCLUDE.FS option, not as a stand-alone option in your dsm.opt file.

You can use the INCLUDE.FS option to set snapshot options on a per file system basis.

The snapshot cache used for open file backup cannot be stored on the same drive which is being backed up. Use the SNAPSHOTCACHELOCATION option to relocate the cache if necessary. You can specify a SNAPSHOTCACHELOCATION for a specific drive using the INCLUDE.FS option.

Configuring the LVSA to Enable Open File Support (Windows 2000, Windows XP)

If the logical volume snapshot agent (LVSA) is installed and configured for open file support, by default Tivoli Storage Manager performs a snapshot backup or archive of files that are open (or locked) by other applications.

To install or configure the LVSA to enable open file support:

1. Select Utilities >> Setup Wizard from the GUI main window. The Client Configuration wizard appears.

2. Select Help me configure the TSM Logical Volume Snapshot Agent to enable open file support and click the Next button. The Logical Volume Snapshot Agent Task panel appears.

3. Select the task you want to perform. You can install the LVSA, update the LVSA after updating the backup-archive client, or remove the LVSA from your system.

4. Complete each panel in the wizard and click the Next button to continue. To go back to a previous panel, click the Back button. To display help information for the panel, click the Help button.

Considerations

After installing, updating, or removing the LVSA, you are prompted to restart your system. To set preferences for open file support, use the Include-Exclude preferences tab.


• • •••



Student Exercise

Open your Student Exercises book and perform exercise 5 for this unit.

33

Student Exercise


• • • ••

Unit 9: Backup and Restore MethodologyBackup Groups


Lesson 8: Backup Groups

You can use the backup group command to create and back up a group containing a list of files from one or more file space origins to a virtual file space on the Tivoli Storage Manager server.

A group backup allows you to create a consistent point-in-time backup of a group of files that is managed as a single logical entity:

• All objects in the group are assigned to the same management class.

• Existing exclude statements for any files in the group are ignored.

• All objects in the group are exported together.

• All objects in the group are expired together as specified in the management class.

• A group backup can be added to a backup set.

• You can perform a full or differential backup using the mode option.

34

Creating and Using Backup Groups

dsmc backup group

LAN

SANServer

Client


• • •••



For example, to perform a full backup of all the files in the C:\dir1\filelist1 file to the virtual file space \virtfs, containing the group leader C:\group1 file, enter the following command:

dsmc backup group -filelist=c:\dir1\filelist1 -groupname=group1 -virtualfsname=\virtfs -mode=full

The Backup Group Command

Filelist

The Tivoli Storage Manager client opens the file you specify with this option and processes the list of files within according to the specific command. With the exception of the restore and retrieve commands, when you use the FILELIST option, Tivoli Storage Manager ignores all other file specifications on the command line.

The files (entries) listed in the filelist must adhere to the following rules:

• Each entry must be a fully or partially qualified path to a file or directory or a relative path.

35

The Backup Group Command


• • • ••



• Each entry must be on a new line.

• Do not use wildcard characters.

• Each entry results in the processing of only one object (file or directory).

• If the file name contains any spaces, enclose the file name with quotation marks.

• The filelist can be an MBCS file or a Unicode file with all Unicode entries.

• Tivoli Storage Manager ignores any entry that is not valid.

Groupname

Use the GROUPNAME option with the backup group command to specify the name for a group. Directory delimiters are not allowed in the group name since the group name is not a file specification, but a name field. You can only perform operations on new groups or the current active version of the group.

Mode

• FULL – Specifies that you want to perform a full backup of NAS, WAS, or group objects. This is the default for WAS and group backups.

• DIFFERENTIAL – Specifies that you want to perform a NAS, WAS, or group backup of files that changed since the last full backup. If an eligible full backup does not exist, a full backup occurs. This is the default for NAS objects.


• • •••



Backup Group Management

Query Group

Use the query group command to display information about a group backup and its members.

Use the PICK option to display a list of groups from which you can select one group to query.

Use the SHOWMEMBERS option to display and select individual group members that you want to query. The SHOWMEMBERS option is not valid with the INACTIVE option. If you want to display members of a group that are not currently active, use the PITDATE and PITTIME options to specify the backup date and time of the member you want to query.

Use the query filespace command to display virtual file space names for your node that are stored on the Tivoli Storage Manager server.

If you perform a full and differential group backup, a query of this group using the INACTIVE option displays two active backups of the same name, one of type FULL and one of type DIFF.

36

Backup Group Management

LAN

SANServer

Client

dsmc query groupdsmc delete groupdsmc restore groupdsmc query filespace


• • • ••



Delete Group

Use the delete group command to delete a group backup on the Tivoli Storage Manager server. You can also delete WAS group backups using this command. After deleting a group, the group leader (virtualfsname) remains on the Tivoli Storage Manager server. It contains no members (file or directories), but is reported in a subsequent query filespace command. It will have no files listed if the SHOWMEMBERS option is added. Deleting a group does not remove the file space that it resides in because there may be other groups in it. Use delete filespace if you want to remove the file space and all the data it contains.


• • •••



Student Exercise


37

Student Exercise


• • • ••

Unit 9: Backup and Restore MethodologyMoving Data By Node


Lesson 9: Moving Data By Node

You can use the move nodedata command to move data in a sequential access storage pool for one or more nodes, or move data in a single node with selected file spaces. You can also use move nodedata to consolidate data for off-site disaster recovery storage, or move data to another storage pool.

For this command, the data can be located on either a primary or copy storage pool. When the source storage pool is a primary pool, you can move data to other volumes within the same pool or to another primary pool. When the source storage pool is a copy pool, data can only be moved to other volumes within that pool.

During the data movement process, the server moves any readable files to available volumes in the specified destination storage pool. It attempts to bypass any files that previously were marked as damaged. You can only move node data for a storage pool using DATAFORMAT=NATIVE or DATAFORMAT=NONBLOCK.

38

MOVE NODEDATA

Client B dataClient A data Client C data


• • •••



Moving Data for All File Spaces for One or More Nodes

Moving data for all file spaces on one or more nodes is useful when:

• You want to optimize performance by reducing the number of volume mounts required during a restore operation by consolidating data for a specific node within a storage pool.

• You want to consolidate data for off-site disaster recovery storage.

• You want to consolidate data into a different storage pool.

• You want to increase performance of client restore processing by moving data to a random access storage pool.

• You want to free up all space on a volume so that it can be deleted from the Tivoli Storage Manager server.

Moving Data for Selected File Spaces for One Node

Moving data for selected file spaces for a single node is useful when:

• You want to prioritize the order of data to be restored, such as first restoring only business critical data and then restoring nonbusiness critical data.

• You want to consolidate specific file spaces into a different storage pool.

Note: You should not run move nodedata and migration processing concurrently on the same storage pool.


• • • ••



Student Exercise


39

Student Exercise


• • •••

Unit 9: Backup and Restore MethodologyAdministration through the Firewall


Lesson 10: Administration through the Firewall

Tivoli Storage Manager Firewall Support

In most cases, the Tivoli Storage Manager server and clients can work across a firewall. Because every firewall is different, the firewall administrator may need to consult the instructions for the firewall software or hardware in use.

To allow clients to communicate with a server across a firewall, the ports described in the following sections must be opened in the firewall by the firewall administrator.

TCP/IP Port

To enable the command-line administrative client to run outside a firewall, the port specified by the server option TCPPORT must be opened by the firewall administrator.

40

Administration through the Firewall

dsmserv.optTCPport 1500tcpadminport

by default is same as TCPport(change in 5.3)

dsm.optTCPserveraddressTCPport 1500 sessioninitiation client


• • • ••



This port is set on the client and the server using the TCPPORT option. The setting must be the same on the client and server. The default TCP/IP port is 1500. Remember, the client may not use the port specified by the TCPADMINPORT option (on the server) for client session. That port may be used for administrative sessions only.

TCP/IP Port for Administrative Sessions

The TCP/IP port for administrative sessions specifies a separate TCP/IP port number on which the server is waiting for requests for administrative client sessions, allowing secure administrative sessions within a private network.

This method requires a Tivoli Storage Manager server version 5.2 or higher. Additional setup on the server is required to enable server-initiated sessions.

The server cannot log events to a Tivoli Enterprise Console server across a firewall.

New Administrative Port

The TCPADMINPORT option is set in the server or the storage agent option file, and has the following considerations:

• The default is the same as the TCPPORT setting on the server. This feature was new in TSM 5.3. The previous default for the TCPADMINPORT setting was (1500) regardless of the TCPPORT setting.

• If value is the same as TCPPORT, a single thread is used, as in prior releases.

– Any session can use this port, although clients without authority to initiate sessions still cannot initiate them.

• If the value is not the same as that of the server, or if the server default is not used, and this option has not been specified, a separate thread will handle administrative-only sessions.

– Common server-to-server functions will use this new port.

– Administrative client option files do not have to be changed.

– Administrative sessions can still use the client port.

This was done to preclude wholesale migration changes.


• • •••



However, consider the following example server option file in an upgraded Tivoli Storage Manager 5.2 installation:

commmethod tcpiptcpport 1520

The server will start a client session handler on port 1520, and the server will start a separate administrative session handler on port 1500. The TCPADMINPORT option defaults to 1500. It does not default to the value of TCPPORT. Again, the TCPADMINPORT port information here is still valid for TSM 5.2, but note that beginning in version 5.3 the TCPADMINPORT defaults to the value of the TCPPORT setting.


• • • ••



The Backup-Archive Client through the Firewall

Tivoli Storage Manager Firewall Support

In most cases, the Tivoli Storage Manager server and clients can work across a firewall. Because every firewall is different, the firewall administrator may need to consult the instructions for the firewall software or hardware in use.

There are two methods for enabling client and server operations through a firewall, as described in the following sections.

Method 1

To allow clients to communicate with a server across a firewall, the following ports must be opened in the firewall by the firewall administrator.

• TCP/IP port – To enable the backup-archive client and the scheduler to run outside a firewall, the port specified by the server option TCPPORT must be

41

The Backup-Archive Client through the Firewall

dsm.optwebport cadport agentportorsessioninitiation serveronlytcpclientport xxxx


• • •••



opened by the firewall administrator. This port is set on the client and the server using the TCPPORT option. The setting must be the same on the client and server. The default TCP/IP port is 1500. Remember, the client may not use the port specified by the TCPADMINPORT option (on the server) for the client session. That port may be used for administrative sessions only (including node administrative client sessions).

• HTTP port – To allow the Web client to communicate with remote workstations across a firewall, the HTTP port for the remote workstation must be opened. Use the HTTPPORT option in the remote workstation's client option file to specify this port. The default HTTP port is 1581.

• TCP/IP ports for the remote workstation – The two TCP/IP ports for the remote workstation client must be opened. Use the WEBPORTS option in the remote workstation's option file to specify these ports. If you do not specify values for the WEBPORTS option, the default zero (0) causes TCP/IP to randomly assign two free port numbers.

Method 2

For the client scheduler, it is unnecessary to open any ports on the firewall. If you set the SESSIONINITIATION option to SERVERONLY, the client will not attempt to contact the server. All sessions must be initiated by server prompted scheduling on the port defined on the client with the TCPCLIENTPORT option. The SESSIONINITIATION option only affects the behavior of the client scheduler running in the prompted mode. By setting the SESSIONINITIATION option to SERVERONLY, the command-line client, native GUI, and Web client GUI cannot be used.

Note: This method (Method 2) will not work for (Tivoli Storage Manager Client Acceptor Daemon) CAD-managed schedulers.

If you set the SESSIONINITIATION option to CLIENT, the client will initiate sessions with the server (Method 1) by communicating on the TCP/IP port defined with the server option TCPPORT (this is the default). Server prompted scheduling may be used to prompt the client to connect to the server.

When using Tivoli Storage Manager across a firewall, consider the following:

• Method 2 requires a Tivoli Storage Manager server version 5.2 or higher. Additional setup on the server is required to enable server-initiated sessions.

• In prompted mode, the Tivoli Storage Manager server needs to contact the client. In order to do this, some software must be installed on the Tivoli Storage Manager server to route the request through the firewall. This software routes the server request through a socks port on the firewall. Proxies are not supported, since they only route a few types of communication protocols (HTTP, FTP, and GOPHER), and Tivoli Storage Manager does not use these communication protocols. It is


• • • ••



important to note that the client creates a new connection to the Tivoli Storage Manager server when prompted. This means that the firewall configuration must be in place.


• • •••



Review Questions

1. True or False: The restore from a backup set is referred to as a local restore.

2. True or False: It is possible to do a multi-session restore from the disk pool.

3. True or False: It is not necessary to have multiple tape drives to establish a multi-session restore.

4. True or False: You do not need the Tivoli Storage Manager server connection to restore from a backup set.

5. True or False: Adaptive subfile backup uses a differential algorithm to do each day’s backup.

6. True or False: The restore of a subfile backup needs 2.7 times the space to be able to do the restore.

7. True or False: The online image backup in Windows uses the LVSA.

8. True or False: Open file support requires the Logical Volume Snapshot Agent (LVSA).

9. True or False: A move nodedata command can move data from any storage pool to any other storage pool.

10. True or False: The move nodedata command makes it possible to prepare the data for a faster restore.

11. True or False: There is firewall support for backup-archive clients and administrative clients.

12. True or False: In the UNIX environment, it is possible to have multiple TCPSERVERADDRESS statements in a single dsm.sys file.


• • • ••





• • •••



Review Answers

1. True or False: The restore from a backup set is referred to as a local restore.

True.

2. True or False: It is possible to do a multi-session restore from the disk pool.

False.

3. True or False: It is not necessary to have multiple tape drives to establish a multi-session restore.

True.

4. True or False: You do not need the Tivoli Storage Manager server connection to restore from a backup set.

True.

5. True or False: Adaptive subfile backup uses a differential algorithm to do each day’s backup.

True.

6. True or False: The restore of a subfile backup needs 2.7 times the space to be able to do the restore.

True.

7. True or False: The online image backup in Windows uses the LVSA.

True.

8. True or False: Open file support requires the Logical Volume Snapshot Agent (LVSA).

True.

9. True or False: A move nodedata command can move data from any storage pool to any other storage pool.

False.

10. True or False: The move nodedata command makes it possible to prepare the data for a faster restore.

True.

11. True or False: There is firewall support for backup-archive clients and administrative clients.


• • • ••



True.

12. True or False: In the UNIX environment, it is possible to have multiple TCPSERVERADDRESS statements in a single dsm.sys file.

True.


• • •••



Summary

42


• Design and create backup-archive client solutions.• Configure and use multi-session backup and restore.• Set up and use backup sets.• Configure and use subfile backup and restore.• Configure and use online image backup and restore.• Enable and use open file support.• Configure and use file grouping.• Set up and use the move node data for faster restores.• Configure and use firewall support.• Configure and use space management on AIX (or HSM).


• • • ••



• • • ••

10-1

Unit 10: Enterprise Administration



• • • ••



Introduction

This unit discusses how to set up and configure server-to-server communications, command routing, and server groups. It also covers enterprise administration and implementation.

Objectives

Enterprise Administration

Server-to-server Communication–Part of the base Tivoli Storage Manager component which gives servers the ability to communicate with each other.

Administrative command routing–Allows an administrator to send commands from one Tivoli Storage Manager server to one or more Tivoli Storage Manager servers that the administrator is authorized to access.

Enterprise configuration–Allows an administrator to centrally define configuration objects and distribute those objects to managed servers.

2


•Set up server groups•Establish server-to-server communications •Configure single administrative logon•Invoke administrative command routing•Configure Enterprise Administration•Administer Enterprise Administration


• • •••

Unit 10: Enterprise AdministrationServer-to-server Overview


Lesson 1: Server-to-server Overview

For a pair of servers to communicate with each other, each server must be defined to the other. For example, if a configuration manager manages three managed servers, there are three server pairs. You can issue separate definitions from each server in each pair, or you can cross define a pair in a single operation. Cross definition can be useful in large or complex networks. When you set up communications among servers for any purpose, ensure that servers have unique names. At installation, a Tivoli Storage Manager server, if you accept the default, will have the same name as that of the machine. You can change this server name if necessary before setting up communication with other servers.

Configuring server-to-server communications is a prerequisite for the following Tivoli Storage Manager functions:

• Administrative command routing

• Central configuration

• Server-to-server virtual volumes

• Server-to-server event logging

• Server-to-server export and import (new to v5.3)

3

Remote servers

Local server

Server-to-server Overview

•Administrative command routing•Central configuration•Server-to-server event logging•Server-to-server virtual volumes


• • • ••



Server-to-server Configuration

Server-to-server communication uses TCP/IP as the communication protocol. Servers are defined with a server name, password, and TCP/IP address. The server name and password are used to authenticate sessions being initiated by the managing server. There are no password prompts when sessions are initiated between servers.

When setting up server-to-server in a one-to-many scenario for centralized management, the main Tivoli Storage Manager server can be referred to as the Local server and the other servers are called Remote. This is the normal setup in a campus or branch type Tivoli Storage Manager implementation. In a hybrid environment where all Tivoli Storage Manager servers can communicate with all other Tivoli Storage Manager servers, each server essentially is a Local server to all the other Remote servers.

There are three steps used to set up server-to-server communications. They are:

• Set the needed parameters on the remote servers including CROSSDEFINE if desired.

• Set the needed parameters on the local server.

• Define the remote servers on the local server, or define all remote servers on each server.

4

Server-to-server Configuration

• Server-to-server protocol– TCP/IP protocol– Server name– Server password– Server TCP/IP address– Authentication

• define server command– Defines network

connections– Cross definitions– Stored in database on

local and remote servers

Remote servers

Local server


• • •••



Set Up a Remote Server

The remote server needs have the parameter dealing with communications set. The parameter settings discussed below also have to be set on the local server with the exception of crossdefine, unless you will have a hybrid type implementation. Remember cross definition performs two commands when you enter one. You define on the local server, and the cross define will distribute the definition of the local server on the remote server automatically.

set servername

This command is used to set an appropriate server name. Each server should be identified with unique server name and should not use the default server name. Server-to-server communication uses this name to identify the server.

set serverpassword

This command is used to define the server password, and once defined, the server password cannot be displayed by an administrator. The server password can be reset by issuing the set serverpassword command with a new password. The server password is used to

5

Remote Server Setup 3 Setup Steps• Remote server settings• Local server settings• Define remote servers on

local server

Parameter settings on remote servers:• Set servername•Set serverpassword•Set serverhladdress 192.168.1.20•Set serverlladdress 1500•Set crossdefine on

set servername greyset serverpassword secret_greySet serverhladdress 192.168.1.20Set serverlladdress 1500set crossdefine on

query statusServer Name: GREYServer High Level Address: 192.168.1.20Server Low Level Address: 1500Server URL: Crossdefine: OnServer Password Set: Yes

The commands would also be performed on the cyan server

Remote servers

grey

cyan

Local server

black


• • • ••



authenticate sessions initiated from other servers. The server password set on the remote server will also be used when the server-to-server definitions are created on the local server.

set serverhladdress xxx.xxx.xxx.xxx

This option sets the IP address of the server you are administrating. This can be in dotted decimal notation or the DNS name of the server.

set serverlladdress 1500

This option sets the TCP port number the server will listen on for communications with clients and other servers.

set crossdefine on

This option enables the managed server to automatically create a server definition for the local server. Crossdefine reduces the number of server definitions administrators must create. When crossdefine is used, each remote server, when defined on the local server, automatically configures a definition for the local server on itself. The alternative to using crossdefine is to define each remote server to the managing server and then separately define the local server on each of the remote servers.

Viewing the parameter settings

The server name, server password, hladdress, lladdress, and crossdefine options can be viewed by using the query status command on the managed server.


• • •••



Set Up the Local Server

The four server definitions listed below are on the local server. These are passed to the remote server, and used by that server to perform the crossdefine operation of the local server on the remote server, if this function is set.

set servername – Used to set the server name on the local server. When the local server creates the server-to-server definitions, this server name is passed to the remote server.

set serverpassword – Used to define the local server password. When the local server creates the server-to-server definitions, this password will also be passed to the remote server.

set serverhladdress – Used to specify the TCP/IP address of the local server in dotted decimal format. When the local server creates the server-to-server definitions, this address is also passed to the remote server.

set serverlladdress – Used to specify the TCP port number of the local server. This is the TCP port where the local server listens for incoming TCP connections from other servers or clients. When the local server creates the server-to-server definitions, this port number will also be passed to the remote server.

6

Local Server Setup

set servername blackset serverpassword secret_blackSet serverhladdress 192.168.1.1Set serverlladdress 1500

query statusServer Name: BLACKServer High Level Address: 192.168.1.1Server Low Level Address: 1500Server URL: Crossdefine: OffServer Password Set: Yes

The commands would also be performed on the CYAN server

3 Setup Steps• Remote server settings• Local server settings• Define remote servers on

local server

Parameter settings on the local server:• Set servername• Set serverpassword• Set serverhladdress• Set serverlladdress

Remote servers

grey

cyan

Local server

black


• • • ••



Server Definitions

define server

The define server command is used to define the network connection between the local and remote servers. The command, issued on the local server, specifies the server name, password, TCP/IP address and TCP port of the remote server being defined. A crossdefine function can be enabled on the remote server being defined, so that the remote server receives this server configuration information and dynamically creates a reciprocal server definition for the local server on the remote server. This function reduces the number of manual entry server definitions required. To use crossdefine you must add the CROSSDEFINE=YES option to the server definition command in order to cause the cross definition.

The following example shows how you would set up a local server and a remote server so they could communicate with each other:

Local server (black) command to define server grey on server black and then defining server black on server grey, without using crossdefine:

define server grey serverpassword=secret_grey hladdress=192.168.1.20 lladdress=1500

define server black serverpassword=horse

7

Remote servers

grey

cyan

Local server

black

Server Definitions on Local 3 Setup Steps• Remote server settings• Local server settings• Define remote servers on

local server

Server black definition sent to remote server using crossdefine

Server Grey defined

define server grey serverpassword=secret_greyhladdress=192.168.1.20 lladdress=1500 crossdefine=yes

query server

Server Comm. High-level Low-levelName Method Address Address ----------- ----------- ------------------ -------------Grey TCP/IP 192.168.1.20 1500

define server command uses these parameters• Remote server name • serverpassword• hladdress TCP/IP address • lladdress TCP/IP port • crossdefine=yes

ping server commanddelete server command


• • •••



hladdress=9.115.4.797 lladdress=1823

Local server (black) command to define server grey on server black and automatically have server black defined on server grey, using crossdefine:

define server grey serverpassword=secret_grey hladdress=192.168.1.20 lladdress=1500 crossdefine=yes

ping server Command

This command can be used to verify that a defined server can be contacted and that the server passwords are valid. It is important to use this command to check the communications and not assume all the settings are correct. The ping server command will initiate a session to a defined server and will inform you of the success or failure of the session. This command uses authentication information that comes from the login of the command-line interface. If you are at a server console there is no authentication information available and the ping server command will fail. If you turn off authentication on the server you are pinging, the communication will succeed. Remember to turn authentication back on. Better yet, use the administrative command line for the command.

Also the administrator ID used to issue the ping server command must be registered on the server being pinged. The command starts an administrative session on the server being pinged, using the administrative ID that issues the command. If the ID is not registered, the ping server command will fail.

ping server grey

delete server Command

This command deletes the definition of a server configured for server-to-server communications. This command also deletes group member definitions for this server in any server group in which it has been defined.


• • • ••

Unit 10: Enterprise AdministrationServer Groups


Lesson 2: Server Groups

Overview of Server Groups

Tivoli Storage Manager servers that have been defined for server-to-server communications can be logically grouped together for administrative purposes. A server group allows the administrator to enter a command once using the server group name, and have the task automatically executed on the member servers defined within that server group. Server groups can be combined with single server names when issuing commands.

Group of Defined Servers—This is a collection of defined Tivoli Storage Manager servers that have common administrative requirements. The server groups can be organized by business organization, operating system, or any required combination. A server group is a logical definition of servers. To be included in a server group there must be a definition of the server on the local server. This means the local server must be defined to itself to be included in a server group.

Flexible Configuration—Any server or server group can be added to any other server group, making this a very flexible configuration. The groups can be defined as:

• Multiple server groups

8

grey

cyan

black

Servers

• Group of defined servers– Administration– Command routing– Organization– Platforms

• Flexible configuration– Multiple server groups– Subgroups– Servers in multiple groups

Overview of Server Groups


• • •••



• Groups within groups

• Individual servers defined in more than one server group

Server Group Configuration

Server Group Commands

• define servergroup – Defines a new server group on the Tivoli Storage Manager server. The positional parameter, group_name, defines the name of a server group.

• define grpmember – Defines one or more servers as members of a server group. The members to be added can be specified as a list of server names, separated by commas, with no spaces. The maximum length of any server or group name is 64 characters. The server must be defined with the define server command on the server where the server group is being defined before it can be added to a server group. The server where the define servergroup command is being run must be defined to itself with the define server command in order to be added to a group. The definition is used for the purpose of communication with other servers, much

9

define servergroup TSM_Win

define grpmember TSM_Win grey,cyan

query servergroup

delete grpmember TSM_Win cyan

Server group commands• define servergroup• query servergroup• define grpmember• delete grpmember

Defining server

TSM_Win

grey

cyan

black

Server Group Configuration


• • • ••



like an e-mail user will have a contact for himself in his own contacts list to facilitate contact with other users and sometimes himself.

• query servergroup – Displays server groups and the servers defined to those groups

• delete grpmember – Deletes one or more servers, as members, from a server group. Multiple members of a group may be deleted by specifying a list of server names to be deleted, separated by commas with no spaces.

• copy servergroup – Creates the new group. If the new group already exists, the command fails.

• update servergroup – Updates a server group to modify its description.

• rename servergroup – Renames an existing server group.

• delete servergroup – Deletes a server group from the Tivoli Storage Manager server.

• move grpmember – Moves a server from one server group to another.


• • •••

Unit 10: Enterprise AdministrationMulti-server Administration


Lesson 3: Multi-server Administration

Multi-server Administration OverviewS

Multi-server administration in a Web-based environment has changed with the advent of the new Administration Center. In the previous Web Administration tool, you would create links to the other servers using the URL option in the server definition.

Now with the Administration Center, you add server connections and provide the administrator name and password along with the IP address and TCP port number for the server. This allows for different passwords to be used for the different connections. Through the Administration Center you will select the radio button of the server you want to perform an action on and the command is built and executed.

From the command line you will use the same command structures of command routing to perform tasks on other servers.

10

Multi-server Administration

In order to manage multiple servers in the new Administration Center you must first add a server connection. This shows the input screen that you fill out to create a connection to another IBM Tivoli Storage Manager server for future management.


• • • ••

Unit 10: Enterprise AdministrationMulti-server Administration


Multi-server Administration

Single administrative login in the previous Web Administration tool has been replaced with the Administration Center add server connection task, to configure a server, and the Enterprise Management portlet. This portlet provides actions you can use to configure server-to-server communications and enterprise administration. Before these can be accomplished you will need to add server connections in the Administration Center for the server you will want to manage. Once the connections are configured then the portlets will display the servers available and by selecting the radio button for the particular server you want to work on, you can use the action menu to select the task.

11Multi-server Administration – Administration Center

The screen from the Administration Center shows how to perform an action on a particular server.•Select column used to select the server name.•Use the Select Action menu to select the task or action.


• • •••

Unit 10: Enterprise AdministrationAdministrative Command Routing


Lesson 4: Administrative Command Routing

Command Routing Setup

This section describes how to set up communications for command routing. You must define the remote servers to the local servers, and the same administrator must be registered with proper authorization, on all servers involved. After the servers are configured, a new administrative command syntax is used to route commands to one or more of them. Routed commands are executed on the defined servers, and the command output is returned and formatted on the server where the commands were issued.

To enable administrative command routing:

• Server definitions must exist on both the server where the commands are issued, and the servers to which the commands are routed.

• The administrative ID and password on the server issuing the commands must be valid on the server where the commands are to be routed, and must have the appropriate authority or the command execution will fail. The password will be encrypted before it is sent across the network.

12

• Configuration• Server definitions• Administrative ID and

password• Appropriate administrative

authority

• New command syntax• Routing prefix

• Individual server• Server groups• No wild cards allowed

• Default is local server

server name,server_group: administrative command(server name,server_group) administrative command

TSM_Win

grey

cyan

Command Routing Setup


• • • ••



Command Syntax for Performing Command Routing

servername,server_group: administrative command

(servername,server_group) administrative command

The routing information is put at the beginning of the administrative command. It consists of one or more servers, server group names or a combination of the two separated by commas. The end of the routing information is denoted by a colon followed by the administrative command. An alternative syntax is to enclose the server or server group names within parentheses, without the colon before the administrative command. This alternative syntax must be used if command routing is used within server scripts. This is because in a script a colon is used as a label designator. Wild cards cannot be used in this routing information. If routing information is not specified, the command will be issued on the local server, where it was entered.

Command Routing Example

Administrative commands can be routed to individual servers, server groups, or any combination of servers and server groups.

13

TSM_Win

grey

cyan

black

tungsten : q occtungsten : q occ

• Individual server cyan:query occupancy

• Server groups TSM_Win:query stgpool

• Combination of servers and groups (TSM_Win,black) query db

• Command output– Returned to issuing server– Formatted by that server

Command Routing Examples


• • •••



Individual Servers

In the example, the query occupancy command is issued against a single server. If there is no definition for this server or the server cannot be accessed, the query command will fail.

Server Groups

In the example, the query stgpool command is issued against the TSM_W2K server group and all of the individual servers that are members of that server group.

Combination of Servers and Groups

In the example, the query db command is issued against the TSM_W2K server group and the BLACK server. The command is issued against the group members and the individual server.

Command Output

The command output from routed commands is returned to the server where the command was issued. The output is formatted so that output from individual servers is contiguous and identifies from which server the output was returned. Duplicate server names are ignored. The local server will resolve the server names to a list that contains no duplicated servers. If a command is routed to two server groups that contain the same server name, the command is issued only once for that server. If a single server fails to execute a command, a return code is displayed with an associated error message for that server. The order in which the results of the command are displayed is based only on the order the servers processed the request in a first done first displayed manner. There is no significance to the order in which the command results are returned.


• • • ••



Student Exercise

Open your Student Exercises book and perform Exercise 1for this unit.

14

Student Exercise


• • •••

Unit 10: Enterprise AdministrationEnterprise Configuration


Lesson 5: Enterprise Configuration

Enterprise Configuration Overview

The Tivoli Storage Manager enterprise configuration allows an administrator to centrally define configuration information, or objects, on a configuration manager and then distribute those objects to managed servers, using server-to-server communications.

A server is selected to be the configuration manager system. The administrator then defines objects in the database on the configuration manager. Once the objects exist, the administrator can define a configuration profile, and associate the objects with the configuration profile. There may be as many profiles as needed to accomplish the needed results.

When the managed server subscribes to this configuration profile, the objects associated with the profile are distributed to the managed server as managed objects. These objects cannot be changed on the managed servers. They must be changed on the configuration manager and then distributed to the managed servers. Notification of the need to refresh the subscribed profiles can be set to a number of minutes.

15

• Configuration manager• Configuration profiles• Configuration profile

commands• Managed servers• Managed objects• Profile subscribers

TSM managed server on Windows

TSM managed server on

AIX

TSM managed server on OS/390

Enterprise Configuration Overview

TSM configuration

managerserver on AIX


• • • ••



Configuration Manager

The configuration manager is the Tivoli Storage Manager server used to define a central configuration and propagate it to one or more managed servers. In an enterprise environment, there could be one or more configuration managers, but a managed server can only be subscribed to by one configuration manager at a time. Normally a configuration manager is a local server in the server-to-server communications realm. In the following pages you will see how to set up enterprise administration from the command line. This will be followed by pages with brief descriptions on how to set up enterprise administration with the new Administration Center utility.

Server Definitions

To propagate the central configuration, the configuration manager must be able to communicate with its managed servers using server-to-server communications. The managed (remote) servers must be defined on the configuration manager (local) using the define server command, and the configuration manager (local) must be defined to the managed (remote) servers.

16

• Server definitionsManaged serversCross defined to configuration manager

• Configuration manager setupSet configmanager onDefault profile created

Configuration Manager


• • •••



Configuration Manager Setup

The server must be enabled as a configuration manager by issuing the set configmanager on command. The query status command can be used to display the values for the CONFIGMANAGER setting. A server cannot be set as a configuration manager if it already subscribed to a configuration profile defined on another configuration manager.

When the set configmanager on command is issued, a configuration profile named default_profile is created. This default profile contains references to all servers and server group objects defined on the configuration manager.

You can turn off the configuration management on the local server by using the set configmanager off command. This command will only work if there are no configuration profiles on the configuration manager. This applies even if you set the configmanager to on because this task creates the default profile.


• • • ••



Configuration Profiles

A configuration profile contains a set of defined objects on the configuration manager that are distributed to managed servers which subscribe to the profile. The configuration profile only contains references to the objects defined on the configuration manager, and not the actual objects.

Default Profile

The default profile is the first profile created on the configuration manager. It is automatically created when the set configmanager on command is issued. The name of the default profile is always default_profile, and it contains associations to all server and server group definitions within the database on the configuration manager.

When a managed server subscribes to a configuration profile, it automatically subscribes to the default profile as well. A managed server can also directly subscribe to the default profile by specifying the default profile name, using the define subscription command.

The default profile can be updated with associations to additional objects for distribution to managed servers. Any additional objects associated with the default profile are distributed to the managed server.

17

• Default profile– Server definitions– Server group definitions

• Additional profiles• Managed objects

– Administrators– Policy domains– Command schedules– Server scripts– Client option sets

• Servers and server groups• Profile subscribers

Configuration Profiles


• • •••



The default profile can be deleted (for example, to avoid the distribution of the server and server group definitions to the managed servers). If the default profile is deleted, it can be recreated by defining a new profile with the same default_profile name. The recreated default profile will not contain the server and server group associations. These will have to be reassociated by using the define profassociation command.

Additional Profiles

In addition to the default profile, other configuration profiles can be defined. For example, a configuration profile could be defined for distributing administrator definitions. The administrators are first defined on the configuration manager, and then are associated with this configuration profile.

When a managed server subscribes to this configuration profile, the administrator definitions associated with it are distributed to the managed server. These administrator definitions become managed objects on the managed server.

Managed Objects

Objects associated with configuration profiles become managed objects on the managed servers, once distributed. They are managed objects because their attributes are determined on the configuration manager where they were originally defined, and they cannot be updated or modified on the managed server. Managed objects are stored in the managed server database and can be used on the managed server, without a connection to the configuration manager. The following managed objects can be received from the configuration manager:

• Administrators

• Policy domains

• Administrative command schedules

• Server scripts

• Client option sets

• Servers and server groups

A managed object cannot be modified or deleted on the managed server. Whenever a managed object has to be modified, the modifications must be performed on the configuration manager. The next time the managed server refreshes the associated configuration profile, the modified object will replace the original on the managed server. Managed servers refresh the configuration profiles to which they are subscribed on an administrator-defined refresh interval. If required, the configuration manager can notify the managed server to immediately refresh its configuration.


• • • ••



Profile Subscribers

All managed servers that subscribe to at least one configuration profile, on the configuration manager, become profile subscribers.


• • •••



Configuration Profile Commands

• define profile–This command creates a configuration profile on the configuration manager. Only the configuration profile definition is created, and the defined profile does not contain any object associations.

define profile cyan_profile description=’Profile for server Cyan’

• update profile–This command updates the description field of a configuration profile.

update profile cyan_profile description=’New profile for server Cyan’

• copy profile–This command copies a configuration profile and its associated objects to another configuration profile.

copy profile cyan_profile grey_profile

• query profile–This command displays all profiles defined on the configuration manager. This command can also be used on managed servers to display profiles defined on the configuration manager.

18

define profile command

define profile cyan_profile description=’ Profile for server Cyan’

update profile command

update profile cyan_profile description=’New profile for server Cyan’

copy profile command

copy profile cyan_profile grey_profile

query profile command lock profile command

query profile cyan_profile lock profile cyan_profile 45

unlock profile command delete profile command

unlock profile cyan_profile delete profile cyan_profile

Configuration Profile Commands (1)


• • • ••



• lock profile–This command locks a configuration profile so that the associated objects are not propagated to subscribing managed servers.

lock profile cyan_profile 45

The optional parameter is the time, in minutes, that the profile will remain locked. After that period it is unlocked. A value of 0 locks the profile permanently. Locking a profile is used to make changes to the profile while keeping the subscribed managed server from refreshing the profile until the work is complete.

• unlock profile–This command unlocks a previously locked configuration profile, allowing the associated objects to be propagated to subscribing managed servers.

unlock profile cyan_profile

• delete profile–This command deletes a profile and its references to any associated objects. It does not delete the objects associated with the profile from the configuration manager or the managed objects from the managed servers.

delete profile cyan_profile

Managed servers should remove their subscription to the profile before it is deleted. If this is not done, the managed server will keep trying to refresh the profile, even though it has been deleted. If managed servers still have subscriptions to the profile, a FORCE=YES parameter can be specified to force the profile deletion. If one or more managed servers still have a subscription to that profile, the FORCE=YES parameter must be specified to delete the profile.


• • •••



Configuration Profile Commands (continued)

• define profassociation–This command associates one or more objects with a configuration profile to distribute them to subscribing managed servers. Multiple object types can be associated with a profile, and a match-all definition can be used to associate all defined objects of a type:

define profassociation cyan_profile admins=* adscheds=dbbackup,copy_stg

If an administrator object is removed on the configuration manager, the corresponding administrator managed object will also be removed on the managed servers when the profile is next refreshed. The objects must be defined on the configuration manager before they can be associated with a configuration profile.

• delete profassociation–This command deletes the association of one or more objects from a specific configuration profile. When the profile association is deleted, the objects are no longer distributed to the subscribing managed server.

delete profassociation cyan_profile adscheds=copy_stg

19

define profassociation

define profassociation cyan_profile admins=* adscheds=dbbackup,copy_stg

delete profassociation

delete profassociation cyan_profile adscheds=copy_stg



• • • ••



Configuration Profile Commands (continued)

• query subscriber–This command shows information about subscribers and their subscriptions to configuration profiles. The information includes the name of the configuration profile, and the name of the subscribing managed server. It also tells whether the configuration information is current or not, and when it was last updated.

• notify subscriber–This command notifies the subscribing managed servers that a configuration profile should be refreshed immediately from the configuration manager. When this command is issued, the managed servers refresh the configuration profile specified.

notify subscriber profile=cyan_profile

• delete subscriber–This command deletes all subscriptions for a managed server from the configuration manager database. This command is used to delete obsolete managed server subscriptions in cases where the managed server no longer exists or is unable to notify the configuration manager after deleting a subscription:

delete subscriber cyan

20

query subscriber command

query subscriber cyan

notify subscriber command

notify subscriber profile=cyan_profile

delete subscriber command

delete subscriber cyan



• • •••



Managed Servers

A managed server is a server that is subscribed to one or more configuration profiles on a configuration manager. A managed server receives objects on the configuration manager that are associated with the configuration profile. A managed server can only subscribe to one configuration manager, and it cannot also be a configuration manager for other managed servers.

Define Managed Servers

A server becomes managed by subscribing to one or more configuration profiles on a configuration manager. This is done by using the define subscription command, specifying the profile name to subscribe to, and the configuration manager name.

define subscription default_profile server=black

The first subscription made by a server does not have to be to the default profile. If the subscription is to another profile, the server is also automatically subscribed to the default profile on the configuration manager. Remember this fact because you may not want the server to also subscribe to the default profile.

21

Managed Servers

• Define managed serversdefine subscription

• Set refresh intervalset configrefresh

• Query configuration profiles query profilequery subscription

define subscription default_profile server=blackset configrefresh 60query profilequery subscription

• Remove profile subscriptiondelete subscription

• Configuration conflicts


• • • ••



Set Refresh Interval

When a managed server subscribes to a configuration profile, it periodically contacts the configuration manager to refresh the subscribed configuration profiles for any updates that have been made on the configuration manager. The configuration refresh interval is set on the managed server with the set configrefresh command and specifies the time in minutes that elapses between attempts by the managed server to contact the configuration manager for configuration refreshes. The query status command can be used to display the value for the CONFIGREFRESH option.

Query Configuration Profiles

Managed servers can query the configuration manager for all defined profiles and those profiles to which it is subscribed. The query profile command is used on a managed server to query profiles defined on the configuration manager.

By default, profiles known locally to the managed server are queried. Profile names are updated in the managed server database every time the managed server refreshes its configuration from the configuration manager. The query profile USELOCAL=NO option specifies that the managed server should contact the configuration manager for the list of profiles defined on that server.

• query subscription – This command is used on a managed server to query profiles on the configuration manager to which it is subscribed.

query subscription

Configuration Profile name Last updatemanager date/time--------------- --------------------------- --------------------------BLACK DEFAULT_PROFILEBLACK W2K_ADMINS 02/14/2005 16:08:17

Remove Profile Subscription

• delete subscription – This command can be used on managed servers to remove the subscription of a managed server from the specified profile on the configuration manager. Managed objects associated with the profile are left on the managed server. These objects are no longer managed and can be modified locally on the managed server.

delete subscription w2k_admins


• • •••



If the managed objects are to be deleted, the DISCARDOBJECTS=YES option can be specified. This option deletes the managed object associated with the subscription being deleted.

Configuration Conflicts

When a managed server receives or refreshes a configuration profile, all objects associated with that profile replace any locally defined objects of the same type on the managed server with the same name. For example, if an administrator ID is locally defined on the managed server and an administrator ID with the same name is also defined and associated with the configuration profile on the configuration manager, the local definitions are replaced by the definitions obtained from the configuration manager.


• • • ••



Using the Administration Center for Enterprise Administration

The replacement of the Web Administration tool with the Administration Center has changed the interface and how enterprise administration is managed in a GUI environment. To set up configuration management there first has to be server connections created to the servers to be managed. Then by accessing the servers using the Administration Center and performing the correct tasks, you can set up the configuration manager, profiles and subscribers.

The following pages will show how to use the Administration Center to set up enterprise administration.

22Using the Administration Center for Enterprise Administration

Use the Enterprise Management portlet to work on configuration management• Expand Tivoli Storage Management• Click Enterprise Management• The servers that have defined connections will be displayed• Add connection to other servers if needed




Set Up the Configuration Manager

To start the setup of enterprise administration, select the server you want to be the configuration manager (radio button) and then from the Select Action menu select Setup Enterprise Configuration. This will start a wizard to help you set up the configuration.

23

Set Up the Configuration Manager

•Select the server you want to be the configuration manager in the Select column• From the Select Action menu select Set Up Enterprise Configuration• A wizard will start to help you with the configuration




Beginning the Enterprise Configurations Wizard

The setup starts with a Welcome screen with information about the task.

24

Beginning the Enterprise Configuration Wizard

The wizard starts and presents the Welcome screen




Selecting the Role of the Server

This window has information about the server and what role you want for the server. If the server has already had a connection created for it, then the communications information will be prepopulated with the information; if not, you will need to provide it.

Then you will select the role at the bottom. Selecting None will change the status of a configuration manager or managed server to none meaning that the server no longer participates in the enterprise administration and will no longer retrieve profiles and objects. To change this status there are stipulations; a configuration manager must have all profiles it has defined deleted before it can be changed to a state where it is no longer a configuration manager. If you try to perform a task and the requirements are not met, you will receive a message in the Administration Center.

Selecting Configuration manager will create a default profile and in the background set the configuration manager on.

Selecting Managed server will allow you to select profiles to subscribe to and refresh rates.

25

Selecting the Role of the Server

• Enter the information for a server or accept the already entered data• Select the role the server is to take, in this case Configuration manager




Set Up Sharing - Create a Profile

The next task in the wizard is to set up a profile other than the default profile. You can enter a name and a description in this window.

26

Set Up Sharing – Create a Profile

Enter the name and description for the first profile on the configuration manager.




Select the Type of Objects for the Profile

The next window presents the available object types that can be included in a profile. You must place a check in the box for each type you want to add during this initial configuration. You can always add or delete objects after this initial configuration is completed. In the graphic above Administrator registration and authorities, and Policy Domains are selected. The next pages will show how the actual objects are added. The other selections are configured the same way.

27Select the Type of Objects to Place in the Profile – Share Objects

• Next add shared objects to the profile• Place a check in the box next to the object type you want to add• In the example above, Administrator registrations and authorities, and Policy Domains are selected (see next two slides)




Adding Administrators to the Profile

The object types checked in the previous window are presented in order for the actual selection of the objects to be included in the profiles. In the case above, the administrator, NEWADMIN, has been selected for inclusion.

28

Adding Administrators to the Profile

•After the object types are chosen, a list of each object type displays for selection.

•Place a check in the box next to each object you want to be included in the profile, in this case NEWADMIN.




Adding Domains to the Profile

After adding the administrators to the profile, the wizard then goes on to the next selected object type, in this case, Policy Domains. On this window you will select the policy domains that you want to include in the policy.

29

Adding Domains to the Profile

• After the object types are chosen a list of each type display forselection

• Place a check in the box next to each object you want to beincluded in the profile, in this case CMGRDOMAIN




The Summary Page

The summary screen will be displayed upon completion of the tasks. On the summary screen you will see a status of the events that occurred during the configuration wizard.

30

The Summary Page

• The summary page of the Enterprise Configuration is displayed• It will show the tasks that were completed as part of the wizard




View the Enterprise Properties

Once configured you can work on profiles by selecting to view the Enterprise Properties for the selected server.

31

View the Enterprise Properties

• Select the server that has been set up as the configurationmanager

• From the Select Action menu select View EnterpriseProperties




Working with Profiles for a Server

After selecting to view the properties of Enterprise Configuration a new portlet is displayed showing the profiles for the server selected. Selecting one of these profiles in the list and accessing the Select Action menu allows you to work on defined profiles. You can also choose to add a new profile from the menu.

32

Working with Profiles for a Server

Use the Select Action menu to create new profiles or select an existing profile and use the menu to make changes.




Profile Modifications

If you select a profile in the list and select to modify it, the screen above is displayed and you will use the list on the left side to perform tasks. Shown above is the General information. You can change the description of the profile selected on this page.

33

Profile Modifications

After selecting to modify the profile ALL_PROFILE you can choose to change the description or change the object associated with the profile.




Managed Administrators

Any administrator ID defined on the configuration manager can be added to a configuration profile with the exception of the SERVER_CONSOLE ID. These administrator IDs are propagated to subscribing managed servers as managed objects. If such an ID is subsequently deleted on the configuration manager, the associated managed object on the managed server is also deleted. Exceptions to this are: If the ID is in use on the managed server, it will not be deleted, and an administrator with system privilege will not be deleted if such deletion would leave the managed server without a system administrator.

Administrative Authority

Whenever an administrator ID is propagated, all privileges of that administrator ID on the configuration manager are propagated with the administrator ID to the managed servers. Therefore, a managed administrator ID always has the same authority on the configuration manager and on all managed servers to which the ID has been propagated.

Lock Attribute

The lock attribute of an administrator ID on the configuration manager is not propagated to the managed servers. This is a useful security feature. Administrator IDs can be defined for

34

Managed Administrators

•Administrative Authority• Same authority as administrator on the configuration manager.

•Lock Attributes• Lock attribute not propagated to the managed servers.• Will stay locked on the configuration manager.• Way to place administrator user IDs on managed servers and not allow

those IDs on the manager to be usable.

•Passwords• Password cannot be changed on the managed server unless it has

expired.• Set password expiration to the same value on manager and managed

servers.


• • •••



use only on managed servers by locking them on the configuration manager. When the ID is propagated, it will remain locked on the configuration manager, but it will not be locked on the subscribing managed servers. Such an ID can be used to administer the managed server but could not be used to log in to, or route commands to, the configuration manager.

Passwords

The password of an administrator ID is propagated to the managed servers. It is not possible to update the password of a managed administrator ID on the managed servers. The password must be updated on the configuration manager, and the managed servers must be notified to refresh the configuration profile.

If the password expiration value is not the same on the configuration manager and the managed servers, the password of a managed administrator ID can be changed if it is expired. As an example, the password expiration value is set to 90 on the configuration manager and to 10 on the managed server. If a managed administrator ID logs in to the managed server and his password has expired, it can be changed. This results in a password on the configuration manager that is different from the password on the managed server for the same administrator ID, until the password is next changed on the configuration manager, and the managed server is notified to refresh the configuration profile.

Note: Set password expiration the same on managed and managing servers.


• • • ••



Managed Policy Domains

Policy domains can be associated with a configuration profile. All objects within the policy domain are propagated as managed objects to the managed servers.

Policy Domains

A policy domain defined on the configuration manager can be associated with a configuration profile by associating the policy domain name. When such a profile is subscribed to by a managed server, the policy domain and its associated objects (policy set, management classes, and copy groups) will be propagated to the managed server and become managed objects on that server. If a locally defined policy domain with the same name exists, it will be overwritten. If that locally defined domain has nodes assigned to it, the propagation will fail. The nodes must be reassigned to another domain before the propagation can complete.

Client schedules defined for the policy domain on the configuration manager are also propagated to the managed server as an entity of the policy domain structure. The propagated client schedules on the managed server are managed objects and cannot be modified or deleted. The only operation allowed on a managed client schedule is its association with one or more client nodes. After a policy domain is associated with and propagated to a configuration profile, any modifications made on the configuration manager are also propagated when the managed server next refreshes the profile.

35

Managed Policy Domains

•Policy Domains• The policy domain and its associated objects (policy set, management

classes, and copy groups) will be propagated to the managed server.• Will overwrite a policy domain by the same name if it exists. Will fail if

nodes are member of domain.• Client schedules are propagated also and must be associated to nodes.

Cannot be modified or deleted on the managed server.

•Destination Storage Pools• Destination storage pool names cannot be changed in the copygroups on

the managed server.• Create storage pool or use rename stgpool command.

•Activated Policy Set• Only inactive policy propagated to the managed servers.• Validate on the managing server to check for warnings and errors.• Activate on managed server once the policy has been propagated.




When a policy domain associated with a configuration profile is deleted on the configuration manager, the managed object on the managed server is also deleted. Any nodes assigned to that domain must be removed before the deletion can complete.

Destination Storage Pool

The copy groups that are propagated with the profile contain pointers to destination storage pool names. These copy groups are managed objects on the managed server, and their destination storage pool name cannot be modified. To handle situations where the destination storage pool name in the propagated copy group does not match the actual storage pool name on the managed server, there is a new rename stgpool command. This command can be used on the managed server to rename an existing storage pool to match the name defined in the copy group managed object.

Activated Policy Set

The policy domain on the configuration server to be distributed can contain an activated policy set. However, the inactive policy set is propagated to the managed servers. After the inactive policy set has been propagated, it must be validated and activated on the managed server.




Managed Administrative Command Schedules

Administrative Command Schedules

When defining administrative command schedules to be propagated, care should be taken in specifying commands or scripts to be executed. All specified commands or scripts must exist on the managed server where the schedule will execute.

Activate the Administrative Command Schedules

An administrative command schedule must be defined on the configuration manager before it can be associated with a configuration profile. It can be defined either as an active or inactive administrative command schedule.

After propagation, the managed administrative command schedules are not activated on the managed server, even if they were on the configuration manager. The schedule must be activated on the managed servers in order for it to be executed. Such activated managed schedules are not automatically deleted on the managed server during a refresh following their deletion on the configuration manager. They must first be deactivated before they can be deleted.

36

Managed Administrative Schedules

•Administrative Command Schedules• When scheduling scripts, make sure the script exists on the managed server.

• Make sure all named objects used in the scripts and schedules exist on the server.

•Activate Administrative Command Schedules• The administrative command schedules are propagated with an inactive setting if set to active on the managing server.

• Must be activated on the managed server.• Must be placed in an inactive state before deletion can occur.




Managed Server Scripts

Server scripts defined on the configuration manager can be associated with a configuration profile for propagation to managed servers.

Operations

As with other managed objects, managed server scripts cannot be updated, renamed, or deleted on the managed servers. All those operations have to be performed on the configuration manager, and the managed servers have to be notified to update their configuration. However, a managed server script can be copied on the managed servers. The copied server scripts are unengaged and can be executed on the managed servers.

37

Managed Server Scripts

•Server Scripts

• Cannot be modified or deleted from the managed server.

• Make sure named objects in the script exist on the managed server.

• Managed scripts can be copied to another name on the managed server.

• Copied scripts are not managed objects and can be modified and deleted.




Managed Client Option Sets

Client option sets are defined on the configuration manager, and can be associated (by name) with a configuration profile. All client options defined in the client option set are propagated to the managed servers.

Node Association

The association of client nodes with client option sets is not part of the client option set and is not propagated to the managed servers. This node association must be performed on the managed server by using the register node or update node command with the CLOPTSET=OPTION.

Client Options

All client options within a managed client option set are managed objects and can neither be modified nor deleted on a managed server. Special attention must be paid to those client options that refer to objects like drives or file systems (DOMAIN option), paths (INCLEXCL option), or management classes. All such objects must exist on the managed servers and it does not matter whether they exist or not on the configuration manager.

38

Managed Client Option Sets

•Node Association• Node associations are not propagated.• Use the register or update node to associate propagated client option set with local nodes.

•Client Options• Option in the option sets in the profiles cannot be modified at the managed server.

• Be extra careful about options in the option sets that use file space, drive, directory or file names, as these must exist locally to the client associated with client option sets. Prime entries for this problem could be include and excludes.




Managed Servers and Server Groups

Server and server group definitions defined on the configuration manager can be associated with a configuration profile for propagation to managed servers.

Default Profile

When the set configmanager on command is issued on a server to create a configuration manager, a default profile is automatically created on that server. This default profile has all server and server group definitions on the configuration manager associated with it. When a managed server subscribes to a configuration profile, it is automatically subscribed to the default profile as well. In addition to the default profile, server and server group definitions can be associated with other configuration profiles.

Server Definitions

When a server is associated with a configuration profile, only a subset of the server definitions is propagated to a managed server. These are the communication options required to establish a server-to-server connection: server name, communications method, TCP/IP address, server password, Web address, and description. The propagated server definition becomes a managed object on the managed server and cannot be modified.

39

Managed Servers and Groups

•Default Profile• Default profile, default_profile, is created on the configuration manager when the setting is turned on.

• Contains all the server and server group definitions.• Automatically subscribed to by managed servers.

•Server Definitions• Server definitions are propagated but only a subset of settings dealing with communications.

• Cannot be modified.• Not automatically replaced. To replace use the Allowreplace=yes option on the define server command.

•Server Groups• Server groups and server definitions must have unique names.




A server definition propagated from a configuration manager is different from other managed objects in that it does not automatically replace a locally defined server definition with the same name. When a server is defined with the define server command, there is an option, ALLOWREPLACE, that determines whether a locally defined server on a managed server can be replaced by a server definition of the same name, propagated from the configuration manager. The default is no.

Server Groups

Server groups defined on the configuration manager can be associated with a configuration profile and propagated to a subscribing managed server.

A server group propagated from a configuration manager replaces a locally defined server group with the same name on the managed server. However, servers and server groups must have unique names. If a server group is propagated and there is a server definition with the same name on the managed server, the server group propagation will fail.




Profile Propagation

There are three ways in which configuration profiles are propagated from the configuration manager to managed servers. The refresh operation is always performed by the managed server.

Profile Subscription

When a managed server subscribes to its first configuration profile, it is also subscribed to the default profile of its configuration manager. This process creates two subscriptions on the managed server and immediately propagates the objects associated with both profiles to the managed server.

Managed Server Refresh Value

The value set by the set configrefresh command on the managed server determines the interval, in minutes, that elapses before the managed server contacts the configuration manager to refresh profiles to which it is subscribed. When the set configrefresh command is issued with a value greater than zero, the managed server immediately contacts the configuration manager to refresh the configuration. Once the refresh is complete, the timer starts counting down to the next refresh, according to the value specified.

40

• Profile subscription– Default profile– Other profiles

• Managed server refresh value• notify subscriber

command

Profile Propagation




notify subscriber – This command can be issued on the configuration manager to notify one or more managed servers to immediately refresh their subscriptions from the configuration manager. The configuration manager cannot push profiles to managed servers.

Profile Subscription View

A managed server administrator uses the query subscription command to view subscriptions for that server. The command can be issued only on managed servers or the Web administrative interface. The administrator can see only current subscriptions for that managed server.




Profile Subscribers

Whenever a managed server subscribes to a configuration profile, it becomes a subscriber on the configuration manager. An administrator on the configuration manager can view all profile subscribers.

Configuration Manager Views

The configuration manager administrator can use the query subscriber command to view all profiles and their subscribers. The command can be issued only on a configuration manager or the Administration Center interface. Views are also available to display all managed servers that are subscribed to a specific configuration profile, and all profiles to which a specific managed server is subscribed. These views display the current status of the subscription between the configuration and subscribing managed servers.

The Current Status

The current status is based on the currency of the configuration profile stored on the database of the configuration manager and on the database of the managed servers. If both the configuration manager and the managed server have the same profile, the current status indicates YES. If the configuration manager profile is newer than that of the managed

41

• Configuration manager views– Defined profiles– All subscribers – Subscribers to a specific profile

• The current status

Profile Subscribers




server, the current status indicates NO. The current status can also indicate UNCERTAIN; that is, the managed server has a more current profile than that of the configuration manager. This status can occur if the configuration manager database is restored to a previous level or a profile is deleted on the configuration manager while subscriptions remain on managed servers.




Student Exercise


42

Student Exercise


• • • ••



Review Questions

1. True or False: Server-to-server communication supports most communication protocols.

2. True or False: Administrative commands can only be routed from a configuration manager.

3. Server groups are (choose two):

a. A way of combining backup-archive clients together

b. A grouping of servers

c. Only for servers on the same platform

d. Used to route commands to multiple servers at once

4. True or False: A server that is defined as a configuration manager can also be a managed server.

5. True or False: Managed servers subscribe to configuration profiles.

6. Which of the following is true about a managed administrative ID?

a. The password can be changed on the managed server at any time.

b. The password can be changed on the managed server if it has expired.

c. Does not require a password.

d. There is no such thing as a managed administrative ID.

7. True or False: Storage pools, referenced by an activated policy set in a managed policy domain, must exist on both the configuration manager and the managed server.

8. True or False: You can delete the default configuration profile.


• • •••





• • • ••



Review Answers

1. True or False: Server-to-server communication supports most communication protocols.

False.

2. True or False: Administrative commands can only be routed from a configuration manager.

False

3. Server groups are (choose two):

b. A grouping of serversd. Used to route commands to multiple servers at once

4. True or False: A server that is defined as a configuration manager can also be a managed server.

False

5. True or False: Managed servers subscribe to configuration profiles.

True

6. Which of the following is true about a managed administrative ID?

b. The password can be changed on the managed server if it has expired.

7. True or False: Storage pools, referenced by an activated policy set in a managed policy domain, must exist on both the configuration manager and the managed server.

False

8. True or False: You can delete the default configuration profile.

True


• • •••



Summary

43


• Set up server groups

• Establish server-to-server communications

• Configure single administrative logon

• Invoke administrative command routing

• Configure Enterprise Administration

• Administer Enterprise Administration


• • • ••



• • • ••

11-1

Unit 11: Server-to-server Virtual Volumes

Unit 11: Virtual Volumes


• • • ••

Unit 11: Server-to-server Virtual Volumes


Introduction

This unit discusses how to configure and use server-to-server virtual volumes for primary storage pools. It also describes various ways to exploit virtual volumes.

Objectives

2


Configure server-to-server virtual volumes.

Administer virtual volumes for primary storage pools.Exploit server-to-server virtual volumes.


• • •••

Unit 11: Server-to-server Virtual VolumesServer-to-server Virtual Volumes Configuration


Lesson 1: Server-to-server Virtual Volumes Configuration

Overview

Server Operations

By using a server device class, server-to-server virtual volumes can be used for any operation that uses a device class, such as server import or export operations or database backups (only available with Disaster Recovery Manger).

Storage Pool Operations

Because the remote server is seen as a device class, a storage pool can be defined to use that device class. Storage pool operations such as client sessions (backup, archive, HSM),

3

OverviewDevice class defines a remote (target) server

Server operations can use remote server

Storage pool operations can use remote server

TCP/IP communications only

PASSWORDACCESS GENERATE automates theprocess

Disaster Recovery Manager


• • • ••



storage pool migration, and storage pool backup (only available with Disaster Recovery Manager) can use such a storage pool, which, in turn, uses the remote server.

TCP/IP Communications

TCP/IP is the only communications method that can be used for server-to-server virtual volumes.

passwordaccess generate

The feature, passwordaccess generate has been built into the servers, to provide secure server-to-server communications. If authentication is turned off for backup-archive clients, a password is still used for server-to-server virtual volumes.

DRM

The use of virtual volumes can be exploited for the storage of database backup and copy storage pool data on other servers. This enables an implementation that can eliminate the need for off-site storage facilities as each Tivoli Storage Manager site is essentially an off-site facility for the other. An appropriate distance between sites and other DRM requirements must be met for this to be acceptable practice.


• • •••



Prerequisites for Virtual Volumes

Server-to-server Communications

The servers involved in virtual volumes will use server-to-server communications. They must have their communication parameters set in order to define the servers to each other.

Archive Storage Pool

The data stored in virtual volumes is stored as archive objects on a target server. There must be an archive storage pool on the target server with enough volumes or space to store the data.

4

Prerequisites for Virtual Volumes

• Virtual volumes require server-to-servercommunications to be established.

• There must be an archive storage pool set upfor a policy domain on the server where thedata is to be stored.

•There are three steps to configuring virtualvolume capability after the prerequisites aremet.


• • • ••



Steps to Configure Virtual Volumes

Three steps to configure virtual volumes are listed on the slide above. Once these steps are accomplished there must still be utilization of the new device class that points to the remote server. This could be database backups or next pool settings for storage pool migration as an example.

5

Steps to Configure Virtual Volumes

• Register the source (local) server as a node ina domain on the target (remote).

• On the source server, define the target server.• Define a device class on the source server with

a devtype=server that points to the targetserver.


• • •••



Configuration - Step 1

Source server refers to the local server, the server from which data will be moved. The source server is the only server to which the client connects—that is, it is the first server in the hierarchy.

Target server refers to the remote server, the server to which data is moved. The client does not connect to this server. If the client chooses to connect to this server, it cannot restore or retrieve data that was backed up to the source server, even though the data resides on the target server. This inability to restore or retrieve is attributable to the different database entries of the two servers.

register node TYPE=SERVER

The first configuration step is to register the source server as a node on the target server. This is done using the register node command with the parameter, TYPE=SERVER. A node name and password are defined for the source server to use when communicating with the target server. Client workstations are still registered as nodes with TYPE=SERVER, which is the default TYPE parameter.

A node registered with a Tivoli Storage Manager server with TYPE=SERVER cannot be registered with the same Tivoli Storage Manager server as TYPE=CLIENT. However, a server node can be given any name.

6


(local)target serversource server

(remote)The source (local) server must be registered as a node on the target (remote) server.

register node TYPE=SERVER

query node command

On the target server issue:register node source passwd type=server

query node type=server

Registersourceserver

asnodehere


• • • ••



query node command

Nodes defined with a NODETYPE=SERVER can be queried with the query node command on the target server, specifying the TYPE=SERVER option:

query node type=server


• • •••




The second step of the configuration is to define the target server as a server on the source server. This is performed using the define server command.

Define Server Command

From the source server, the target server is defined as a server. When defining a server, the following parameters can be used:

• SERVERNAME—Specifies the name of the server. This parameter is used when a device class is defined to use the target server.

• COMMMETHOD—Specifies the communications protocol to use when connecting to the target server. The only entry can be TCPIP.

• HLADDRESS—Specifies the high address of the target server. For TCP/IP communications, this is the IP address in dotted decimal format.

• LLADDRESS—Specifies the low address of the target server. For TCP/IP communications, this is the port number of the target server.

7



(remote)The target server must defined as a server on the source (local) server

define server target

query server command

On the source server issue:

registernodehere

server

define server target commmethod=tcpip hladdress=1.2.3.4 lladdress=1500 nodename=source password=passwd delgraceperiod=7

query server


• • • ••



• DELGRACEPERIOD—Specifies the number of days that an object remains on the target server after it has been marked for deletion. The minimum value is five days; there is no maximum.

• NODENAME—Specifies the node name used by the source server to connect to the target server. This name must match the source server node name specified on the target server. If the parameter is not supplied, the name of the source server, as reported by a query status command, is used.

• PASSWORD—Is used to sign on to the target server. This must also match the source server node password defined on the target server.

Query Server Command

The administrative command, query server, can be used to query server definitions on the source server:

query server


• • •••




The third and final step of the configuration is to define a device class on the source server with the DEVTYPE=SERVER parameter. When defining a device class to the source server, the device type of SERVER can be used. The following parameters can be used:

• SERVERNAME—Specifies the name of the target server defined on the source server using the define server command.

• MAXCAPACITY—Specifies the maximum size of objects that can be created on the target server. An object can be a storage pool virtual volume, database backup, or exported Tivoli Storage Manager database data. The default is 500 MB. The amount of space specified by MAXCAPACITY is reserved on the target server before the source server sends any data. Therefore, if this value is greater than the size of the target server storage pool, the data will not be stored on that storage pool; it will be directed to the NEXTPOOL or the operation will fail.

• MOUNTLIMIT—Specifies the maximum number of simultaneous sessions that can be used between the source server and target server. If this is not specified, it defaults to 1 which will limit the number of server-to-server sessions to 1.

• MOUNTRETENTION—Specifies the amount of time to retain an idle virtual volume. If a client requests data that is held on a target server, from a source server, retaining the virtual volume mount can help to reduce the total time taken for communication to the client.

8



(remote)Define a device class on the source (local) server that points to the target (remote) server, devtype of SERVER

define devclass devtype=SERVERquery DEVC command


registernodehere

server

define devclass targetclass devtype=server servername=target\mountlimit=5 maxcapacity=10M

query devclass targetclass

Server-to-server virtual volumes are now configured


• • • ••



• PREFIX—Specifies the beginning portion of the high-level archive file name on the remote server, to control the volume names that are generated.

• RETRYPERIOD—Specifies the retry period in minutes to restart communications in the event of a communications failure.

• RETRYINTERVAL—Specifies how often the retries are done within the retry period.

Query devclass

The query devclass command can be used to display information about device classes defined using a device type of server:

query devclass targetclass


• • •••



Primary Storage Pool

The storage pool is defined in the usual way, referencing a device class. In this case, the device class will be the target server.

Storage Pool Volumes

As data is migrated from the backuppool on the source server, the data is stored in volumes in the target pool (also on the source server). These volumes are actually created as archived objects on the target server. The size of the volume is determined by the MAXCAPACITY parameter of the device class. These volumes appear as archived objects on the target server, belonging to the file space of the source server node.

9

Primary Storage Pools


(remote)• Storage pool defined on source

server using target server devclass• Storage pool volumes stored as

archive objects on target server


registernodehere

server definitiondevclassdefined

storagepool

volumesarchive objects

define stgpool targetpool targetclass \maxscratch=100 reusedelay=7

update stgpool backuppool next=targetpool

update stgpool backuppool hi=0 low=0

Storagepool


• • • ••



Database Backup - Disaster Recovery Manager Only

The device class can be used to back up the source server database. The database backup is stored as a number of objects on the target server. The number of objects is determined by the value of MAXCAPACITY of the device class.

The output from the database backup is stored in volumes, with size determined by the MAXCAPACITY parameter of the device class. These volumes appear as archived objects on the target server, belonging to a file space of the source server node.

Note: This is only supported with Disaster Recovery Manager.

10

Database Backup - DRM Only

backup db devclass=targetclass type=full


(remote)register

nodehere


storagepool

volumes

archive objects

• Source server is node on target server• Target server defined on source server• Device class using DEVTYPE=SERVER

pointing to target server


Database backup volumes are stored on the target server as archive objects


• • •••



Copy Storage Pool - Disaster Recovery Manager Only

The storage pool is defined in the usual way, referencing a device class. In this case, the device class will use the target server.

Copy Pool Volumes

As data is backed up from the backuppool on the source server, the data is stored in volumes in the copy pool also on the source server. These volumes are actually created as archived objects on the target server. The size of the volume is determined by the MAXCAPACITY parameter of the device class. These volumes appear as archived objects on the target server, belonging to the file space of the source server node. Knowledge of where the data is stored is on the source server database.

11

Copy Storage Pools - DRM Only


(remote)• Storage pool defined on source server

using device class pointing to targetserver

• Storage pool volumes stored asarchive objects on the target server


registernodehere


storagepool

volumesarchive objects

define stgpool targetcopy targetclass \pooltype=copy maxscratch=100 ...

backup stgpool backuppool targetcopy

StoragePool


• • • ••



Virtual Volumes

The source server is registered as a node to the target server, using TYPE=SERVER to differentiate it from client nodes. The target server is defined on the source server with the define server command. These two definitions, and a device class on the source server using TYPE=SERVER, are the only definitions on the source and target that define the relationship between servers.

Note: It is not possible for a node to be both TYPE=CLIENT and TYPE=SERVER. For the same machine to be both types of node, different node names must be used.

Nodes defined as TYPE=SERVER can only create archive objects on the target server. These archive objects reside in a file space on the target server, owned by the source server node.

The file space on the target server that belongs to the source server node is named adsm.server. It has a file space type adsm_fs to differentiate it from other client file space types such as NTFS (Windows NT), JFS (AIX), or DB2 (DB2). This file space can contain multiple storage pool volumes, database backups, or database exports from the source server.

Storage pools defined on the source server with a DEVTYPE=SERVER device class contain volumes in the same manner as for other device classes. These volumes on the source server are not real, they are virtual volumes.

12

Virtual Volumes


(remote)• Source server is a node of the target

server.• Server node file space on target server

is adsm.server.• Source server storage pool contains

virtual volumes.• Individual files can be retrieved• Client files in virtual volumes managed

on source server.•Cannot see individual files from the

target server.• Requires source server for restores.

registernodehere


storagepool

volumes

archive objects

StoragePool


• • •••



Virtual volumes are used in the same way as any other sequential media. However, it is not possible to append to a virtual volume. When a virtual volume is created it is not rewritten to. They always have a volume status of Full on the source server. Virtual volumes are defined using the device class defined on the source server. They appear on the source server in a similar format to those created with a DEVTYPE=FILE device class. Issuing the query volume command on the source server displays all volumes, both real and virtual:

query volume

The first two volumes shown above are real volumes in storage pools. The third volume is a scratch virtual volume. Virtual volumes are scratch volumes if the MAXSCRatch parameter of the storage pool is greater than zero. In this case, the names of the volumes created will take the form SERVER.ABC.123456789 where:

• server—The target server name, as defined on the source server.

• ABC—The three letters that denote the type of data stored:

– BFS Storage pool data

– EXP Export data

– DBB Database backup

– DMP Database dump

– DBS Database Snapshot

• 123456789 are the nine numeric digits used to give the virtual volume a unique name.

It is also possible to create private virtual volumes. This is done in the normal manner, using the define volume command and specifying a volume name. Private volumes are created empty and used when required.

The size of these volumes is the maximum of the MAXCAPACITY values associated with the device class and the size of the data operation. For example, if the data operation is more than twice the size of the MAXCAPACITY value, three volumes are created, and these appear as three archived objects in the adsm.server file space on the target server.

A Tivoli Storage Manager client can restore or retrieve as few files from the source server as needed, even when those files are held in the storage pool located on the target server.

Volume Status Storage PoolName Device Class Name Est.

Cap(MB)Pct.

UtilizedVolumeStatus

D:\Tivoli\Server\ARC01.DSM ARCHIVEPOOL DISK 100.0 0.1 On-Line

D:\Tivoli\Server\BAK01.DSM BACKUPPOOL DISK 100.0 77.3 On-Line

TARGET.BFS.870897083 BACKUPTAPE TARGETCLASS 20.0 100.0 Full


• • • ••



When files are held in the storage pool on the target server, the source server uses the Partial Object Retrieve mechanism to retrieve only those files from the storage pool, not the entire storage pool.

The target server sees the virtual volumes as archived objects. If these objects were managed by the target server's policy management, as with regular client archive files, there would be only one management class and copy group associated with each object

What is needed is the ability to keep multiple files in the same storage pool, such that each file can have a different management class. This can only be done if the source server controls the management of these files, because only the source server has database entries for each file. The target server has only one database entry for the entire storage pool volume.


• • •••



Managing Virtual Volumes

The policy management of the source server and target server is separate; each is unaware of the other's existence. To ensure that the source server manages the files held in virtual volumes, the target server's policy management is disabled for server nodes. The backup copy group attributes and Hierarchical Storage Manager attributes in the management class are never used for server nodes because server nodes can only have archived files.

All attributes of the archive copy group on the target server are ignored, with the exception of the DESTINATION attribute, which specifies the target storage pool.

The target storage pool is the storage pool on the target server where the source servers virtual volumes will be held. It is determined by the archive copy group of the default management class in the policy domain in which the source server is registered. It may be desirable to have a policy domain on the target server for server nodes, separate from the policy domain for regular client nodes, to ease administration.

Although remote server storage is seen as a single storage pool and device class, it may in fact be several storage pools. For example, the target server may accept the data from the source server into the target storage pool and later migrate the data to a different storage pool. The source server will be unaware of this migration, just as a Tivoli Storage Manager client is unaware of the location of its data.

The source server controls the expiration of virtual volume data. A target server cannot delete virtual volume information until requested to do so by the source server.

13

Managing Virtual Volumes

• Use separate policy management• Target storage pool for virtual

volumes• Expiration• Reclamation

Primary storage poolsCopy storage pools

• Reconciliation uses command reconcile volumes


(remote)

registernodehere


storagepool

volumes

archive objects

StoragePool


• • • ••



The source server retains individual files in the storage pool for the period specified by their policy management. After the specified period, the source server deletes its database entries.

Storage pools containing virtual volumes have reclamation options defined in the same way as other storage pools. When a virtual volume contains expired data, it is a candidate for reclamation which begins when a threshold is reached.

• Primary storage pool virtual volumes

For primary storage pool virtual volumes, data being reclaimed on the volume is transferred from the target server to the source server and back to the target server into a new virtual volume.

When reclamation has completed the reclaimed virtual volume will be marked empty and ready for deletion. The source server will request the target server to delete the volume.

• Copy storage pool virtual volumes

For copy storage pool virtual volumes, reclaimed data is copied from the primary storage pool wherever possible. For copy storage pool virtual volumes this copy operation minimizes the network traffic between the two servers to a single session, assuming that the primary volume is on the source server.

Reconciliation can be performed using the reconcile volumes command when differences in the virtual volume definitions on the source and target servers have occurred. For example:

• The target server was temporarily unavailable when the source server ran an expiration process on data that resides on the target server.

• One of the servers was subject to a disaster and a database restore had to be done to an earlier point in time.

• A delete volhist command was issued on the source server at a time when the target sever was unavailable.

• A virtual volume definition was deleted from the source server at a time when the target server was unavailable.

The reconcile volumes DEVICE_CLASS_NAME command can be issued from the source server to reconcile differences in the virtual volume definitions on the source server and archive objects on the target server. Both servers must be running to use this command.


• • •••



Server-to-server Virtual Volumes Considerations

Tivoli Storage Manager server-to-server virtual volumes duplicates data over the network when:

• One storage pool is backed up to another storage pool and both storage pools are held on the same target server.

In this situation, data is moved from a primary storage pool volume (which resides on the target server) to the source server, directly back to a copy storage pool volume (which also resides on the target server).

• Primary storage pool held on a target server is reclaimed.

In this situation, data is moved from one primary storage pool volume (which resides on the target server) to another primary storage pool volume (which also resides on the target server).

14

Server-to-server Virtual Volumes Considerations

• Server-to-server network load

• Performance tuning• Monitoring sessions source server

begins the backup storage operation

target serverholds the data

STGPOOL1 STGPOOL2


• • • ••



Performance Tuning

Two types of performance tuning can be done to optimize server-to-server virtual volumes:

• TCP/IP

TCP/IP system parameters such as TCP_SENDSPACE and TCP_RECVSPACE (examples from AIX) can be optimized with values that depend on the network topology being used. These parameters must be optimized at both Tivoli Storage Manager servers.

TCP/IP parameters in the target servers option file can be tuned, namely TCPWINDOWSIZE and TCPNODELAY.

• Server transactions

All server data is moved in transactions controlled by the parameters MOVEBATCHSIZE and MOVESIZETHRESH server options on the source server. MOVEBATCHSIZE specifies the number of files that are to be moved within the same server transaction and MOVESIZETHRESH specifies, in megabytes, the amount of data to be moved within the same server transaction. When either threshold is reached, a new transaction is started.

The target server treats data movement to or from the source server in the same way as for client nodes. The TXNGROUPMAX parameter in the server options file determines when a new node transaction is started. For best performance, this transaction size should be the same as the source server transaction size determined by MOVEBATCHSIZE and MOVESIZETHRESH server options.

Monitoring Sessions

When a server begins a session with another server, the session can be monitored just like any client node session:

• Tivoli Storage Manager Administration Center

• Tivoli Storage Manager administrative command-line interface


• • •••

Unit 11: Server-to-server Virtual VolumesUsing Server-to-server Virtual Volumes


Lesson 2: Using Server-to-server Virtual Volumes

Exploiting Server-to-server Virtual Volumes

15

Exploiting Server-to-server Virtual Volumes

UPSPower failure

Media failure

With Disaster Recovery Manager• Site disaster protection• Disaster recovery• Branch office model• Campus-wide implementation


• • • ••



Site Disaster Protection

Storage Pool Backup

The backup of a server's storage pool can be performed from that server to a copy storage pool residing on another server.

Database Backup

The backup of a server's database can be performed from that server to another server.

Peer Servers

Server storage pool and Tivoli Storage Manager database backup operations can be performed in any combination of servers: for example, two servers can back up each other. In this way, the relationship of the servers is equal, not hierarchical.

Note: Using server-to-server virtual volumes with database backup and copy storage pools is only supported with Disaster Recovery Manager.

16

Site Disaster ProtectionBackup storage pool from one server to another.Backup IBM Tivoli Storage Manager database of

one server to another.Peer-to-peer basis

–Different data centers or different sites–Any number of servers


• • •••



Remote Disaster Recovery Manager File

Disaster Recovery Manager is able to make use of server-to-server virtual volumes for storing the Disaster Recovery Manager plan file on a remote server. The following terms are used to differentiate between the two servers shown in this configuration:

• Source server–The server that runs Disaster Recovery Manager to prepare for disaster recovery. The source server creates the Disaster Recovery Manager plan files.

• Target server–The remote server that is the server which stores the Disaster Recovery Manager plan files of the target server.

The Disaster Recovery Manager plan file is accessible from both the target and the source servers.

17

Remote Disaster Recovery Manager File

Source server is node of target serverTarget server defined on source serverDevice class points to target server

prepare devclass=targetclass

Recovery plan of source server written as archived objects on target server.

prepare plan

Recovery plan file InputRecovery

plan file

node definition

server definition

anddevice class

target server(local)(remote)

source server

BlackGrey


• • • ••



Branch Office Model

Branch offices or workgroups may be constrained by the available network bandwidth. A branch office model can be found in the banking, insurance, and retail industries. Client data could be backed up to the local servers, which could be deployed on a regional server basis. The number of regional servers may be determined by the number of clients such a server is capable of backing up, given the limited network bandwidth available. To provide site disaster protection, a small amount of data determined to be business critical could be migrated or directly backed up to a central Tivoli Storage Manager server.

Multiple Server Storage Hierarchy

The Tivoli Storage Manager storage hierarchy can be implemented across multiple Tivoli Storage Manager servers. The simplest example of this implementation is for the first local Tivoli Storage Manager server to have a disk storage pool only and migrate data to the second Tivoli Storage Manager server with a tape storage pool.

The small local Tivoli Storage Manager servers could have their own storage hierarchy of disk and tape to manage some data. A Tivoli Storage Manager server hierarchy can be integrated with a Tivoli Storage Manager storage hierarchy.

18

Branch Office ModelIBM Tivoli Storage Manager storage hierarchy across multiple serversData directed to multiple TSM servers based on policyData location transparent to TSM clientsTape management minimizedScalability increased

TSM servers

TSM clients


• • •••



Policy-Based Data Placement

Different data of Tivoli Storage Manager clients can be bound to different management classes on the local Tivoli Storage Manager server. These management classes can direct the data to storage pools on that server or on a different server. A management class uses a remote storage pool in the same way as a local storage pool. When backing up files that are bound to such management classes, a single client session can back up data to multiple Tivoli Storage Manager servers.

Transparent Data Location

Just as Tivoli Storage Manager clients do not need to know which storage pool volumes hold their data, they also do not need to know which Tivoli Storage Manager server holds their data. However, a client always uses the same Tivoli Storage Manager server as a single point of contact. Only that server has the necessary database entries for that client's data. The same Tivoli Storage Manager server to which the client backed up must be available for the client to perform a restore.

Tape Management Minimized

Tape management can be minimized if only one Tivoli Storage Manager server has a tape storage pool. The local Tivoli Storage Manager servers may only have a disk storage pool and may migrate or back up storage pool data to a central Tivoli Storage Manager server with a tape storage pool.

Increased Scalability

This server hierarchy can increase Tivoli Storage Manager scalability because the number of local Tivoli Storage Manager servers is almost unlimited. These servers could act as a staging area to hold the data before it is migrated to another Tivoli Storage Manager server. One server alone may not be able to handle a workload where there are a large number of clients due to limited network bandwidth of the wide area network.


• • • ••



Campus-wide Implementation

Service Levels

Different service levels could be offered to the users (Tivoli Storage Manager clients) by configuring Tivoli Storage Manager policy management and storage management. Not only do these service levels state how long data is retained, but also the physical location of the data storage. This encompasses the previous example, where server-to-server virtual volumes are an extension of an existing server hierarchy.

Exploiting Hardware

By distributing Tivoli Storage Manager server implementation, existing server hardware can be utilized. Multiple servers can be implemented to spread workload. Centralized automated tape libraries can be utilized for multiple Tivoli Storage Manager servers.

19

Campus-wide Implementation

Disaster recoveryService levelsExploit existing hardware

–Servers–Tape libraries

Remote server volume sharing


• • •••



Volume Sharing

A Tivoli Storage Manager server can store its database backups and storage pool backups on another server. This server could send these objects to the same storage pool. Such a storage pool may use high capacity volumes such as IBM 3590, in which case both Tivoli Storage Manager database backups and storage pool backups of the source server node may be stored on the same 3590 tape. This provides the best utilization of such tape media, and is not possible without server-to-server virtual volumes.


• • • ••



Student Exercise


20

Student Exercise


• • •••



Review Questions

1. True or False: The target server must be defined as a node (type=server) to the source server.

2. True or False: devtype=server is a device class parameter for server-to-server virtual volumes.

3. List the three steps necessary to set up server-to-server virtual volumes.

______________________________________________________________

______________________________________________________________

______________________________________________________________

4. True or False: A source server primary storage pool can be defined on the target server.

5. Data created on the source server and stored on the target server consists of (choose two):

a. Virtual volumes on the source server

b. Backed up objects on the target server

c. Archived objects on the target server

d. Virtual volumes on the target server

6. Which statement is true concerning the management of virtual volumes?

a. Management is performed from the source server.

b. Management is performed from the target server.

c. Tivoli Storage Manager does not manage virtual volumes.

d. Virtual volume management is a manual process.

7. True or False: Server-to-server virtual volumes have little impact on your network.


• • • ••



Review Answers

1. True or False: The target server must be defined as a node (type=server) to the source server.

False

2. True or False: devtype=server is a device class parameter for server-to-server virtual volumes.

True

3. List the three steps necessary to set up server-to-server virtual volumes.

Register a node type=server on the target serverDefine the target server on the source serverDefine a device class of devtype=server

4. True or False: A source server primary storage pool can be defined on the target server.

False

5. Data created on the source server and stored on the target server consists of (choose two):

a. Virtual volumes on the source serverc. Archived objects on the target server

6. Which statement is true concerning the management of virtual volumes?

a. Management is performed from the source server.

7. True or False: Server-to-server virtual volumes have little impact on your network.

False


• • •••



Summary

21


Configure server-to-server virtual volumes.

Administer virtual volumes for primary storage pools.Exploit server-to-server virtual volumes.


• • • ••



• • • ••

A-1

Appendix A:Related Tivoli Storage Manager Products

Appendix A: Related Tivoli Storage Manager Products

A-2 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••



Introduction

This unit describes the various Tivoli Storage Manager related products, their implementation and usage. There are five data protection agents. Each of them will be briefly addressed.

Tivoli Storage Manager products covered will be:

• Tivoli Storage Manager for Application Servers

• Tivoli Storage Manager for Mail

• Tivoli Storage Manager for Databases

• Tivoli Storage Manager for Enterprise Resource Planning

• Tivoli Storage Manager for Hardware

• Tivoli Storage Manager Space Manager

• Tivoli Storage Manager Hierarchical Storage Management for Windows

• Tivoli Storage Manager for System Backup and Recovery (SYSBack).

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two A-3

• • •••



Objectives

2


• Describe Tivoli Storage Manager Data Protection products.• Utilize database backup methodology.• Describe the Storage Manager data protection architecture.• Automate database backup operations.• Describe how to use each of the data protection products.• Describe Space Management on AIX and HSM on Windows.• Understand basic storage area network concepts.• Describe Storage Manager for System Backup and Recovery.


• • • ••

Appendix A: Related Tivoli Storage Manager ProductsTivoli Storage Manager Data Protection Architecture


Lesson 1: Tivoli Storage Manager Data Protection Architecture

This slide shows you the basic architecture of all of the data protection solutions.

An application program interface (API) is linked and uses the Tivoli Storage Manager API to communicate with the Tivoli Storage Manager backup-archive client, the storage agent, or with the Tivoli Storage Manager server.

There is always some configuration work to be done on the application to set up communications. If you are going to do LAN-free backups and there is also configuration to be done between the TDP and the storage agent, you must add the appropriate entries to the dsm.opt file of the agent, the backup-archive client, and Tivoli Storage Manager API.

Most of the TDPs have their own specific requirements for the domain, and their own data management rules. It is recommended that you refer to the readme file and the installation and configuration manuals to help determine which rules apply for which TDP. Some of the TDPs create a unique file name for each backup of the database. Other TDPs put all the backups of the files under one file space name. Special utilities are used to delete old backups or Tivoli Storage Manager data management rules are used to expire old data.

3

Data Protection Architecture


• • •••

Appendix A: Related Tivoli Storage Manager ProductsDatabase Backup and Restore Overview


Lesson 2: Database Backup and Restore Overview

Tivoli Storage Manager Data Protection products are specifically designed to protect the large amounts of data that belong to special applications, such as databases and mail servers. Tivoli Storage Manager Data Protection products communicate between the application and the Tivoli Storage Manager server.

This makes online backup possible. Since databases can be very large, even an online backup can effect performance or cause the log files to fill up too fast. In such cases, special hardware devices or a special backup method like a full, incremental or differential backup would enable the customer to meet his service level agreement (SLA).

You should always check the IBM Web site to see which hardware or software is supported and which Data Protection products are available for your applications and platforms.

4

Database Backup and Restore Overview

application servers

hardware

mail

databases(oracle,sybase,informix...)

enterprise resource planning

OnlineOfflineDifferentialIncremental

storagepool


• • • ••



Tivoli Storage Manager for Databases

Tivoli Storage Manager protects the database by four basic methods.

The first method, Image Backup, backs up the database at the disk level to an image on the Tivoli Storage Manager server. This method is fast and supports raw volumes. However, to perform the backup, the database may need to be taken offline.

The second method backs up the data into an intermediate file using tools supplied by the database. Then, Tivoli Storage Manager backs up the data into storage pools. This could be an online backup. The application utility secures the data, and provides a record of the backup. The record will not be in sync with the file backup done by Tivoli Storage Manager, however. Since the database will not have the information necessary to find the backup file in Tivoli Storage Manager, restoring the database will be a two-step method. Storing intermediate files on the disk could also require almost twice the disk capacity, and processor resources. The work is done by the same processor that runs the application.

The third method is the basic way Tivoli Storage Manager backs up files. The consequence is again that the application must be stopped in order to do a consistent backup of those files. Another consideration is the fact that the database administrator does not really have control over which files are backed up. In fact, the database administrator has to ask the Tivoli Storage Manager administrator to provide the requested files based upon information only the administrator knows. It is not available with any tool or mechanism.

The fourth method is the method to be discussed in this unit. The backup is done in a very structured way by the tools of the application itself. All control files are under control of

5

Tivoli Storage Manager for Databases

raw devices

files

raw devices or files

2-stepprocess

typicalofflinebackup

Tivoli Data Protection

user program with TSM API

TSMcommand

intermediatefile

raw devices or files

TSMcommand


• • •••



the database administrator, and the tool takes care of the consistency of all the files backed up. This means the backup can be done online without having to bring down the database. When a restore needs to be done it communicates with Tivoli Storage Manager to request the files for the restore.


• • • ••

Appendix A: Related Tivoli Storage Manager ProductsDatabase Backup Automation


Lesson 3: Database Backup Automation

To automate backups, there is a function used to create command schedules in the Tivoli Storage Manager server. The purpose for this is to have a single point of control. This single point of control is very important, because most sites have limited resources available to back up their data.

The other reason is that an API has no scheduler. You have to use the backup-archive client scheduler to initiate these TDPs to start their backups.

You should also be aware that there is an option called schedcmddisabled which can be used to prevent the command schedule from running.

6

Database Backup Automation


• • •••



Backing Up and Restoring Database Servers

Data Protection Application Client for Oracle

Data Protection Application Client for Oracle supports both Oracle8 and Oracle9. In order to install and use Data Protection Application Client for Oracle, it is necessary to register the license with Tivoli Storage Manager. Root authority is also necessary to install the Data Protection Application Client on an AIX system.

Recovery Manager (RMAN) supports both online and offline backups and restores. The RMAN component also initiates the backup or restore process. The Data Protection Application Client for Oracle then acts as a pipeline into and out of the Tivoli Storage Manager server. Once the files are stored in Tivoli Storage Manager, they are managed the same way as any other backup files.

It is possible to have Data Protection Application Client for Oracle running on a system that also has a backup-archive client. Although it is possible for them to share the same option files, it is easier to manage the environment if you separate the files used by API applications from the interactive client files.

An API application has its own error log as well.

7

Database Backup and Restore Overview

application servers

hardware

mail

databases(oracle,sybase,informix...)

enterprise resource planning

OnlineOfflineDifferentialIncremental

storagepool


• • • ••



To separate the API application files from the interactive client, each has its own environmental variable to use.

When you no longer need an Oracle backup, you issue an RMAN command to delete backups from the Oracle backup catalog. A delete object API call is then sent to Tivoli Storage Manager, where the files are marked as inactive. These deleted files then expire from Tivoli Storage Manager during the next expiration process.

Note: In order to delete backup objects, the Data Protection Application Client for Oracle node should be registered with backdelete=yes specified.

Data Protection Application Client for Oracle Schedule

The Tivoli Storage Manager scheduler can be used to schedule an operating system command or script that initiates the Oracle backup process.

Microsoft SQL Server

Microsoft SQL Server allows for full database backups as well as incrementals. An incremental is a backup of the transaction log.

The Data Protection Application Client for SQL Server simplifies backing up to the Tivoli Storage Manager server, as well as restoring and deleting backup files from it.

The Data Protection Application Client must run on the same server as the SQL Server.

The SQL agent is registered as a client node, but does not require any special policy definitions. The files backed up by the agent are stored as backup objects rather than as archive objects.

The SQL agent must be registered by the Tivoli Storage Manager server.

Data Protection Application Client for Microsoft SQL Server Policy Management

Old policy management recommendations:

• Each object that gets backed up has a unique name. Because of this, it will always be active and never expire. Setting verdeleted and retonly to 0 helps remove the backed up files once they are marked inactive by the agent.


• • •••



• An incremental backup of the SQL Server means the transaction log was backed up. This is not the same as a Tivoli Storage Manager incremental backup, which backs up the changed database files only.

New policy management recommendations:

The way in which Data Protection for SQL Version 5.1.5 names the backup objects it stores on the Tivoli Storage Manager Server differs significantly from that used by Version 1. These changes were made in order to allow exploitation of the automatic policy-based expiration capabilities provided by the Tivoli Storage Manager server. As a result, do not use the same management class parameters recommended for Data Protection for SQL Version 1. The following are the current recommended policy settings.

Archive copy group

Data Protection for SQL stores all objects as backup objects on Tivoli Storage Manager in backup storage pools, so an archive copy group is not required, although it can exist.

Version control values

Set the following copy group parameters as desired to define the version limit and retention periods for SQL database backup objects:

– versions data exists

– versions data deleted

– retain extra versions

– retain only version

Copy group values

You should accept default values for the following backup copy group parameters because they are not applicable to Data Protection for SQL:

– copy mode

– copy serialization

– copy frequency

Data Protection Application Client for Microsoft SQL Server Administration

Data Protection Application Client for SQL Server has both a GUI and a command-line interface available. The agent uses an option file for the operation of the GUI as well as the Tivoli Storage Manager client option file.


• • • ••



Deleting a database backup is done through the Data Protection Application Client. It deletes the entry out of the backup history and tells the Tivoli Storage Manager server to mark the file as inactive. That file will not be deleted from Tivoli Storage Manager until expiration processing is run.

The Tivoli Storage Manager scheduler can be used to schedule command scripts, in order to automate the backup process.

Data Protection Application Client for Informix

Data Protection Application Client for Informix provides centralized, online, incremental backup capabilities and enables you to restore and manage Informix server databases and logical logs.

Data Protection Application Client for Informix is supported on the AIX and Sun Solaris platforms. The following Informix databases are supported:

• Informix 7 – Informix Dynamic Server

• Informix 9 – Universal Data option

Data Protection Application Client for Informix is certified by Informix, and provides data protection with the use of the Informix ON-Bar utility. ON-Bar provides both parallel backup and restore, and automatic backup of logical logs. Backups are stored on the Tivoli Storage Manager server. ON-Bar uses the X/Open Backup Services API (XBSA) to communicate with Tivoli Storage Manager.

This is not a Tivoli Storage Manager Data Protection application but is now an Informix solution that integrates with the Tivoli Storage Manager server.


• • •••

Appendix A: Related Tivoli Storage Manager ProductsData Protection for Applications


Lesson 4: Data Protection for Applications

Backing Up and Restoring Application Servers

Tivoli Storage Manager for Application Servers

The Tivoli Storage Manager Data Protection Application Client for WebSphere Application Server (WAS) is a solution for managing the backup and restore of an entire IBM WebSphere Application Server domain, including the administrative database, using Tivoli Storage Manager.

The Data Protection Application Client for WebSphere Application Server lets you manage consistent backups of your WebSphere Application Server domain as a single entity.

You select one of the WebSphere Application Server nodes as your master node. This will be the node from which you start the backup and restore operations for the entire WebSphere administrative domain, including the administrative database. This node acts

8

Backing Up and Restoring Application Servers

WebSphere application servers

storagepool


• • • ••



as the coordinator of all activities. All other nodes within your domain are considered to be controlled nodes. The backup and restore operations on these nodes are controlled by the master node. However, the Data Protection Application Client for WebSphere Application Server must be installed on all nodes within the WebSphere administrative domain.

Overview for WAS

Data Protection for WAS functions

Data Protection for WAS allows you to back up, query, and restore WAS Version 5.0 components using the Tivoli Storage Manager backup-archive client command-line interface and Web client.

WAS backup

Data Protection for WAS allows you to back up stand-alone application servers and Network Deployment configurations of WebSphere Application Servers. A Network Deployment configuration is backed up from the node that contains the Network Deployment Manager. Data Protection for WAS can also back up multiple instances of the Network Deployment Manager and Application Server concurrently. However, multiple concurrent backup sessions of the same node or cell are not supported.

9

Setup Overview for WASWebSphere Application Server (Master Node)

DB Server

DB2 UDB(Admin DB)

Log FileManager(db2uext2)ProleProle

Prole datamover

datamover

sharedvendorlibraryDB2 redo

log file

datamover

Storage Manage WebSphereApplication Server(controlled node)

StorageManager

File System(servlets, jars)

File System(servlets, jars)


• • •••



Data Protection for WAS backs up the following Network Deployment Manager and Application Server data:

• The properties directory

• WAS Version 5.0 Web applications:

– Java archive files (JAR)

– Class files

• Configuration information from the configuration repository

The following types of backup are available:

Full

A full backup is a complete backup of the following configuration files of the selected Application Server or Network Deployment Manager node:

• Configuration information from the WAS Configuration Repository

• All files in the properties directory

• WAS Version 5.0 installed Web applications

Note: The temp and backup directories are recreated every time WAS starts. As a result, these directories are excluded from a Data Protection for WAS backup. The backup directory is backed up only if it contains data.

This is the default backup.

Differential

A differential backup is a backup of files that have changed on the WAS node since the last full backup. A differential backup backs up a subset of files that are otherwise included in a full backup. Files that have not changed are not resent.

• For concept and strategy information related to backing up WAS data, see Backup Strategies in the documentation.

• For instructions on how to back up WAS data, see How to back up WAS data in the documentation.

WAS query

Data Protection for WAS allows you to query the Tivoli Storage Manager Server to display WAS backups and WAS instances. You can query both active and inactive backups. The query function is available using the query was command. The Web client automatically queries the Tivoli Storage Manager Server when the Restore window is refreshed.


• • • ••



WAS restore

Data Protection for WAS allows you to restore full or differential WAS backups that match the node name and type of WAS backup that you specify. You can also restore WAS backups that were backed up at a particular date and time by using the Tivoli Storage Manager backup-archive client pittime and pitdate options. This option allows you to specify the date and time at which you want to restore the latest version of your backup. Groups backed up on or before the specified date and time, and which were not deleted before the date and time you specified, are processed. When using the Web client, restoring data other than at the Network Deployment Manager or Application Server group level can corrupt your WAS installation. It is strongly recommended that you restore the entire Network Deployment Manager or Application Server group only.


• • •••

Appendix A: Related Tivoli Storage Manager ProductsData Protection for Mail Servers


Lesson 5: Data Protection for Mail Servers

Data Protection Application Client for Exchange assists in doing backups and restores of the Exchange directory and the information store database. It also deletes these backups from the Tivoli Storage Manager server. The difference between the types of backups has to do with whether the transaction logs are deleted or not. The Data Protection Application Client for Exchange is able to work with all four types of backups.

Configuring the Data Protection Application Client for Microsoft Exchange

The Data Protection Application Client must be registered as a node and must be registered with Tivoli Storage Manager.

The Data Protection Application Client for Microsoft Exchange does not require any special policy considerations.

All objects are stored as backed up objects as opposed to archived objects.

10

Data Protection for Mail Servers

Exchange and Domino servers

storagepool


• • • ••



Data Protection Application Client for Microsoft Exchange Policy Management

Old policy management recommendations:

• Each object that is backed up has a unique name. Because of this, it will always be active and will never expire.

• Setting verdeleted and retonly to 0 helps clear out backup files once they are marked inactive by the agent.

New policy management recommendations:

Data Protection for Exchange Version 5.1.5 differs significantly from Data Protection for Exchange Version 1 in the manner backup objects stored on the Tivoli Storage Manager Server are named. This difference allows exploitation of automatic, policy-based expiration capabilities provided by the Tivoli Storage Manager server. As a result, do not use Data Protection for Exchange Version 1 recommended management class parameters with Data Protection for Exchange Version 5.1.5. Use the following recommended policy settings for Data Protection for Exchange Version 5.1.5:

• Data Protection for Exchange stores all objects as backup objects on Tivoli Storage Manager in backup storage pools, so an archive copy group is not required, although an archive copy group can exist.

• Set the following copy group parameters as desired to define the version limit and retention periods for Exchange Server database backup objects:

– Versions Data Exists

– Versions Data Deleted

– Retain Extra Versions

– Retain Only Version

It is recommended that you discuss these parameters with your Tivoli Storage Manager server administrator in order to accomplish your backup strategy.

• You can accept default values for the following backup copy group parameters as they are not applicable to Data Protection for Exchange:

– Copy mode

– Copy serialization

– Copy frequency

• Incremental backups are always uniquely named. As a result, incremental backups do not participate in expirations (due to version limit) because there is never more than one version of an incremental backup object. However, all backup objects for an Exchange Server storage group are inactivated when a new full backup of that


• • •••



Exchange Server storage group is performed. Therefore, the retention period set in the Retain Only Version parameter controls the expiration of incremental backup objects.

• When setting the value of the Retain Only Version parameter for incremental backups, the value must be (at a minimum) as long as the value set for the full backup objects to which the incremental backups are associated. You can use the same management class for incremental backups and the full backup objects (that are retained the longest) to be sure an adequate value is used.

• A single restore can require a full backup, a differential backup, and possibly multiple incremental backups. Use collocation if these backups are stored on removable media. Use collocation by file space if you plan to restore multiple storage groups in parallel because all data for any one storage group is stored within one Tivoli Storage Manager server file space.

Data Protection Application Client for Microsoft Exchange Administration

The Data Protection Application Client for Exchange has both a GUI and a command-line interface. There is an option file for the operation of the GUI, as well as the client option file that is used by the agent.

A database backup is deleted through the Data Protection Application Client, which deletes the entry from the backup history and tells the Tivoli Storage Manager server to mark the file as inactive. That file is not deleted from Tivoli Storage Manager until expiration processing is run.

The Tivoli Storage Manager scheduler can be used to schedule command scripts to automate the backup process.

Data Protection for Lotus Domino

Data Protection for Domino is integrated with the Domino server, and has backup and restore capabilities. Data Protection for Domino does backups at the database level. Incremental backups are achieved by backing up the transaction log.

Data Protection for Lotus Domino Environmental Variables

Several environmental variables are used to locate various files:


• • • ••



• DOMI_ xxxx variables are Domino-specific.

• DSMI_ xxxx variables are for any API agent.

• DSM_ xxxx variables are for backup-archive clients.

Data Protection for Lotus Domino–Policy Management

Each database backup is a complete backup, not incremental. Incremental backups are accomplished by doing a backup of the database transaction log. Recovery using the transaction log requires a proper full database backup as well, so the data from an incremental backup alone is not sufficient to complete a restore process.

When a database backup is performed, a unique name is not generated. Because of this, normal version controls apply.

Domdsmc inactivatelogs expires transaction log files from backup storage. Because there is a single shared transaction log for all logged databases on a Domino server, log files cannot be inactivated (and allowed to expire) until all databases that require that log file for recovery are inactive. This command queries the database backups on the Tivoli Storage Manager server to determine which log files are required by any active database backup. This command also inactivates log files that are no longer required (because the database backups were inactivated). This command should be run after full database backups are completed, to inactivate the transaction logs at the same time the database backups requiring them are inactivated.

Since all objects are backed up, there is no need for an archive copy group.


• • •••

Appendix A: Related Tivoli Storage Manager ProductsTivoli Storage Manager for Enterprise Resource Planning


Lesson 6: Tivoli Storage Manager for Enterprise Resource Planning

Data Protection Application Client for mySAP.com

Data Protection Application Client for mySAP.com, also known as mySAP.com (and formerly known as BACKINT/ADSM) is an intelligent interface program that seamlessly connects the mySAP.com database administration utilities, SAP DBA (Oracle), with Tivoli Storage Manager.

Tivoli Storage Manager and Data Protection Application Client for mySAP.com provide a cost-effective, high-performance solution for backup and restore of mySAP.com systems with an Oracle database. This solution enables mySAP.com users to meet key business requirements for protecting their vital data, using a reliable and SAP-certified solution.

Data Protection Application Client for mySAP.com has an optional feature called Administration Assistant that provides the capability for customization and performance monitoring. It can be used free for 60 days during initial setup. For production use beyond the 60 days, a registration string will be supplied on CD-ROM if the feature is licensed.

11Tivoli Storage Manager for Enterprise Resource Planning

mySAP.com (R/3)

storagepool


• • • ••



Administration Assistant consists of the Assistant Server, which collects data and files from multiple Data Protection Application Client for mySAP.com instances, and Assistant Clients, which provide the administrator's Java-based interface. One license is required per Assistant server.

The combination of Data Protection Application Client for mySAP.com and Tivoli Storage Manager provides reliable and reproducible operation procedures so that system administrators can manage large volumes of data more efficiently. It enables mySAP.com administrators to back up or restore mySAP.com through the accustomed interfaces and functions of SAP DBA while using all the professional automated data management capabilities of Tivoli Storage Manager. Data Protection Application Client for mySAP.com is the leading external solution for mySAP.com data protection.

Incremental Backup (RMAN) Support

Data Protection Application Client for mySAP.com now supports incremental database backups using Oracle's Recovery Manager (RMAN). An incremental backup saves only those blocks within the database that have been changed since the last full backup, reducing the total amount of data to be backed up. Depending on the system environment, this may result in a significant decrease in backups. It may also lead to a reduction in the network load and to reduced backup system capacity needs.

Administration Assistant

The objective of the optional Administration Assistant (formerly Administration Tools) is to assist in configuration, monitoring, and administration of Data Protection Application Client for mySAP.com. With its Java-based GUI, it can operate from either a local or remote workstation.

The Administration Assistant is now enhanced to provide:

• Administrative support for incremental backup (Oracle).

• Administrative support for DB2 UDB.

• Simplified installation of the Administration Assistant system.

The Administration Assistant continues to provide:

• A fast overview of the status of all backups.

• A one-step customization process, including SAP DBA and Tivoli Storage Manager, instead of native editing of multiple initialization files.

• A replication capability eliminating the need for individual manual updates.


• • •••



• A dynamic visualization of backup and restore performance instead of analysis of log listings.

• A single control point instead of multiple logins.


• • • ••

Appendix A: Related Tivoli Storage Manager ProductsTivoli Storage Manager for Hardware


Lesson 7: Tivoli Storage Manager for Hardware

Data Protection Application Client for EMC Symmetrix

Data Protection Application Client for EMC Symmetrix integrates the power of EMC Symmetrix TimeFinder with the effective functions of Tivoli Storage Manager. These products offer exciting new options for implementing high-efficiency backup and recovery of business-critical applications, while virtually eliminating backup-related downtime or user disruption on the production host.

Using these features, you can benefit from:

• The support of EMC's TimeFinder Business Continuance Volumes (BCVs).

• Integration with Tivoli Storage Manager Media Management.

• High-performance, multiple-datastream backups.

12

Tivoli Storage Manager for Hardware

database servers

storagepool

hardware

sourcevolume

targetvolume

IBM ESS

database


• • •••



• Online, outboard database backup support for Oracle that exploits the RMAN utility and enables centrally administered and scheduled backup operations.

• Support for mySAP.com (Oracle), that is fully integrated with key SAP utilities, provides an automated restore process, and exploits Data Protection Application Client for mySAP.com performance and functionality.

Data Protection Application Client for ESS

Data Protection Application Client for IBM ESS for mySAP.com (hereafter Data Protection Application Client for ESS) minimizes the impact on mySAP.com Oracle database servers while allowing automated database backups to the Tivoli Storage Manager Server.

Integrated in the mySAP.com DBA tools, Data Protection Application Client for ESS off-loads the transfer of backup data from the database server. The Oracle database must reside on an ESS storage subsystem.


• • • ••



Functions of Data Protection Application Client for ESS

Operating Environment

The operating environment consists of the Oracle DBMS executing on an AIX server attached to an IBM Enterprise Storage Server (ESS). This AIX server is the production system. Another AIX server, the backup system, is also attached to the same ESS to back up ESS point-in-time copies of the production system to the Tivoli Storage Manager server. This is done by the concerted action of the programs of the products mySAP.com DBA tool (brbackup), Data Protection Application Client for ESS (idscntl), and Data Protection Application Client for mySAP.com (backint and prole). Data Protection Application Client for ESS requires Data Protection Application Client for mySAP.com to perform the actual backup or restore to or from the Tivoli Storage Manager server.

Data Protection Application Client for ESS provides certain inherent functions, referred to as active functions. As an agent operating in conjunction with mySAP.com DBA tool brbackup and Data Protection Application Client for mySAP.com, it also provides additional, passive functions, as described in the following sections.

13Functions of Data Protection Application Client for ESS

•Operating environment

•Active functions

•Passive functions

•Integration with mySAP.com DBA backup tool (brbackup)

•Types of backups


• • •••



Active Functions

• High-availability Oracle database backup using ESS FlashCopy devices.

• Integration with mySAP.com DBA backup tool functions to support running ESS Copy Services functions (FlashCopy, withdraw, and so forth).

• Keeps the progress of the Data Protection Application Client for ESS functions in a housekeeping file to monitor the proper sequential usage of the functions.

• Sends information to the Data Protection Application Client for mySAP.com Administration Assistant while Data Protection Application Client for ESS is running.

Passive Functions

• Seamless augmentation of the functions of Data Protection Application Client for mySAP.com.

• Centrally administered and scheduled backup operations.

• Integration with Tivoli Storage Manager media management functions.


• • • ••



Data Protection for Hardware and DB2 Overview

TDP for ESS for DB2 includes the following:

• Exploitation of ESS FlashCopy.

• Multiple datastream backups.

• Online, outboard database backups.

Data Protection for ESS provides these features:

• Backs up DB2 UDB databases with minimal impact and downtime on the production DB2 UDB database server.

• Restores DB2 UDB databases from Tivoli Storage Manager storage to your production system or your backup system.

• Performs a Quick Restore (FlashCopy restore) of your DB2 UDB database from the backup image on the Enterprise Storage Server target volumes to the production system.

• Automates backup operations.

• Integrates with Tivoli Storage Manager media management functions.

14Data Protection for Hardware and DB2 Overview


• • •••



• Supports IBM Subsystem Device Driver (SDD) functions.

• Includes globalization support.

• Backs up DB2 UDB database.

Data Protection for ESS uses the IBM Enterprise Storage Server FlashCopy feature to create a point-in-time copy of database volumes from the DB2 UDB production system. The copied database volumes are then made available for back up to a Tivoli Storage Manager server by a secondary host (backup system). Because the backup system performs most of the processing, the production system can dedicate processor time to other applications. This greatly reduces any backup-related performance impact on the production system.

DB2 UDB Database Restore

Use the DB2 restore command to restore a database from Tivoli Storage Manager storage. You can restore a database to either the source volumes or the target volumes on the Enterprise Storage Server.

DB2 UDB Database Quick Restore

Data Protection for ESS uses the IBM Enterprise Storage Server FlashCopy feature to restore a DB2 UDB database from the latest FlashCopy image available on the Enterprise Storage Server target volumes.

Scheduled Backups

You can schedule automated DB2 UDB backups from the Tivoli Storage Manager server. You can select when the backups occur without waiting for off-peak hours or maintenance downtime.

Integration with Tivoli Storage Manager

All Tivoli Storage Manager storage devices and media management capabilities are available to Data Protection for ESS. You can share the devices used for other backups, or give DB2 UDB exclusive use of certain devices and media. Data Protection for ESS supports media and tape management for off-site vaulting.


• • • ••



DB2—UDB Backup Alternatives

How UDB DB2 Stores Data Objects

Database objects stored on the Tivoli Storage Manager server by the DB2 backup command with the use Tivoli Storage Manager option are stored as backup objects. Each DB2 backup is stored as a unique object by specifying a time stamp as part of the low-level qualifier (ll_name). This means that the DB2 backups must be manually inactivated.

This is done by using the db2adutl delete command. This also means that the management class which the backup objects are bound to should have retention settings that change the inactivated backup objects to be expired immediately. The retention settings for a backup copy group that would facilitate this is retonly=0 and verdeleted=0.

The user exit that can be compiled to automate storage of the DB2 log files stores the logs as archive objects on the Tivoli Storage Manager server. The log archives should be manually deleted. This is done by using the db2adutl delete command. The management class that the archive objects are bound to should have retention settings that keep the archive objects forever until they are deleted manually. The retention settings for an archive copy group that would facilitate this is retver=nolimit.

15

DB2—UDB Backup AlternativesDatabase

storagepool

• DB2 uses the Tivoli Storage Manager API• Full online and offline backups• Table space backups• Log file backups


• • •••



Define a New Management Class in an Existing Domain

If you choose to define a new management class within an existing policy domain (which is not the default management class for that domain), then you must add an INCLUDE statement in the client options file (dsm.opt for Windows; dsm.sys for UNIX) that is used by the DB2 node. This include statement binds the DB2 backup objects to that management class that you have defined for managing these backups. The include statement would be:

include * ManagementClassName

With DB2 you can also specify the management class by setting the TSM_MGMTCLASS option within the DB2 database. For example, using TSM_MGMTCLASS the DB2 command would be:

db2 update cfg

Backup Copy Group Considerations

Normally, Tivoli Storage Manager backup copy groups are designed to hold multiple versions of files and directories to restore not only the latest (active) but also older versions that had been changed or deleted (inactive). The DB2 database will send its full and table space backups through the Tivoli Storage Manager client API directly to the backup copy group of the default management class to which the node is assigned. DB2 assigns a unique name to each database backup. The settings that pertain to multiple versions do not apply. The following retention settings should be used for the management class that will be bound to the DB2 backups:

• VEREXISTS=1–Keeps only one version of the backup file as the name of the backup is unique. There will not be a newer version of the backup image with the same name.

• VERDELETED=0–If the backup file has been deleted (via db2adutl), then Tivoli Storage Manager should not keep an inactive version of this file.

• RETEXTRA=0 (the same value as RETONLY)–The RETEXTRA parameter will never be used as you will never have more than one version of the backup file. To prevent confusion set this parameter to the same value as RETONLY.

• RETONLY=0–When a backup image file becomes inactive it will be purged from the Tivoli Storage Manager server at the next expiration.

Special Considerations

The values of VERDELETED and RETONLY can be changed so that DB2 database and table space backups become inactive for a period of time before becoming expired. To do


• • • ••



this, change VERDELETED=1 and RETONLY=number of days to keep inactive. You can then use the db2adutl command with the show inactive option to access the inactivated backup objects. Doing this is not recommended for a number of reasons. Archive logs do not have the same ability to be kept, once deleted. Online backups without the associated logs are worthless. Space is wasted on the Tivoli Storage Manager server during normal operation.

Archive Copy Group Considerations

Archive copy groups are designed to hold archive objects for a dedicated time and then to automatically delete the object. The DB2 database will send its log files through the Tivoli Storage Manager client API directly to the archive copy group of the default management class to which the node is assigned. The log files are numbered from S0000000.LOG to S9999999.LOG. During normal operation, it is unlikely to need to access such a vast number of log files. However, be aware that when doing a point-in-time restore of a DB2 database, some log files may be reused and therefore stored twice. It is recommended that you set the retention for the archive copy group to NOLIMIT.

RETVER: NOLIMIT

This will hold log files forever unless they are deleted by db2adutl. Like the deletion of the backup images, the deletion of the log files should be the responsibility of the DB2 administrator.


• • •••

Appendix A: Related Tivoli Storage Manager ProductsTivoli Storage Manager for System Backup and Recovery


Lesson 8: Tivoli Storage Manager for System Backup and Recovery

Tivoli Storage Manager for System Backup and Recovery 5.6 (also called SysBack) empowers you with a flexible backup method to help protect your data and provide bare machine recovery (BMR) capabilities for your AIX systems. It offers a comprehensive system backup, restore, and reinstallation tool. SysBack is a simple to use, yet highly effective, tool. Any feature may be executed from either the AIX command line or by using the SMIT menu interface.

Protect Your Most Vital Asset—Your Data

Whether your business is large or small, your company needs to protect your data under any circumstances. Data is the lifeblood of any business. The loss of critical data from equipment failure or environmental factors can be a damaging or fatal event. These circumstances make strong backup and recovery systems imperative, but due to the high cost of personnel time, these services need to be as simple and automated as possible.

16

Bare Machine Recovery with Tivoli Storage Manager for System Backup and RecoveryTivoli Storage Manager for System Backup and Recovery (SysBack) ensures that the AIX OS, LVMstructures and file systems can be recovered in the event of a disaster.


• • • ••



Tivoli Storage Manager Integration with SysBack

Comprehensive Backup and Recovery Options through a Simple, Efficient Interface

Tivoli Storage Manager for System Backup and Recovery (SysBack) allows you to choose from several types of backups, including full system (installation image), volume group, file system, file or directory, and raw logical volume. SysBack allows for recovery of all or part of the system, and is flexible enough to allow a system installation image from one system to be installed onto another system with either identical or different hardware configurations (cloning). Backup and recovery options are performed through a simple SMIT menu-driven interface or through a command line.

Central Management and Automation Tools

Tivoli Storage Manager for System Backup and Recovery provides utilities to create backup scripts and schedules for easier task automation. Furthermore, backup, list, and verify operations are quickly assessed using a completion status tracking log. SysBack's

17Tivoli Storage Manager Integration with SysBack

Local backupBackup to a locally attached disk file, tape or CD/DVD, or TSM server

Push backupRemote backup from the client to the server's disk file, tape or CD/DVD drive, or TSM server

TCP/IP network

tape

TSMserver

diskCD/DVD

servertape

TSMserver

diskCD/DVD

server


• • •••



pull client backups allow the administrator to centrally manage backup operations from a single server (remote or local). By combining all of these features, administrators are provided with a complete central management system.

Integrating with Tivoli Storage Manager

SysBack version 5.6 and later allows for the storage of backup objects into a Tivoli Storage Manager server. Combining the SysBack backup, restore, and network boot and installation functions with a Storage Manager server provides a bare machine recovery capability for Storage Manager configurations. SysBack will back up and recover a system’s volume group, logical volumes, and file system information. Additionally, SysBack has the capability to back up any non-root volume group data specified. You can use SysBack simply to recover the root volume group and then use Storage Manager to back up and manage other user data. Backups to a Storage Manager server may be manipulated like any other SysBack backup. They may be listed, verified, restored, and used for system reinstallation.

Integrating Network Boot and Installation Features in Support of RS/6000 SP Systems

Tivoli Storage Manager for System Backup and Recovery provides network boot features for those administrators that choose not to boot from locally attached tape devices. Network boot options include remote SysBack functions (classic booting using media), and utilization of existing NIM resources (NIM resource boot, network based). Using the SysBack NIM resource boot functionality, SysBack provides IBM RS/6000 SP system specific network boot and installation utilities for nodes. The SP boot and installation utilities allow for smooth installation and cloning processes for nodes within the same, or between different, SP complexes.

Provides Offline Mirror Backup Options

The offline mirror backup option splits specified AIX mirrors to allow Tivoli Storage Manager for System Backup and Recovery access to inactive copies of data. This allows simultaneous user and system access to the active copies.


• • • ••



Two-way Pull Backup Diagram

Pull backups help you centrally manage your backups by enabling you to initiate, and thereby control, backup operations from a single location. You can further automate this central management of backups using the SysBack scheduling and scripting functions.

The machine that pulls the backup is called the initiator while the machine that is backed up is called the backup system. The machine that receives that backup data is called the destination system. Pull backup enablement requires not only remote services configuration, but also remote command access configuration. The remote services configuration enables the data to be sent across the network from the backup system to the data destination machine. The remote command access configuration enables the initiator system to pull the backups from the backup system.

Pull backups can be done in a two-way or three-way manner. When the initiator system is also the data destination system, this is a two-way backup.

18

Two-way pull backupThe initiator (server) initiates the backup on the backup system (client).The backup system sends the backup to the data destination system (initiator).

Two-way Pull Backup Diagram

.

TCP/IP network

backupsystem(client)

initiator(server)

tape

TSMserver

diskCD/DVD


• • •••



Three-way Pull Backup Diagram

A three-way backup is when the initiator system is a different machine than the data destination system and different than the backup system. Note that for three-way backups, remote services must be configured between machine A, the backup system, and machine B, the data destination. Remote command access must be also configured between machine A, the backup system, and machine C, the backup initiator.

19

Three-way Pull Backup Diagram

.

TCP/IP network

backupsystem(client)

initiator(server)

tape

TSMserver

diskCD/DVD

Three-way pull backupThe initiator (server) initiates the backup on the backup system (client).The backup system sends the backup to the data destination system.

data destination(server)


• • • ••

Appendix A: Related Tivoli Storage Manager ProductsHierarchical Storage Management Overview


Lesson 9: Hierarchical Storage Management Overview

E

Tivoli Storage Manager for Space Management and Tivoli Storage Manager HSM for Windows

Tivoli Storage Manager for Space Management or Tivoli Storage Manager HSM for Windows, also known as Space Manager or HSM automates the movement of seldom-used files to and from near-line storage while presenting to the user the impression the data is still on disk.

HSM or Space Manager, maximizes usage of existing storage resources by transparently migrating data off of the workstation and file server hard drives based on size and age criteria, leaving only a stub file. When the migrated data is accessed, HSM transparently migrates data back onto the local disk. In doing so, HSM relieves the user from the task of manual deleting and archiving of data on his workstation.

20Tivoli Storage Manager for Space ManagementTivoli Storage HSM for Windows

//usr

/data

HSMclient

Migrate

Recall

Hierarchical StorageManagement (HSM)

stub

/data

TSMserver

LAN


• • •••

Appendix A: Related Tivoli Storage Manager ProductsHierarchical Storage Management Overview


You can find more information about this feature of Tivoli Storage Manager Extended Edition at the following Web sites:

AIX, HPUX, Linux, Solaris:

www.ibm.com/software/tivoli/products/storage-mgr-space/

Windows:

www.ibm.com/software/tivoli/products/storage-mgr-hsm/


• • • ••

Appendix A: Related Tivoli Storage Manager ProductsContinuous Data Protection for Files


Lesson 10: Continuous Data Protection for Files

Tivoli Continuous Data Protection (CDP) for Files provides continuous, real-time data protection of files. The files can be stored locally or saved periodically on a Tivoli Storage Manager server, network-attached file server or a removable disk.

When saved in real-time mode files are saved in the native operating system format not in a database. Files can also be saved in real time on shared file systems.

CDP also provides:

• Centralized reports of all machines protected.

• Global configuration changes and management.

• Select install to upgrade all computers running Tivoli Continuous Data Protection for Files.

• Delta-block transmission (to reduce the number of bits moved).

• Multiple versions per file for restore based on a point in time.

21

Continuous Data Protection for Files•Continuous Data Protection (CDP) for Files provides real-time protection of files.

•Files saved locally in native format.

•Two modes for saving files:• Real-time backups (continuous protection).• Scheduled backups.

•CDP can be configured to work with Tivoli Storage Manager using Tivoli Storage Manager client configurations.

•Configured and maintained using a browser-based GUI.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two A-41• • •••



• Archiving (locking down coherent groups of files from any future alteration).

CDP supports LAN-free backups and allows for support of snapshots and open file backup.

A-42 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007• • • ••



Review Questions

1. True or False? The Storage Manager API must be installed for every Storage Manager Data Protection solution.

2. True or False? All the Storage Manager Data Protection products provide multi-session backup and restore.

3. True or False? Each Storage Manager Data Protection Client must be associated with its own policy domain.

4. True or False? To schedule the backup of the database, the Storage Manager administrator must create a command schedule.

5. True or False? There is no Storage Manager Data backup solution available for DB2.

©Copyright IBM Corp. 2007 IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two A-43• • •••



Review Answers

1. True or False? The Storage Manager API must be installed for every Storage Manager Data Protection solution.

True.

2. True or False? All the Storage Manager Data Protection products provide multi-session backup and restore.

True.

3. True or False? Each Storage Manager Data Protection Client must be associated with its own policy domain.

False.

4. True or False? To schedule the backup of the database, the Storage Manager administrator must create a command schedule.

True.

5. True or False? There is no Storage Manager Data backup solution available for DB2.

True.


• • • ••



Summary

22


• Describe Tivoli Storage Manager Data Protection products.• Utilize database backup methodology.• Describe the Storage Manager data protection architecture.• Automate database backup operations.• Describe how to use each of the data protection products.• Describe Space Management on AIX and HSM on Windows.• Understand basic storage area network concepts.• Describe Storage Manager for System Backup and Recovery.

• • • ••

I

Tivoli Professional Certification

“As an individual, becoming a Tivoli Certified Professional has helped me

to progress further in my career and gain job satisfaction. To the company,

it has helped build up the confidence and trust to the customer. At the end of

the day, you win, the company wins, and the customer wins.”

Special Offer For Having Taken This Course

For having completed this course, you are entitled to a 15% discount on your next exam at any Thomson Prometric or Pearson VUE testing center. Use this special promotion code when registering online or by phone to get the discount: 15CSWR. (This offer may be withdrawn. Please check with the testing center as described later in this section.)

Why Get Certified?

• It is a demonstration of value to your customer through increased overall performance with shorter time cycles to deliver applications.

• Technical certifications assist technical professionals to obtain more visibility to potential customers.

• In tough economic times it is good to be able to make yourself stand out from your competition.

What Is Role-based Certification?

All IBM certifications are based on job roles. They focus on a job a person must do with a product, not just the product’s features and functions. The two primary job roles used by Tivoli Professional Certification are the IBM Certified Deployment Professional and the IBM Certified Advanced Deployment Professional.

II IBM Tivoli Storage Manager 5.4 Implementation and Administration Part Two ©Copyright IBM Corp. 2007

• • • ••

Tivoli Professional Certification


When Should I Attempt a Certification Exam?

After completing this course, you should review the test objectives and sample test on the IBM Certification Web site (URL follows) that correspond with the exam you wish to take. A general guideline is to play a major role in at least two product deployments, getting as much hands-on experience working with the product as possible before you attempt the exam. Exam questions are written by experts in the deployment of this product and most of the material covered is taken from their experience working with the product in real world commercial environments. You will find supplemental readings and other study material on the Web site as well.

Where Can I Take a Certification Exam and How Much Do They Cost?

All IBM certification exams are available at Thomson Prometric and Pearson VUE testing centers worldwide. In addition, testing is offered at a discount or for free at many IBM conferences and events, including many Tivoli User Group (TUG) meetings. Exam costs vary from country to country. Please see the Thomson Prometric or Pearson VUE Web sites for the exact cost at the testing center you plan to use.

Where Can I Get More Information?

• The Tivoli Certification Programhttp://www.ibm.com/certify/ or contact us at [email protected]

• Tivoli Certification Study Guideshttp://www.redbooks.ibm.com (search for “certification study guide”)

• Tivoli Support Technical Exchange Callshttp://www.ibm.com/software/sysmgmt/products/support/supp_tech_exch.html

• Tivoli Certification Study GroupsSend an e-mail inquiry to [email protected]

• Thomson Prometric testing centershttp://www.prometric.com

• Pearson VUE testing centershttp://www.vue.com/ibm/

Ibm tivoli storage manager v5.4 implementation and administration part two tm526inst

Technology

trademarks of ibm corporation

permission of ibm corporation

tivoli products

ibm products

planet tivoli

tivoli enterprise

ibm tivoli storage manager

tivoli ready