Advisor

Performance Advisor Administration GuideFor Use with DataFabric® Manager Server 4.0

NetApp, Inc.495 East Java DriveSunnyvale, CA 94089 U.S.A.Telephone: +1 (408) 822-6000Fax: +1 (408) 822-4501Support telephone: +1 (888) 4-NETAPPDocumentation comments: [email protected] Web: http://www.netapp.comPart number: 210-04801_A0February 2010

Contents

Copyright information ................................................................................. 9Trademark information ............................................................................. 11About this guide .......................................................................................... 13

Audience .................................................................................................................... 13

Terminology .............................................................................................................. 13

Keyboard and formatting conventions ...................................................................... 14

Special messages ....................................................................................................... 15

How to send your comments ..................................................................................... 15

What is new in this release ......................................................................... 17Overview of new and changed features ..................................................................... 17

User interface changes ............................................................................................... 18

New and changed CLI commands ............................................................................. 20

What Performance Advisor is ................................................................... 23Topology of Performance Advisor ............................................................................ 23

Components of the Performance Advisor topology ...................................... 23

Limitations of Performance Advisor ............................................................. 24

Authentication of a Performance Advisor instance ....................................... 24

Performance Advisor support for hierarchical groups ............................................... 24

Performance Advisor support for secure connections ............................................... 25

Administrator roles in Performance Advisor ............................................................. 25

Performance Advisor support for vFiler units ........................................................... 25

About NetApp Management Console ....................................................... 27What NetApp Management Console is ..................................................................... 27

Applications that run in NetApp Management Console ............................................ 27

Licenses ..................................................................................................................... 28

Window layout and navigation .................................................................................. 29

Window customization .............................................................................................. 30

Data filtering .............................................................................................................. 31

Performance objects and counters ............................................................ 33What performance objects are ................................................................................... 33

List of performance objects ....................................................................................... 33

What a performance counter is .................................................................................. 34

Table of Contents | 3

List of performance counters ..................................................................................... 35

What access permission is .......................................................................... 45RBAC for configuration of views ............................................................................. 45

RBAC for changing storage system settings ............................................................. 46

RBAC for thresholds ................................................................................................. 46

RBAC for threshold templates .................................................................................. 47

RBAC for configuration of events ............................................................................. 48

RBAC for configuration of alarms ............................................................................ 48

Performance Advisor and server setup .................................................... 51Enabling the performance-monitoring server ............................................................ 51

Secure connections for Performance Advisor ........................................................... 52

Setting up transport between the performance-monitoring server and

Performance Advisor ............................................................................... 52

Setting up transport protocols between Performance Advisor and

monitored devices .................................................................................... 53

Setting up appliance-level transport between Performance Advisor and

monitored devices .................................................................................... 53

List of options for appliance-level setting ..................................................... 54

Introduction to the Performance Advisor user interface ........................ 55Starting Performance Advisor ................................................................................... 55

Stopping Performance Advisor ................................................................................. 56

Unavailable storage systems or vFiler units in Performance Advisor ....................... 56

Unavailable menu options in the Performance Advisor interface ............................. 56

Zero-value data in a Performance Advisor chart ....................................................... 56

Absence of data in a Performance Advisor chart ...................................................... 57

Display of new storage systems or vFiler units in Performance Advisor ................. 57

Overview of the NetApp Management Console dashboards .................. 59Dashboard panel descriptions .................................................................................... 59

What events are ......................................................................................................... 60

How to know when an event occurs .......................................................................... 60

Monitoring events .......................................................................................... 61

Responding to and acknowledging events ..................................................... 61

Deleting events .............................................................................................. 62

Views in Performance Advisor .................................................................. 63What the performance view window is ..................................................................... 63

What default views are .................................................................................. 64

4 | Performance Advisor Administration Guide For Use with DataFabric Manager Server 4.0

What canned views are .................................................................................. 64

What custom views are .................................................................................. 64

What advanced views are .............................................................................. 65

What real-time data is .................................................................................... 66

What historical data is ................................................................................... 66

Hierarchy of objects ....................................................................................... 66

Analyzing historical data ............................................................................... 66

What synchronizing time axes is ................................................................... 67

Viewing performance data ............................................................................. 67

Displaying tabbed views ................................................................................ 67

Selecting a performance view ........................................................................ 67

Closing a performance view .......................................................................... 68

Data in performance views ............................................................................ 68

Changing the time period in a chart ............................................................... 69

Resetting both axes ........................................................................................ 69

Synchronizing time axes ................................................................................ 69

Undoing synchronization of time axes .......................................................... 70

How to use the legend to interpret performance data .................................... 70

How to view thresholds ................................................................................. 70

Creating advanced views ............................................................................... 70

Adding thresholds using a view ..................................................................... 71

Changing the chart type ................................................................................. 71

Playing or pausing data collection ................................................................. 71

Printing a chart from a view .......................................................................... 72

Enabling charts to scale the y axis ................................................................. 72

Diagnosis of performance issues ................................................................... 72

Diagnosing a performance issue .................................................................... 73

Performance-monitoring with custom views ............................................................. 73

When you must create a custom view ........................................................... 74

How flexible the custom views are ................................................................ 74

Counter data across views ............................................................................. 74

Counters with different units in a single chart ............................................... 74

Viewing real-time data .................................................................................. 75

Creating a custom view ................................................................................. 75

Opening a custom view ................................................................................. 76

Editing a custom view ................................................................................... 76


Deleting a custom view ................................................................................. 76

CPU utilization issues on storage systems or vFiler hosts ............................ 77

Analysis of real-time data .......................................................................................... 77

Methods for preserving performance data ..................................................... 77

Saving a chart ................................................................................................ 78

Exporting a view to a .csv file ....................................................................... 78

Limitations to exporting performance data ................................................... 79

Canned views at various levels .................................................................................. 79

Canned views at the global or group level .................................................... 80

Canned views at the storage system level ..................................................... 82

Canned views at the vFiler level .................................................................... 87

Canned views at the processor level .............................................................. 88

Canned views at the network interface level ................................................. 88

Canned views at the dataset level .................................................................. 89

Canned views at the resource pool level ....................................................... 89

Canned views at the volume level ................................................................. 90

Canned views at the qtree level ..................................................................... 92

Canned views at the LUN level ..................................................................... 93

Canned views at the aggregate level .............................................................. 94

Canned views at the disk level ...................................................................... 94

Canned view at the target level ...................................................................... 95

Alarms, hosts, and thresholds .................................................................... 97What alarms are ......................................................................................................... 97

List of alarm management tasks .................................................................... 97

Overview of alarm management tasks ....................................................................... 98

Monitoring alarms ......................................................................................... 99

Editing alarms ................................................................................................ 99

Testing alarms .............................................................................................. 100

Enabling or disabling alarms ....................................................................... 100

Adding alarms .............................................................................................. 101

Deleting alarms ............................................................................................ 102

Responding to alarms .................................................................................. 102

How hosts become visible to Performance Advisor ................................................ 103

List of fields in the hosts table ..................................................................... 103

Editing host credentials ............................................................................... 104

Configuration of data collection .................................................................. 104


Copying the data collection configuration ................................................... 105

Effect of increasing sample rates ................................................................. 105

Setting retention periods and sample rates .................................................. 105

What thresholds are ................................................................................................. 106

What resource properties are ....................................................................... 107

Viewing threshold values ............................................................................ 107

When to use thresholds in combination ....................................................... 107

What the Details tab is ................................................................................. 107

Adding thresholds ........................................................................................ 108

Editing thresholds using the Summary tab .................................................. 108

Editing thresholds using charts .................................................................... 109

Editing thresholds using the Details tab ...................................................... 109

Deleting thresholds ...................................................................................... 110

Creating alarms ............................................................................................ 110

What a threshold template is ....................................................................... 110

Glossary ..................................................................................................... 113Index ........................................................................................................... 117


Copyright information

Copyright © 1994–2010 NetApp, Inc. All rights reserved. Printed in the U.S.A.

No part of this document covered by copyright may be reproduced in any form or by any means—graphic, electronic, or mechanical, including photocopying, recording, taping, or storage in anelectronic retrieval system—without prior written permission of the copyright owner.

Software derived from copyrighted NetApp material is subject to the following license anddisclaimer:

THIS SOFTWARE IS PROVIDED BY NETAPP "AS IS" AND WITHOUT ANY EXPRESS ORIMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIEDWARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE,WHICH ARE HEREBY DISCLAIMED. IN NO EVENT SHALL NETAPP BE LIABLE FOR ANYDIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIALDAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTEGOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESSINTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHERIN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OROTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IFADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

NetApp reserves the right to change any products described herein at any time, and without notice.NetApp assumes no responsibility or liability arising from the use of products described herein,except as expressly agreed to in writing by NetApp. The use or purchase of this product does notconvey a license under any patent rights, trademark rights, or any other intellectual property rights ofNetApp.

The product described in this manual may be protected by one or more U.S.A. patents, foreignpatents, or pending applications.

RESTRICTED RIGHTS LEGEND: Use, duplication, or disclosure by the government is subject torestrictions as set forth in subparagraph (c)(1)(ii) of the Rights in Technical Data and ComputerSoftware clause at DFARS 252.277-7103 (October 1988) and FAR 52-227-19 (June 1987).

Copyright information | 9

Trademark information

NetApp; the NetApp logo; the Network Appliance logo; Cryptainer; Cryptoshred; DataFabric; DataONTAP; Decru; Decru DataFort; FAServer; FilerView; FlexCache; FlexClone; FlexShare; FlexVol;FPolicy; gFiler; Go further, faster; Manage ONTAP; MultiStore; NearStore; NetCache; NOW(NetApp on the Web); ONTAPI; RAID-DP; SANscreen; SecureShare; Simulate ONTAP;SnapCopy; SnapDrive; SnapLock; SnapManager; SnapMirror; SnapMover; SnapRestore;SnapValidator; SnapVault; Spinnaker Networks; Spinnaker Networks logo; SpinAccess;SpinCluster; SpinFlex; SpinFS; SpinHA; SpinMove; SpinServer; SpinStor; StoreVault; SyncMirror;Topio; vFiler; VFM; and WAFL are registered trademarks of NetApp, Inc. in the U.S.A. and/or othercountries. Network Appliance, Snapshot, and The evolution of storage are trademarks of NetApp,Inc. in the U.S.A. and/or other countries and registered trademarks in some other countries. TheStoreVault logo, ApplianceWatch, ApplianceWatch PRO, ASUP, AutoSupport, ComplianceClock,DataFort, Data Motion, FlexScale, FlexSuite, Lifetime Key Management, LockVault, NOW,MetroCluster, OpenKey, ReplicatorX, SecureAdmin, Shadow Tape, SnapDirector, SnapFilter,SnapMigrator, SnapSuite, Tech OnTap, Virtual File Manager, VPolicy, and Web Filer aretrademarks of NetApp, Inc. in the U.S.A. and other countries. Get Successful and Select are servicemarks of NetApp, Inc. in the U.S.A.

IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International BusinessMachines Corporation in the United States, other countries, or both. A complete and current list ofother IBM trademarks is available on the Web at http://www.ibm.com/legal/copytrade.shtml.

Apple is a registered trademark and QuickTime is a trademark of Apple, Inc. in the U.S.A. and/orother countries. Microsoft is a registered trademark and Windows Media is a trademark of MicrosoftCorporation in the U.S.A. and/or other countries. RealAudio, RealNetworks, RealPlayer,RealSystem, RealText, and RealVideo are registered trademarks and RealMedia, RealProxy, andSureStream are trademarks of RealNetworks, Inc. in the U.S.A. and/or other countries.

All other brands or products are trademarks or registered trademarks of their respective holders andshould be treated as such.

NetApp, Inc. is a licensee of the CompactFlash and CF Logo trademarks.

NetApp, Inc. NetCache is certified RealSystem compatible.

Trademark information | 11

About this guide

You can use your product more effectively when you understand this document's intended audienceand the conventions that this document uses to present information.

This document describes the Performance Advisor software product and how to use it to monitor,administer, and optimize NetApp storage systems that run the Data ONTAP operating system. Theinformation in this guide applies to all supported storage system models.

This guide does not cover basic system or network administration topics, such as IP addressing andnetwork management.

Next topics

Audience on page 13

Terminology on page 13

Keyboard and formatting conventions on page 14

Special messages on page 15

How to send your comments on page 15

AudienceThis document is written with certain assumptions about your technical knowledge and experience.

This document is for system administrators and others interested in managing and monitoring storagesystems with DataFabric Manager and Performance Advisor.

This document is written with the assumption that you are familiar with the following technology:

• Data ONTAP operating system software• The protocols you use for file sharing or transfers, such as NFS, CIFS, iSCSI, FCP, or HTTP• The client-side operating systems (Linux or Windows)

TerminologyTo understand the concepts in this document, you might need to know how certain terms are used.

Storage terms

array LUN Refers to storage that third-party storage arrays provide to storage systemsrunning Data ONTAP software. One array LUN is the equivalent of one disk ona native disk shelf.

About this guide | 13

LUN (logicalunit number)

Refers to a logical unit of storage identified by a number.

storagecontroller

Refers to the component of a storage system that runs the Data ONTAPoperating system and controls its disk subsystem. Storage controllers are alsosometimes called controllers, storage appliances, appliances, storage engines,heads, CPU modules, or controller modules.

storage system Refers to the hardware device running Data ONTAP that receives data from andsends data to native disk shelves, third-party storage, or both. Storage systemsthat run Data ONTAP are sometimes referred to as filers, appliances, storageappliances, V-Series systems, or systems.

third-partystorage

Refers to the back-end storage arrays, such as IBM, Hitachi Data Systems, andHP, that provide storage for storage systems running Data ONTAP.

Keyboard and formatting conventionsYou can use your product more effectively when you understand how this document uses keyboardand formatting conventions to present information.

Keyboard conventions

Convention What it means

The NOW site Refers to NetApp On the Web at http://now.netapp.com/.

Enter, enter • Used to refer to the key that generates a carriage return; the key is namedReturn on some keyboards.

• Used to mean pressing one or more keys on the keyboard and then pressing theEnter key, or clicking in a field in a graphical interface and then typinginformation into the field.

hyphen (-) Used to separate individual keys. For example, Ctrl-D means holding down theCtrl key while pressing the D key.

type Used to mean pressing one or more keys on the keyboard.


http://now.netapp.com/

Formatting conventions

Convention What it means

Italic font • Words or characters that require special attention.• Placeholders for information that you must supply.

For example, if the guide says to enter the arp -d hostname command,you enter the characters "arp -d" followed by the actual name of the host.

• Book titles in cross-references.

Monospaced font • Command names, option names, keywords, and daemon names.• Information displayed on the system console or other computer monitors.• Contents of files.• File, path, and directory names.

Bold monospaced

fontWords or characters you type. What you type is always shown in lowercaseletters, unless your program is case-sensitive and uppercase letters arenecessary for it to work properly.

Special messagesThis document might contain the following types of messages to alert you to conditions that youneed to be aware of.

Note: A note contains important information that helps you install or operate the systemefficiently.

Attention: An attention notice contains instructions that you must follow to avoid a system crash,loss of data, or damage to the equipment.

How to send your commentsYou can help us to improve the quality of our documentation by sending us your feedback.

Your feedback is important in helping us to provide the most accurate and high-quality information.If you have suggestions for improving this document, send us your comments by e-mail to [email protected]. To help us direct your comments to the correct division, include in thesubject line the name of your product and the applicable operating system. For example, FAS6070—Data ONTAP 7.3, or Host Utilities—Solaris, or Operations Manager 3.8—Windows.

About this guide | 15

mailto:[email protected]

What is new in this release

The "What is new in this release" section describes new features and changes in PerformanceAdvisor for DataFabric Manager server 4.0. Detailed information about the features is providedelsewhere in this guide.

Next topics

Overview of new and changed features on page 17

User interface changes on page 18

New and changed CLI commands on page 20

Overview of new and changed featuresPerformance Advisor in DataFabric Manager server 4.0 contains new and changed features.

New features

Data collectionconfiguration

You can view the counters for which data collection is enabled or disabled on aper storage system basis. The following are the other advantages that areavailable with this feature:

• Disabling or enabling data collection for selected counters per storagesystem

• Configuring sample rate and retention period for selected counter groups perstorage system

• Copying the data collection configuration to other storage systems

Advanced views You can use advanced views to establish a baseline for the data collected byapplying statistical parameters such as minimum, maximum, mean, andpercentile. Time filters enable you to retrieve the baseline data based on selectedbusiness hours and week days. You can use the baseline data in advanced viewsto set appropriate thresholds.

Data export This feature helps you to export the Performance Advisor data collected on theDataFabric Manager server. The data is collected in CMA format. However, youshould use this feature only when NetApp requests the data from you.

Performancediagnosis

You can diagnose the storage system for any performance-related issue during agiven period and get recommendations to resolve the issue. The eventcorrelation chart allows you to plot events of related objects. Also, the chartplots the Operations Manager events, which can help you in diagnosing thecause of the performance degradation.

What is new in this release | 17

Client statistics You can collect the statistics related to CIFS and NFS client operations to findthe cause of latency. Also, you can use this feature to automatically collect per-client statistics when a set of thresholds on a storage system is breached.

Changes

Flexiblecustomviews

You can apply custom views on either a particular object or every instance of theselected object. The following are the other advantages that are available with thisfeature:

• Creating copies of canned views and modifying them• Displaying data for historical periods in custom views• Comparing the data collected for a counter on multiple instances of an object• Configuring bar chart preferences such as the number of objects and the kind of

chart

User interface changesPerformance Advisor introduces the Add Threshold Template wizard, threshold on graphs, thresholddetails, set default view, event and bar charts in custom view, event correlation, and usabilityimprovements.

The following are the other changes in the interface:

Terminologychange

"Sample buffer" is now called "Retention period."

A distinction is made between canned views and default views. Cannedviews are views that are available in Performance Advisor and are notcreated by you.

New window tomonitor clientstatistics

You can use the Monitor Client Statistics window (Monitor > ClientStatistics) to collect statistics related to the CIFS and NFS client operationson storage systems.

Command buttonchanges in Viewwindows

The following user actions are consolidated in the View Actions list:

• Switch the view to real-time or historic• Launch advanced view• Diagnose performance issues on the selected storage system• Synchronize the time axis of all charts• Play or pause the data collection• Collect CIFS and NFS client statistics

Command buttonchanges in charts

The following user actions are consolidated in the Actions list:

• Add counters


• Show events• Reset both axes• Add thresholds• Save the chart• Print the chart

Improvements inthe Set Up Hostswindow

• A new command button, Data Collection, is added to the Set Up Hostswindow to configure data collection on storage systems. Starting fromthis release, data collection is disassociated from views. Hence, you canperform tasks such as setting sample rates and retention periods by usingthis command button. You can also edit and copy data collectionconfiguration using this button.

• The following new columns are added to the hosts list:

• Max Projected Space• Used Space• Percentage of Used Space

• A new tab, Data Collection, is added to the bottom-half of the Set UpHosts window. You can click this tab to display the maximum projectedspace, used space, sample rate, and retention period of a counter groupfor a selected host. Also, the window displays the data collection enabledfor counters belonging to a selected counter group.

Improvements inthe Set Up CustomViews window

• Starting from this release, data collection is not part of custom views.You cannot set sample rates and retention periods using this window.

• The following new command buttons are added:

• Copy• Refresh

• The following new columns are added to the Views list:

• Applies to• View Type

• You can apply custom views to one or more particular objects or for allinstances of objects by selected type.

• The details area displays the counter, the counter group, and the kind ofchart associated with the selected view.


New and changed CLI commandsDataFabric Manager server 4.0 includes new and changed CLI commands to support the new andchanged features. For detailed information about these commands, see the DataFabric Manager manpages.

New CLI commands to support new features

Client statistics • dfm perf clientstat collect

Collects CIFS and NFS client operation statistics for a storage system• dfm perf clientstat list

Lists collections of client operation statistics• dfm perf clientstat purge

Deletes one or more collections of client statistics from the DataFabricManager database

Performancediagnosis

• dfm perf diag troubleshoot

Diagnoses the storage system for performance issues

Data export • dfm perf data export

Exports the counter data for the specified objects in the Counter ManagerArchive (CMA) file format (binary)

Advanced views • dfm perf data retrieve

Retrieves performance data for the specified counters and instances ofobject with statistical parameters

Data collectionconfiguration

• dfm perf data enable

Enables the counter data collection for the selected storage system• dfm perf data disable

Disables the counter data collection for the selected storage system• dfm perf data copy

Copies the data collection configuration from one storage system to one ormore storage systems.

Flexible customviews

• dfm perf view create

Creates a new performance view• dfm perf view chart add

Adds a chart to a view• dfm perf view counter add


Adds a counter to a view• dfm perf view describe

Lists all the counters associated with a view• dfm perf view list

Lists all the views• dfm perf view counter delete

Deletes a counter from a view• dfm perf view chart delete

Deletes a chart from a view• dfm perf view destroy

Destroys a view


What Performance Advisor is

Performance Advisor is a Java-based client application that runs in NetApp Management Console.

Using Performance Advisor, you can view comprehensive performance information about storagesystems and vFiler units. You can use the information to achieve the following objectives:

• Monitor storage systems or vFiler units for usage levels and optimal functioning, by viewinghistorical and real-time data for all monitored devices recognized by the DataFabric Managerserver

• Identify potential blockages in the data infrastructure• Perform short-term trend analysis for the data infrastructure

Next topics

Topology of Performance Advisor on page 23

Performance Advisor support for hierarchical groups on page 24

Performance Advisor support for secure connections on page 25

Administrator roles in Performance Advisor on page 25

Performance Advisor support for vFiler units on page 25

Topology of Performance AdvisorThe topology of Performance Advisor is designed specifically to help you gather performance data.

Next topics

Components of the Performance Advisor topology on page 23

Limitations of Performance Advisor on page 24

Authentication of a Performance Advisor instance on page 24

Components of the Performance Advisor topologyThe Performance Advisor topology includes NetApp Management Console (which provides theinterface to Performance Advisor), DataFabric Manager server (which includes a performance-monitoring server), and one or more storage systems or vFiler units.

These components interact in the following ways:

• The performance-monitoring server collects and stores performance data from one or morestorage systems or vFiler units on the network. The monitoring server is installed and enabled bydefault as part of the DataFabric Manager server installation.

What Performance Advisor is | 23

• Performance Advisor reads the performance data stored on the monitoring server and generatesperformance-related views and charts. You must install Performance Advisor on a Windows orLinux system, separate from the system on which the DataFabric Manager server is installed.

Note: Performance Advisor is available only on the host on which NetApp Management Consoleis installed. You cannot view Performance Advisor remotely through a browser.

• The DataFabric Manager server monitors one or more storage systems or vFiler units.

Limitations of Performance AdvisorTo minimize CPU utilization, Performance Advisor does not collect real-time data for any view thatincludes a bar graph.

Although the performance-monitoring server supports simultaneous connections to multiplePerformance Advisor instances, connection to many instances of Performance Advisorsimultaneously can result in slower collection of data.

Authentication of a Performance Advisor instanceThe NetApp Management Console platform in which Performance Advisor runs must supply validuser name and password credentials to connect with the performance-monitoring server.

The server obtains the user name and the password from the database on the workstation, andvalidates the credentials with the HTTP or HTTPS authentication mechanisms.

Performance Advisor support for hierarchical groupsPerformance Advisor supports hierarchical groups and other groups created on DataFabric Manager3.3 or later.

A group can be one of the following:

• An empty group with no storage systems or vFiler units or subgroups• A group containing storage systems or vFiler units only, with no subgroups• A group containing subgroups only, without direct storage system or vFiler units• A group containing one or more direct storage systems or vFiler units and one or more subgroups• The Global group, whose properties and configurations apply on a default basis to all of its

subgroups

The groups are visible in the View navigation pane. The Global group might contain one or moregroups. You can select a group from the Group drop-down list in the menu bar. All the operationsthat you can perform on the Global group are valid for subgroups too. For example, you can click asubgroup folder to display the Summary View window.


Performance Advisor support for secure connectionsPerformance Advisor supports both HTTP and HTTPS connections between the performance-monitoring server, Performance Advisor, and the monitored storage systems and vFiler hosts.

The DataFabric Manager server generates an SSL certificate for the performance-monitoring serverautomatically at start-up. You do not need to take any action to generate the certificate.

Note: Using HTTPS transport for the connection between the performance-monitoring server andthe storage system or vFiler host might adversely affect storage system performance. In suchcases, use HTTP transport only.

Administrator roles in Performance AdvisorYou can use the administrator roles such as the GlobalRead, GlobalWrite, GlobalDelete, andGlobalPerfManagement roles to control access to performance data in the DataFabric Managerserver.

The administrator roles determine the availability of menu options in Performance Advisor. Greyedout options in the application interface indicate that the options are not available to you for theselected storage system or vFiler unit.

The following roles help you to work with performance views in the interface:

• The GlobalRead role allows you to view all performance views.• The GlobalWrite and GlobalDelete roles allow you to create, edit, and delete custom views within

any group.• The GlobalPerfManagement role allows you to manage views, event thresholds, and alarms, apart

from viewing performance information in the application.You can use this role to work with custom views.

Note: You can execute the performance-monitoring tasks if you have the GlobalRead,GlobalWrite, and GlobalPerfManagement roles.

Performance Advisor support for vFiler unitsPerformance Advisor supports the MultiStore feature in DataFabric Manager 3.3 or later. You canuse the Data ONTAP MultiStore feature to view performance data on vFiler units.

Performance Advisor displays performance data pertaining to a vFiler unit’s resource, such as LUNs,volumes, and qtrees. It also displays some physical performance data that applies to the hostingstorage system for a vFiler unit (for example, CPU usage).

What Performance Advisor is | 25

The vFiler counters pertaining to its hosting storage systems are displayed for hosting storagesystems running Data ONTAP 7.1 or later.

Note: The vFilers, All report at Control Center > Home > Member Details > Virtual Systemsin Operations Manager lists your network’s vFiler units and their hosting storage systems.

See the Operations Manager Administration Guide for information about DataFabric Managersupport for the MultiStore feature.

Related information

Operations Manager Administration Guide - now.netapp.com/NOW/knowledge/docs/DFM_win/dfm_index.shtml


http://now.netapp.com/NOW/knowledge/docs/DFM_win/dfm_index.shtml


About NetApp Management Console

NetApp Management Console is used by administrators to carry out management tasks aided byDataFabric Manager, but it runs on a Windows or Linux system separate from the server on whichDataFabric Manager is installed.

NetApp Management Console allows storage, application, and server administrators to performmanagement tasks such as data backup protection, space management, resource provisioning, datamigration, and performance tuning, without having to switch between separate user interfaces.

The DataFabric Manager server provides infrastructure services (such as discovery, monitoring, role-based access control (RBAC), auditing, and logging for products in the storage and data suites) forNetApp Manageability Software client applications. The DataFabric Manager software runs on aseparate server and is managed itself through Operations Manager, the Web-based user interface ofDataFabric Manager. For more information about DataFabric Manager and Operations Manager, seethe Operations Manager Administration Guide.

Next topics

What NetApp Management Console is on page 27

Applications that run in NetApp Management Console on page 27

NetApp Management Console Licenses on page 28

NetApp Management Console window layout and navigation on page 29

NetApp Management Console window customization on page 30

NetApp Management Console data filtering on page 31

Related information

Operations Manager Administration Guide - http://now.netapp.com/NOW/knowledge/docs/DFM_win/dfm_index.shtml

What NetApp Management Console isNetApp Management Console is the client platform for NetApp Manageability Softwareapplications.

Applications that run in NetApp Management ConsoleThe Performance Advisor and the licensed Protection Manager and Provisioning Managerapplications run in NetApp Management Console.

PerformanceAdvisor

This application provides a single location from which you can viewcomprehensive information about storage system and MultiStore vFiler unit

About NetApp Management Console | 27



performance and perform short-trend analysis. The application also helps youidentify the data infrastructure causes and potential causes of reducedperformance.

Performance Advisor is automatically enabled with the Operations Manager Corelicense.

ProtectionManager

This application provides a policy-based management tool to help you unify andautomate backup and mirroring operations. The application uses a holisticapproach to data protection. It provides end-to-end, workflow-based design andseamless integration of SnapVault, SnapMirror, and Open Systems SnapVault toenable you to manage large-scale deployments easily.

The disaster recovery feature of the protection application enhances your dataprotection services by enabling you to continue to provide data access to yourusers, even in the event of mishap or disaster that disables or destroys the storagesystems in your primary data node. If the disaster recovery license is installed, youcan quickly enable your secondary storage systems to provide primary datastorage access to your users with little or no interruption, until your primarystorage systems are reenabled or replaced.

ProvisioningManager

This application helps you simplify and automate the tasks of provisioning andmanaging storage. The application provides policy-based provisioning andconformance of storage in datasets. The application also enables you to manuallyadd volumes or qtrees to a dataset at any time, provides manual controls for spaceand capacity management of existing storage and newly provisioned storage, andallows you to migrate datasets and vFiler units to a new storage destination.

The deduplication feature of the provisioning application enhances your dataprovisioning services by enabling you to eliminate duplicate data blocks to reducethe amount of storage space used to store active data.

NetApp Management Console LicensesSome NetApp Management Console applications require a license.

You must enter a software license code on the DataFabric Manager server to enable thecorresponding functions. You do not need to indicate which license the code enables. The code ismatched automatically to the appropriate service license.

• Performance functionsPerformance Advisor is automatically enabled with the Operations Manager Core license. Noadditional license is required to enable the performance functions.

• Protection functions


To enable the protection functions, you must install the Protection Manager license on theDataFabric Manager server. The Protection Manager license also enables all the provisioningfunctions provided by Provisioning Manager.

• Disaster recovery functionsTo enable the disaster recovery functions, you must install the Protection Manager license and thedisaster recovery license.

• Provisioning functionsProvisioning functions are automatically enabled if you have the Protection Manager licenseinstalled. If that license is not installed, you must install the Provisioning Manager license on theDataFabric Manager server. This license enables only the provisioning features.

• Deduplication functionsTo enable the deduplication functions, you must install the Provisioning Manager license and thededuplication license.

NetApp Management Console window layout and navigationYou can navigate NetApp Management Console to move within and between applications, to displayHelp, and to log out of the console.

Learning aboutProtectionManager

The Get Started section of the user interface provides an overview ofProtection Manager, a short tutorial that demonstrates how to set upprotection, and a list of frequently asked questions (FAQs).

Learning aboutdisaster recovery

The Get Started section of the user interface provides an overview of disasterrecovery, a short tutorial that demonstrates how to set up disaster recovery,and a list of frequently asked questions (FAQs).

Learning aboutProvisioningManager

The Get Started section of the user interface provides an overview ofProvisioning Manager, a short tutorial that demonstrates how to set upprovisioning, and a list of frequently asked questions (FAQs).

Moving back andforth betweenpanes

The Back and Forward arrow buttons toggle between your last and previouspanes, even if the pane was in a different application.

Toggling betweenapplications

• To use Protection Manager or Provisioning Manager, select Tasks >

Manage Data or go to the Tasks Bar and click .• To use Performance Advisor, select Tasks > Manage Performance or

go to the Tasks Bar and click .

Displaying help • To display the Help for all applications on NetApp Management Console,click Help > Help Contents.


• To display Help for the specific window that is currently displayed, click

Help > Help For This View or click .

Logging out • To log out, select File > Log Out or click the logout button.• To log out and close NetApp Management Console, select File > Exit.

Viewing productlicense information

To view information about the product licenses you have installed, clickHelp > About > Licenses. Removed or added licenses are reflected in thelicense list after you restart NetApp Management Console.

Applications that run in NetApp Management Console vary in the specifics of their window layout.However, the windows are generally divided into two panes:

• Use the navigation pane on the left to move within an application.• Use the content pane on the right to view and manage data. The content pane is generally divided

into two areas: lists of selectable data and details of the selected data. Click , when it isdisplayed, for additional details.

Note: Specific applications might have additional navigation features not described in this section.In addition, software version incompatibility may result in some reduced functionality, which maycause some menu options or buttons to be disabled.

NetApp Management Console window customizationNetApp Management Console includes features you can use to customize the application windowlayout. By customizing application windows, you can control which data is viewable or how it isdisplayed.

Note: Not all customization features are available for every application window.

Hiding orredisplaying theTasks Bar

You can hide the Tasks Bar to increase the space available for the contentpane. (The overall width of the window remains the same when the bar ishidden.) You use the View menu to hide or display the bar.

Selecting columnsto display

In content panes that include a list divided into columns, you can choosewhich columns you want to display. To display or hide a column, click the

column-selection icon in the upper-right corner above the scroll bar ( ),then click the name of a column heading from the selection list. Columnsalready displayed are identified with a check mark.

Note: The default version of a list does not necessarily include allavailable columns. You might find additional columns available in thecolumn selection list.


Readjustingcolumn widths

In content panes that include a list divided into columns, you can adjust thewidth of individual columns by horizontally dragging the vertical linebetween column headings. If accommodating all the columns you want toinclude results in column widths that obscure data, you can still see theentire value of a column field by putting your cursor over the field. A tool tippops up, displaying the entire value of the field. You can maximize data

visibility across all columns by clicking the column selection icon ( ) andchoosing Pack All Columns.

Rearrangingcolumn order

In content panes that include a list divided into columns, you can rearrangethe column order for a session by dragging and dropping column headingshorizontally within the table. The arrangement persists until you exit theconsole.

Reapportioning acontent pane

A splitter bar divides a content pane list area from the details area. You canmove the splitter bar to increase or decrease the proportion of the contentpane that displays the list of selectable data.

NetApp Management Console data filteringYou can use the data filtering features to display only data meeting the criteria you specify.

In large-scale environments, the content pane might list so much data that it becomes difficult tolocate the information you want. Filtering the data by specified criteria helps you focus on the issuesmost important to you.

Note: Not all filtering features are available for every application window.

Filtering bygroup

The Group selection list in the toolbar enables you to display only the data thatpertains to objects in a selected group. This setting persists until you log out orchoose a different group.

Filtering byregularexpression

You can filter columns displaying site-specific values, such as storage systemnames or dataset names, by regular expression. To filter a column by regularexpression, click the filter icon in the column heading ( ) and specify theregular expression to match against values in the column field. Headings offiltered columns are highlighted to remind you that some data is not currentlydisplayed.

NetApp Management Console uses Java regular expression syntax for filteringdata. For example:

• To view only items beginning with the letters "sch," type sch in the filterfield, which matches Schedule but not SnapMirror.


• To view only items containing "space" somewhere in their string,type .*space in the filter field, which matches strings such as Volume SpaceNormal.

• To view only items ending with the string "ok," type *ok in the filter field,which matches strings such as SnapMirror: Date Ok.

Filtering bycolumn values

When a column displays a set of predefined possible values, you can choose todisplay only the rows of data that include a selected value. To filter a column by aspecific, predefined value, click the filter icon in the column heading ( ) andselect one or more of the predefined values from the drop-down list. Headings offiltered columns are highlighted to remind you that some data is not currentlydisplayed. Click "All" to select or clear all the values.

Filtering bycolumn view

You can click in the upper-right corner of the list to select which columns youwant displayed.

Sorting bycolumn values

You can click on the column header to change the sort order of the columnentries. When you click the column header, the sort arrow appears for thatcolumn.


Performance objects and counters

You can monitor the performance of storage systems by understanding performance objects andcounters in Performance Advisor.

Next topics

What performance objects are on page 33

List of performance objects on page 33

What a performance counter is on page 34

List of performance counters on page 35

What performance objects arePerformance objects are collections of counters for specific subsystem types.

The names of the objects correspond to the subsystem names that are displayed in the navigationpane of Performance Advisor.

List of performance objectsYou can identify the performance objects by using the list of the performance objects most likely tobe monitored by an administrator.

Except for the qtree object, the application displays a default performance view for everyperformance object listed here.

Note: In the following table, performance objects that apply to both storage systems and vFilerunits are noted in parentheses. All other listed performance objects apply to storage systems only.

Object name Type of data contained

aggregate Aggregate statistics (includes aggregates containing traditionalvolumes)

cifs CIFS statistics for systems with a CIFS license

disk Disk statistics (per disk)

fcp FCP statistics for systems with an FCP license

ifnet Statistics on various network interface cards in the system

Note: It is possible that a system can have all target cards and nointerface cards.

Performance objects and counters | 33

Object name Type of data contained

iscsi iSCSI statistics for systems with an iSCSI license

LUN (applies to vFiler units also) LUN statistics (per LUN)

nfsv3 NFS statistics for systems with an NFS license

nfsv4 NFS statistics for systems with an NFS license

processor Processor statistics (per processor)

qtree (applies to vFiler units also) qtree statistics (per qtree)

Note: To minimize negative performance impact on storagesystems that it monitors, the application does not automaticallygenerate default performance views for its qtree performanceobjects; however, you can use the application to defineperformance views for individual qtrees.

system Summary of general performance health of the system

target Statistics on various target cards in the system

vFiler (applies to vFiler units and theirhosting storage system)

Statistics that apply to vFiler units and their hosting storage system

volume (applies to vFiler units also) Volume statistics (per volume)

What a performance counter isA performance counter is a statistical measurement of activity on a storage system or storagesubsystem.

The DataFabric Manager performance-monitoring server collects the data from the monitored storagesystems or vFiler units. Performance Advisor uses this data to generate a performance chart or part ofa performance chart.

For example, the application might use the volume's total_ops counter of a storage system to generatea chart that displays total operations-per-second on that volume on a minute-to-minute basis.

The server collect the data for a counter from the time the counter is added to a view. When the viewis deleted, data collection for the counter stops. When you add a counter to a view, data collectionbegins with the counter in that view. The new view does not display the data collected for the counterin the other views.

Note: Each type of storage system, vFiler unit, or subsystem has a set of counters. However, aparticular storage system or a vFiler unit might not have non-zero data for all the countersavailable to it.


For example, every storage system has an http_ops counter. However, if the storage system doesnot have HTTP enabled, the counter contains only zero-value data. If SnapMirror is not licensedon the storage system, the counters for the SnapMirror statistics contain no data.

List of performance countersYou can identify counters by using the list of performance counters that collect data for eachperformance object.

The following are the descriptions of counters for objects. Availability of counters and theirdescriptions might vary with the version of Data ONTAP.

Aggregate object

total_transfers Total number of disk operations serviced by the aggregate per second

user_reads Number of user read operations serviced by the aggregate per second

user_writes Number of user write operations serviced by the aggregate per second

cp_reads Number of read operations to the aggregate per second during consistencypoint processing

user_read_blocks Number of blocks transferred for user read operations per second

user_write_blocks Number of blocks transferred for user write operations per second

cp_read_blocks Number of blocks transferred for consistency-point read operations per second

cifs object

cifs_ops Total number of CIFS operations

cifs_op_count Array of select CIFS operation counts

cifs_ops_pct Array of select CIFS operation counts as a percentage of total CIFS operations

cifs_latency Average latency for CIFS operations in milliseconds

cifs_latency_base Total observed CIFS operations to be used as a base counter for CIFS averagelatency calculation

cifs_read_ops Total number of CIFS read operations per second

cifs_write_ops Total number of CIFS write operations per second

disk object

total_transfers Total number of disk operations serviced by the aggregate per second


user_reads Number of user read operations serviced by the aggregate per second

user_writes Number of user write operations serviced by the aggregate per second

cp_reads Number of read operations to the aggregate per second duringconsistency point processing

guaranteed_reads Number of disk read operations initiated each second for RAIDreconstruct or scrubbing activities

Note: This counter is available on storage systems running DataONTAP versions earlier than 7.1.

guarenteed_reads Number of disk read operations initiated each second for RAIDreconstruct or scrubbing activities

Note: This counter is available on storage systems running DataONTAP 7.1 or later.

guaranteed_writes Number of write operations initiated each second for RAIDreconstruct or scrubbing activities

Note: This counter is available on storage systems running DataONTAP versions earlier than 7.1.

guarenteed_writes Number of write operations initiated each second for RAIDreconstruct or scrubbing activities

Note: This counter is available on storage systems running DataONTAP 7.1 or later.

user_read_chain Average number of blocks transferred in each user read operation

user_write_chain Average number of blocks transferred in each user write operation

cp_read_chain Average number of blocks transferred in each consistency-point readoperation during a consistency point

guaranteed_read_chain Average number of blocks transferred in each guaranteed readoperation

guaranteed_write_chain Average number of blocks transferred in each guaranteed writeoperation

user_read_blocks Number of blocks transferred for user read operations per second

user_write_blocks Number of blocks transferred for user write operations per second

cp_read_blocks Number of blocks transferred for consistency-point read operationsper second

guaranteed_read_blocks Number of blocks transferred for guaranteed read operations persecond


guaranteed_write_blocks Number of blocks transferred for guaranteed write operations persecond

user_read_latency Average latency per block in microseconds for user read operations

user_write_latency Average latency per block in microseconds for user write operations

cp_read_latency Average latency per block in microseconds for consistency-point readoperations

guaranteed_read_latency Average latency per block in microseconds for guaranteed readoperations

guaranteed_write_latency Average latency per block in microseconds for guaranteed writeoperations

base_for_disk_busy Time base for disk_busy calculation

disk_busy Percentage of time there was at least one outstanding request to thedisk

skip_blocks Number of blocks skipped in skip-mask write operations per second

user_skip_write_ios Number of disk skip-write operations initiated each second for storingdata or metadata associated with user requests

user_writes_in_skip_mask Number of disk write IOs that were executed as part of a skip-maskwrite

fcp object

fcp_ops FCP operations per second

fcp_write_data FCP bytes written per second

fcp_read_data FCP bytes read per second

fcp_read_latency Average latency of read operations observed over all LUNs in the systemaccessed over FCP in milliseconds

fcp_read_ops Total number of read operations observed over all LUNs in the systemaccessed over FCP

fcp_write_latency Average latency of write operations observed over all LUNs in the systemaccessed over FCP in milliseconds

fcp_write_ops Total number of write operations observed over all LUNs in the systemaccessed over FCP

fcp_latency Average latency for FCP operations in milliseconds


ifnet object

recv_packets Packets received per second

recv_errors Errors per second while receiving packets

send_packets Packets sent per second

send_errors Errors per second while sending packets

collisions Collisions per second on CSMA interfaces

recv_data Total bytes received per second

send_data Total bytes sent per second

recv_mcasts Packets received per second via multicast

send_mcasts Packets sent per second via multicast

recv_drop_packets Receive packets dropped per second

iSCSI object

iscsi_ops iSCSI operations per second

iscsi_write_data iSCSI bytes written per second

iscsi_read_data iSCSI bytes read per second

iscsi_read_latency Average latency of read operations observed over all LUNs in the systemaccessed over iSCSI in milliseconds

iscsi_read_ops Total number of read operations observed over all LUNs in the systemaccessed over iSCSI

iscsi_write_latency Average latency of write operations observed over all LUNs in the systemaccessed over iSCSI in milliseconds

iscsi_write_ops Total number of write operations observed over all LUNs in the systemaccessed over iSCSI

iscsi_latency Average latency for ISCSI operations in milliseconds

LUN object

read_ops Read operations per second

write_ops Write operations per second

other_ops Other operations per second

read_data Read bytes per second

write_data Write bytes per second


queue_full Queue full responses per second

total_ops Total number of operations on the LUN per second

avg_latency Average latency in milliseconds for all operations on the LUN

queue_depth_lun queue depth for the LUN

avg_read_latency Average read latency in milliseconds for all operations on the LUN

avg_write_latency Average write latency in milliseconds for all operations on the LUN

avg_other_latency Average other ops latency in milliseconds for all operations on the LUN

nfsv3 object

nfsv3_ops Total number of NFSv3 operations

nfsv3_op_count Array of select NFSv3 operation counts

nfsv3_op_latency_base Array of select NFSv3 operation counts for latency calculation

nfsv3_op_percent Array of select NFSv3 operations as a percentage of total NFSv3operations

nfsv3_op_latency Array of latencies of select NFSv3 operations

nfsv3_avg_op_latency Average latency of the NFSv3 operations

nfsv3_read_size_histo Histogram of NFSv3 read sizes

nfsv3_write_size_histo Histogram of NFSv3 write sizes

nfsv3_read_latency Average latency for NFSv3 read operations in milliseconds

nfsv3_read_ops Total observed NFSv3 read operations

nfsv3_write_latency Average latency for NFSv3 write operations in milliseconds

nfsv3_write_ops Total observed NFSv3 write operations

nfsv4 object

nfsv4_ops Number of total NFSv4 operations per second

nfsv4_op_latency Array of latencies of select NFSv4 operations

nfsv4_read_ops Number of NFSv4 read operations per second

nfsv4_avg_latency Average latency of an NFSv4 request

nfsv4_write_ops Number of NFSv4 write operations per second


Processor object

sk_switches Number of sk switches

hard_switches Number of context switches

domain_busy Array of processor time in percentage spent in various domains

processor_busy Percentage of elapsed time that the processor is executing non-idleprocesses

processor_elapsed_time Wall-clock time since boot, used for calculating processor utilization

qtree object

Note: To minimize negative performance impact on storage systems that it monitors, PerformanceAdvisor does not automatically generate default performance views for its qtree performanceobjects; however, you can use the following counters to define performance views for individualqtrees.

nfs_ops NFS operations per second

cifs_ops CIFS operations per second

System object

nfsv3_ops NFS operations per second

nfsv4_ops NFS operations per second


http_ops HTTP operations per second

dafs_ops DAFS operations per second

fcp_ops FCP operations per second

iscsi_ops iSCSI operations per second

net_data_recv Network KB received per second

net_data_sent Network KB sent per second

disk_data_read Disk KB read per second

disk_data_written Disk KB written per second

cpu_busy Percentage of time one or more processors is busy in the system

Note: For storage systems running Data ONTAP 7.2 or earlier, thecpu_busy counter is the amount of time that any one CPU is busy. Thisresults in a value for cpu_busy that is inflated. For storage systems


running Data ONTAP 7.2.1 or later, the cpu_busy counter is the greaterof either average CPU utilization or the busiest domain.

avg_processor_busy Average processor utilization across all processors in the system

total_processor_busy Total processor utilization of all processors in the system

num_processors Number of active processors in the system

cpu_elapsed_time Elapsed time since boot

nfs_ops NFS operations per second



time Time in seconds since the Epoch (00:00:00 UTC, January 1, 1970)

uptime Time in seconds that the system has been up

Target object




other_ops Other operations per second

read_data Read bytes per second

write_data Write bytes per second

queue_full SCSI queue full responses per second

protocol_errors FCP protocol errors per second

invalid_xmit_words Invalid words transmitted per second

crc_errors CRC errors per second

lr_sent LRs sent per second

lr_received LRs received per second

discarded_frames Number of frames discarded per second

nos_received NOSs received per second

ols_received OLSs received per second

queue_depth_hba queue depth for the HBA

avg_latency_hba Average latency in milliseconds for all operations on the HBA


total_ops_hba Total number of operations on the HBA per second

avg_read_latency_hba Average read latency in milliseconds for all operations on the HBA

avg_write_latency_hba Average write latency in milliseconds for all operations on the HBA

avg_other_latency_hba Average other ops latency in milliseconds for all operations on the HBA

vFiler object

vfiler_cpu_busy Percentage of CPU capacity utilized per second by the vFiler unit’s hostingstorage system

vfiler_net_data_recv Network data received in kilobytes per second

vfiler_net_data_sent Network data sent in kilobytes per second

vfiler_read_ops Read operations per second on the vFiler unit

vfiler_write_ops Write operations per second on the vFiler unit

vfiler_misc_ops Miscellaneous operations per second on the vFiler unit

vfiler_read_bytes Data read in kilobytes per second on the vFiler unit

vfiler_write_bytes Data written in kilobytes per second on the vFiler unit

Volume object

total_ops Number of operations per second serviced by the volume

avg_latency Average latency in milliseconds for all operations on the volume

read_ops Number of reads per second to the volume

read_data Bytes read per second from the volume

read_latency Average time for reads to the volume

write_ops Number of writes per second to the volume

write_data Bytes written per second (b_per_sec) to the volume

write_latency Average time for writes to the volume

other_ops Number of other operations per second to the volume

other_latency Average time for other operations to the volume

internal_msgs Number of internal backdoor messages per second to the volume

read_blocks Number of blocks read per second from the volume

write_blocks Number of blocks written per second to the volume

nfs_read_ops Number of NFS reads per second to the volume


nfs_read_data Bytes read per second via NFS from the volume

nfs_read_latency Average time for NFS reads to the volume

nfs_write_ops Number of NFS writes per second to the volume

nfs_write_data Bytes written per second via NFS to the volume

nfs_write_latency Average time for NFS writes to the volume

nfs_other_ops Number of other NFS operations per second to the volume

nfs_other_latency Average time for other NFS operations to the volume

cifs_read_ops Number of CIFS reads per second to the volume

cifs_read_data Bytes read per second via CIFS from the volume

cifs_read_latency Average time for CIFS reads to the volume

cifs_write_ops Number of CIFS writes per second to the volume

cifs_write_data Bytes written per second via CIFS to the volume

cifs_write_latency Average time for CIFS writes to the volume

cifs_other_ops Number of other CIFS operations per second to the volume

cifs_other_latency Average time for other CIFS operations to the volume

san_read_ops Number of block protocol reads per second to the volume

san_read_data Bytes read per second via block protocol from the volume

san_read_latency Average time for block protocol reads to the volume

san_write_ops Number of block protocol writes per second to the volume

san_write_data Bytes written per second via block protocol to the volume

san_write_latency Average time for block protocol writes to the volume

san_other_ops Number of other block protocol operations per second to the volume

san_other_latency Average time for other block protocol operations to the volume


What access permission is

Access permissions allow you to customize views, change storage system settings, set eventthresholds, and configure events and alarms.

You can use role-based access control (RBAC) to grant access to the user. The access to resources isdefined and controlled based on the role or job function of a user.

Next topics

RBAC for configuration of views on page 45

RBAC for changing storage system settings on page 46

RBAC for thresholds on page 46

RBAC for threshold templates on page 47

RBAC for configuration of events on page 48

RBAC for configuration of alarms on page 48

RBAC for configuration of viewsYou can create, modify, and delete a custom view using role-based access control (RBAC).

The following table lists the tasks, the capabilities, and the objects you need to configure views.

Task Capability Object

See the performance data of acounter in a view

Database Read Managed object associated with thecounter instance in the view

Create a custom view PerfView Write Global

Modify a custom view: Add acounter

PerfView Write Global

Modify a custom view: Remove acounter

Database Read Managed object associated with thecounter instance in the view

Modify a custom view: Renaminga view

PerfView Write Global

Delete a custom view PerfView Delete Global

View real-time performance data PerfView RealTimeRead Global

Note: To view real-time counter data in a view, you require the Database Read capability.

What access permission is | 45

RBAC for changing storage system settingsYou can change the settings of a storage system using role-based access control (RBAC).

The following table lists the tasks, the capability, and the object you need to change the settings ofstorage system.


Modify a custom view

• Change credentials

• Transport protocol of a storagesystem

Database Write Storage system

RBAC for thresholdsYou can create, view, or edit thresholds using role-based access control (RBAC).

The following table lists the threshold operations that you can perform on objects, the capabilitiesyou need, and the objects on which you can perform the operation.


Create or edit thresholds on thecounters of the following objects:

• vFiler unit

• Aggregate

• Volume

• Qtree

• LUN

• Disk

• Network interface

Database Write On each instance of the object



View thresholds on the counters ofthe following objects:

• vFiler unit

• Aggregate

• Volume

• Qtree

• LUN

• Disk


Database Read On each instance of the object

Create or edit thresholds on thecounters of all the other objectssuch as the following:

• System

• NFSv3

• CIFS

• FCP

• iSCSI

Database Write Storage system

View thresholds on the counters ofall the other objects such as thefollowing:

• System

• NFSv3

• CIFS

• FCP

• iSCSI

Database Read Storage system

RBAC for threshold templatesYou can create, view, or edit threshold templates using role-based access control (RBAC).

The following table lists the threshold template operations that you can perform on objects, thecapabilities you need, and the objects on which you can perform the operation.


Create or edit threshold templates PerfThreshTemplate Write



View threshold templates PerfThreshTemplate Read

Apply or remove a thresholdtemplate on the following objects:

• Storage system or a group ofstorage systems

• Aggregate

• Volume

• Qtree

• LUN

• Disk


Database Write On each instance of the object

RBAC for configuration of eventsYou can configure events using role-based access control (RBAC).

The following table lists the tasks, the capabilities, and the objects you need to configure events.


View event Event Read Source of the event

Delete event Database Write Source of the event

Acknowledge event Event Write Source of the event

RBAC for configuration of alarmsYou can configure alarms using role-based access control (RBAC).

The following table lists the tasks, the capabilities, and the objects you need to configure alarms.


Creating an alarm Alarm Write Group associated with the alarm

Modifying an alarm Alarm Write Group associated with the alarm

Deleting an alarm Alarm Delete Group associated with the alarm

Listing all alarms Alarm Read Group associated with the alarm



Performance Advisor and server setup

Before you use Performance Advisor, you must install NetApp Management Console and ensure thatthe prerequisites are met.

The prerequisites to access Performance Advisor are as follows:

• Install NetApp Management Console.• Enable the performance-monitoring server.• Install DataFabric Manager on the DataFabric Manager server.

• Install Data ONTAP 6.5 or later on all storage systems and the hosts of all vFiler units for whichyou want to monitor performance data.

• Expose the storage systems and vFiler units for which you want to monitor performance data toDataFabric Manager.

Note: For details on downloading and installing NetApp Management Console, see the DataFabricManager Installation and Upgrade Guide. On upgrading to DataFabric Manager 3.6 or later, theapplication graphs that show the calculated stats display data for the period after the upgrade.

Next topics

Enabling the performance-monitoring server on page 51

Secure connections for Performance Advisor on page 52

Enabling the performance-monitoring serverBefore you start NetApp Management Console and log in to Performance Advisor, you must enablethe performance-monitoring server.

Steps

1. In Operations Manager, select Control Center > Setup > Options > Performance Advisor todisplay the Performance Advisor options.

2. If the application is disabled, in Performance Advisor Options, select Performance AdvisorEnabled > Enabled.

By default, the DataFabric Manager server enables the performance-monitoring server.

3. Click Update.

Performance Advisor and server setup | 51

Secure connections for Performance AdvisorPerformance Advisor supports both HTTP and HTTPS connections between the performance-monitoring server, Performance Advisor, and the monitored storage systems and vFiler hosts.

Note: Although HTTPS connections are the most secure, unnecessarily configuring HTTPS as thetransport protocol between the performance-monitoring server and the monitored devices mightadversely affect the performance of those devices.

Next topics

Setting up transport between the performance-monitoring server and PerformanceAdvisor on page 52

Setting up transport protocols between Performance Advisor and monitored devices on page 53

Setting up appliance-level transport between Performance Advisor and monitoreddevices on page 53

List of options for appliance-level setting on page 54

Setting up transport between the performance-monitoring server andPerformance Advisor

By default, both HTTP and HTTPS are enabled as transport protocols between the performance-monitoring server and Performance Advisor. You can change the transport protocol andperformance-monitoring server ports by using the Performance Advisor options in OperationsManager.

Steps


2. Select among the following options to accept or modify the settings as appropriate to yourenvironment:

• Server HTTP Port—Sets the port used by the performance-monitoring server to receive HTTPconnections from the application. The default value is 8088.

• Server HTTP Enabled—Enables or disables HTTP as a transport protocol used by theapplication. The default value is Yes.

• Server HTTPS Port—Sets the port used by the performance-monitoring server to receiveHTTPS connections from the application. The default value is 8488.

• Server HTTPS Enabled—Enables or disables HTTP as a transport protocol used by theapplication. The default value is Yes.

3. Click Update.


Setting up transport protocols between Performance Advisor andmonitored devices

You can set up transport protocols between Performance Advisor and monitored devices by enablingeither both HTTP and HTTPs or HTTPs only as transport protocols in Operations Manager.

Before you begin

You must ensure that there is no conflicting transport setting between Performance Advisor and theindividual storage systems or vFiler hosts.

Steps


2. Select among the following Performance Advisor Transport options to accept or modify thesettings as appropriate to your environment:

Options Description

httpOnly Specifies HTTP only as the global transport protocol between the performance-monitoringserver and all monitored devices. This is the default setting

httpsOk Enables both HTTP and HTTPS as global transport protocols between the performance-monitoring server and all monitored devices.

Note: When the transport protocol for the host is HTTPS, you must not set PerformanceAdvisor Transport to HTTP.

3. Click Update.

Setting up appliance-level transport between Performance Advisor andmonitored devices

You can edit the Performance Advisor Transport setting for an individual storage system or avFiler host using Operations Manager. You can specify whether the individual device can use HTTPonly or HTTP and HTTPS as the transport protocol between Performance Advisor and theperformance-monitoring server.

Steps

1. Click the name of the device in Operations Manager.

2. Click Edit Settings.

3. Select the options you want from the Performance Advisor Transport list.

Performance Advisor and server setup | 53

4. Click Update.

List of options for appliance-level settingDataFabric Manager uses SNMP to determine if each storage system or vFiler host requires HTTPStransport. If HTTPS is required, the performance-monitoring server must connect to the storagesystem or vFiler host by using HTTPS transport. You cannot override that requirement using theoptions for appliance-level setting.

Setting Description

Empty[default]

Allows the global Performance Advisor Transport setting to take precedence.

httpOnly If... Then...

DataFabric Manager has determined that HTTP isused as the transport to connect with this storagesystem

Performance-monitoring is enabled forthis storage system or vFiler host.

DataFabric Manager has determined that HTTPS isused as the transport to connect with this storagesystem

Performance-monitoring is disabled forthe storage system or vFiler host.

httpsOk Enables performance-monitoring for the storage system or vFiler host regardless of whetherthe storage system is using HTTP or HTTPS as a transport.

disabled Disables performance-monitoring for the storage system.


Introduction to the Performance Advisor userinterface

Using NetApp Management Console, you can log in to the DataFabric Manager server to access andstart or stop the Performance Advisor application.

Next topics

Starting Performance Advisor on page 55

Stopping Performance Advisor on page 56

Unavailable storage systems or vFiler units in Performance Advisor on page 56

Unavailable menu options in the Performance Advisor interface on page 56

Zero-value data in a Performance Advisor chart on page 56

Absence of data in a Performance Advisor chart on page 57

Display of new storage systems or vFiler units in Performance Advisor on page 57

Starting Performance AdvisorYou can start Performance Advisor by logging in to the DataFabric Manager server using NetAppManagement Console.

Steps

1. Click Start > All Programs > NetApp > Management Console.

Alternatively, start NetApp Management Console on Windows by clicking the NetAppManagement Console icon on the desktop.

2. Enter the fully qualified domain name, user name, and the password for the DataFabric Managerserver you want to log in to.

The login screen remembers the user name, domain name, transport, and port number from yourmost recent successful login. If you are logging in as the same user as before, you need to enteronly your password.

3. Select HTTP or HTTPS transport.

The default port is selected automatically. If you log in using any other port, you must set theserverHTTPSPort option to the matching port.

4. Click Connect.

5. Click the Manage Performance icon to display the top-level groups configured for the DataFabricManager server that you have logged in to.

Introduction to the Performance Advisor user interface | 55

Stopping Performance AdvisorYou can stop Performance Advisor by closing NetApp Management Console.

Step

1. From the menu, Select File > Exit.

Unavailable storage systems or vFiler units in PerformanceAdvisor

Storage systems or vFiler units that are not available are not shown in the View navigation pane.Instead, they are shown in the Set Up Hosts window.

Devices might be unavailable under the following conditions:

• Storage systems or vFiler units are running Data ONTAP versions earlier than 6.5.

• The DataFabric Manager server is not managing storage systems or vFiler units.

• Data collection from the DataFabric Manager server by the application is not enabled.

• You do not have sufficient access privileges to these storage systems or vFiler units.

Unavailable menu options in the Performance Advisorinterface

The availability of menu options in the Performance Advisor interface is determined by yourDataFabric Manager administrator roles. Options that are dimmed are not available to you for thestorage system or vFiler unit you have currently selected.

Zero-value data in a Performance Advisor chartWhen the performance-monitoring server sends nothing but zeros to Performance Advisor for thecounter or counters in a chart, this is called zero-value data.

Zero-value data is displayed for any counters monitoring protocols that are not enabled on the storagesystem or vFiler unit. A chart displaying only zero-value data shows a vertical axis numbered from 0to 1, and vertical lines, but no graph.


Absence of data in a Performance Advisor chartAbsence of data indicates that the performance-monitoring server has not sent any data toPerformance Advisor for the counters in a chart during the period displayed.

An empty chart indicates that you cannot access the data. This situation is unlikely, unless one of thefollowing conditions is true:

• You are viewing SnapMirror counters for a storage system that does not have a SnapMirrorlicense.

• "Top N luns" chart displays no data when there are no LUNs available.

• "Top qtrees in a dataset" chart displays no data when there are no qtrees available.

• "Priority queue" chart displays no data when the FlexShare option is off.

Display of new storage systems or vFiler units inPerformance Advisor

If the DataFabric Manager server discovers new storage systems or vFiler units, or if you manuallyadd storage systems or vFiler units, you must refresh the application display before the applicationincludes the new devices.

You must press F5 to refresh the display.

Introduction to the Performance Advisor user interface | 57

Overview of the NetApp Management Consoledashboards

Dashboards provide a broad overview of various aspects of your data management environment.

Performancedashboard

The Monitor dashboard in Performance Advisor displays information about thecurrent overall performance status of your systems, such as top performanceevents, top storage systems by the total number of operations, top storage systemsby network throughput, and top storage systems by CPU utilization. You canmodify the default settings to configure any view on the global group as adashboard.

To access an application dashboard, you must have appropriate privileges. Items that you do not haveprivileges for are not displayed in the dashboard. If you encounter access problems, contact theadministrator who maintains your DataFabric Manager roles and privileges.

Next topics

Dashboard panel descriptions on page 59

What events are on page 60

How to know when an event occurs on page 60

Dashboard panel descriptionsThe dashboard panels provide cumulative at-a-glance status information about your system.

Performance Advisor dashboard

You can view more details about the Performance Advisor dashboard by clicking in thedashboard panel title bar.

Top PerformanceEvents

Displays the five events with the highest severity levels. Items are orderedfirst by severity, then by time of the events. More detail about each event isprovided in Monitor > Events.

Top StorageSystems byNetworkThroughput

Displays a bar chart of the top five DataFabric Manager storage systems,sorted by the highest network throughput. The number above each bar chartdisplays the exact value of the throughput per second from that system. Thevertical axis displays the megabytes of throughput per second. Thehorizontal axis displays the name of the storage system.

Overview of the NetApp Management Console dashboards | 59

Top StorageSystems by TotalOPs

Displays a bar chart of the top five DataFabric Manager storage systems,sorted by the highest total operations. The vertical axis displays the totaloperations per second for that storage system. The horizontal axis displaysthe storage system name.

Top StorageSystems by CPUUtilization

Displays a bar chart of the top five DataFabric Manager storage systems,sorted by the highest average CPU utilization. The number above each barchart displays the exact value of the throughput per second from that system.The vertical axis displays the percentage CPU usage of the storage system.The horizontal axis displays the name of the storage system.

What events areEvents are generated automatically when a predefined condition occurs or when an object crosses athreshold. Event messages inform you when specific events occur. All events are assigned a severitytype and are automatically logged in the Events window.

You can configure alarms to send notification automatically when specific events or severity typesoccur. If an application is not configured to trigger an alarm when an event is generated, you can findout about the event by checking the Events window.

It is important that you take immediate corrective action for events with severity level Error or worse.Ignoring such events can lead to poor performance and system unavailability.

Note: Event types are predetermined. Although you cannot add or delete event types, you canmanage notification of events.

How to know when an event occursYou can learn about event occurrences by viewing the event log or by configuring alarms toautomatically notify you when events occur.

• Viewing the event logYou can use the Events window to view a list of all events that occurred and to view detailedinformation about any selected event.

• Configuring alarmsYou can use the Alarms window to add an alarm that sends notifications automatically when anevent occurs.

Next topics

Monitoring events on page 61

Responding to and acknowledging events on page 61

Deleting events on page 62


Monitoring eventsYou can view a list of all events that occurred and view detailed information about any event. Youcan also view deleted events that are marked for deletion.

Before you begin

Ensure that you are authorized to perform all steps of this task. Your RBAC administrator canconfirm your authorization in advance.

Steps

1. For licensed protection and provisioning applications, from the navigation pane,click Notifications > Events. For Performance Advisor, from the navigation pane, click Monitor> Events.

2. (Optional) You can customize the Events window in any of the following ways:

• Select an event to view details about that event.• Click in a column header to control which event entries you want displayed.• Click the sort arrows in a column header to change the sort order of the entries in that column.•

Click in the upper-right corner of the list to select which columns are displayed.• Drag the bottom of the events list area up or down to resize that area.

3. (Optional) You can view events that are marked for deletion by displaying hidden columns andremoving a text filter, as follows:

a.Click in the upper-right corner of the list and select the Deleted By and Deleted Timecolumns.

b. Click in the Deleted By column header and delete the default search string.

The default filtering of deleted events is stopped and the deleted event entries are displayed inthe list.

Responding to and acknowledging eventsWhen an event occurs, you must take appropriate action to acknowledge and manage it and correctthe problem. Responding to an event includes viewing it, acknowledging it, and correcting theproblem.

Before you begin


Overview of the NetApp Management Console dashboards | 61

Steps

1. For licensed applications, from the navigation pane, click Notifications > Events. ForPerformance Advisor, from the navigation pane, click Monitor > Events.

2. In the Events window, select an event to view the details about it.

3. Select an event and click Acknowledge to indicate that you are taking responsibility formanaging that event.

4. In the Acknowledge Events dialog box, click Yes to acknowledge the selected event.

Your user name and the time are entered in the Events list for the selected event.

5. Find the cause of the event and take corrective action.

Deleting eventsYou can delete an event from the event list. Typically, you delete an event only after youacknowledge it and resolve the problem. However, you can also delete unacknowledged events ifthey are not important.

Before you begin


Steps


2. In the Events window, select an event and click Delete.

3. In the Delete Events dialog box, click Yes to delete the selected event.

4. Verify that the event is deleted by viewing the Events window.

After you finish

You can view events that are marked for deletion by clicking in the corner of the list.


Views in Performance Advisor

You can use views in Performance Advisor to analyze and interpret performance data.

Next topics

What the performance view window is on page 63

Performance-monitoring with custom views on page 73

Analysis of real-time data on page 77

Canned views at various levels on page 79

What the performance view window isThe performance view window enables you to view the performance data of the selected group or theobject in the hierarchy.

Next topics

What default views are on page 64

What canned views are on page 64

What custom views are on page 64

What advanced views are on page 65

What real-time data is on page 66

What historical data is on page 66

Hierarchy of objects on page 66

Analyzing historical data on page 66

What synchronizing time axes is on page 67

Viewing performance data on page 67

Displaying tabbed views on page 67

Selecting a performance view on page 67

Closing a performance view on page 68

Data in performance views on page 68

Changing the time period in a chart on page 69

Resetting both axes on page 69

Synchronizing time axes on page 69

Undoing synchronization of time axes on page 70

How to use the legend to interpret performance data on page 70

How to view thresholds on page 70

Creating advanced views on page 70

Views in Performance Advisor | 63

Adding thresholds using a view on page 71

Changing the chart type on page 71

Playing or pausing data collection on page 71

Printing a chart from a view on page 72

Enabling charts to scale the y axis on page 72

Diagnosis of performance issues on page 72

Diagnosing a performance issue on page 73

What default views areYou can use default views to view performance data as soon as the performance-monitoring serverand Performance Advisor start working.

From the navigation pane, you can click the object to get to the default view. You can set any view asdefault by using the Set Default button.

For example, you might want to monitor the top physical objects. The default view is StorageSystem Summary View. Using the Views list and Set Default button, you can set the StorageSystem - Top Physical Objects view for the storage system to appear by default. The StorageSystem Summary View is no longer the default view that appears when you click the storagesystem.

Note: If you change the default view of the global group, then the Dashboards window uses thenew view to display the panels.

What canned views areCanned views are views that are available in Performance Advisor and are not created by you. Youcannot customize, edit, or delete canned views such as the Storage System Summary View.However, you can save a copy of the canned view and create a custom view.

What custom views areYou can customize performance data monitoring through custom views. You can create viewsthrough the Performance Advisor interface.

There are no restrictions on which counters you can combine to display in a single view.Additionally, you can include multiple counters in a single chart, and multiple charts in a single view.By default, custom views are associated with the Global group.

Note: On upgrading to DataFabric Manager 3.6 or later, multiple custom views with the samename in the earlier versions of Performance Advisor are shown in the view name>_view Id format.For example, if there are multiple custom views with the same name cv1, then after the upgrade,these custom views are shown in the cv1_view Id format. Since the view Id is unique, the customview name is unique after the upgrade.

Because custom views offer so much flexibility, it is advisable to consider the following objectivesbefore creating a custom view:


• Access real-time data, other than what appears in the default view• Group counters into combinations that are not offered by default views• Use a chart type different from the charts in the default view

You can temporarily select a different chart type in a default view through the right-click menu.However, the view returns to the default chart type each time you invoke it from the navigationpane.

Note: Although you can combine any counters in a single chart, if you combine counters withdifferent units, the chart shows multiple vertical (“y”) axes with the respective unit labels of thecounters. For example, combining the disk_data_read counter (KB per second) with theprocessor_busy counter (percentage) results in a chart with multiple vertical (“y”) axes, whereeach axis displays the respective unit label of the counter.

What advanced views areAdvanced views enable you to view the performance data based on statistical parameters for a periodof time. You can set parameters such as minimum, maximum, mean, and percentile on the collecteddata for further consolidation.

These views enable you to set the baseline for data collection and thus help you in setting appropriatethresholds. The following are the three types of computations that are applicable to the statisticalparameters:

• SimpleThis is a single value of the statistical parameter for the whole period of time. A horizontal line isdisplayed in the view.

• StepThe values are accumulated for a specified period. This calculation yields multiple values anddisplays a series of horizontal lines in steps.

• RollingThe values are accumulated for a specified period in regular intervals. This calculation yieldsmultiple values and displays smoothened lines.

The statistical computations are based on the Boost.Accumulators library.

Note: Computation types in advanced views are calculated based on the unconsolidatedperformance data. In the other views, data is graphically represented based on the consolidatedperformance data (using the performance settings.) Therefore, the baseline might not meet thegraph in the advanced views. This might result in inconsistencies in the advanced views.

Related information

Boost.Accumulators - www.boost.org/doc/libs/1_38_0/doc/html/accumulators.html


http://www.boost.org/doc/libs/1_38_0/doc/html/accumulators.html

What real-time data isReal-time data is data that is passed through the performance-monitoring server for display.However, the data is not stored.

Real-time data is collected only as long as the Summary View window displaying it is open in thePerformance Advisor interface. The data is accessible only to the application from which it isretrieved. The data is suitable for diagnosing immediate performance issues. You must either save orexport the real-time data for analysis.

What historical data isHistorical data is data that is archived on the DataFabric Manager server. Performance Advisoraccesses the data by connecting to the server.

Historical data is collected on an ongoing basis, independent of whether the Summary View windowin the application is open or not. The data can be used for diagnosing past performance problems orfor short-term trend analysis.

Hierarchy of objectsYou can view the hierarchy of logical and physical objects in a navigation tree from the navigationpane.

By default, only the hierarchy of logical objects appears in the navigation pane. From the navigationpane, click View > Physical to view the hierarchy of physical objects.

Analyzing historical dataYou can analyze the historical data by selecting the time range on a chart. You can update the timerange on all charts by synchronizing time axes.

Steps

1. Select the chart in a view.

2. Click Range > Custom.

3. Edit Start Date and End Date.

4. Click OK.


What synchronizing time axes isYou can synchronize the axes when you want the time axes of all line charts in a view to be in sync.It helps in detecting performance blockage by interpreting data through views.

For example, if you observe a spike in one of the charts in the Summary View window, you canprobe further by interpreting data in the related charts of the window. You must synchronize the timeaxes on all charts to investigate the data for a particular time period.

Viewing performance dataYou can view the performance data of an object by selecting the object in the navigation tree.

Steps

1. From the navigation pane, click View.

2. Expand the container objects in the navigation tree to view the objects within them.

The parent object displays a tree of objects.

3. Click an object.

The default view of the selected object appears.

Note: You can change the default view by using the Set Default button.

Displaying tabbed viewsYou can view the performance data of multiple groups or objects at the same time using tabbedviews.

Steps


2. Right-click an object in the navigation pane.

3. Click Open in New Tab.

Selecting a performance viewYou can select a performance view from a list to view the performance view of your choice. TheAvailable views list in performance view window displays a list of various types of performanceviews.

Step

1. In the Summary View window, select the view from the Available Views list.


Closing a performance viewYou can close any current view using the x icon.

Step

1. Click the x icon.

You should click the x icon to close the tab in which you are currently viewing the performancedata.

Data in performance viewsYou can examine and interpret data in performance views using charts. The charts show performancedetails in the form of graphs.

You can take the following actions to examine and interpret data in charts:

• List all related objects using the arrow icon in the upper-right corner of the title bar in the chart.The tree view of the objects, which is a tree of objects related to the objects of the chart, appears.

• Select the range of data to be displayed, such as displaying the data of performance detailscollected for the last one hour.

• Zoom in to the chart by 10%.• Zoom out of the chart by 10%.• Plot the Performance Advisor events on the chart and identify the cause of spikes.

You can identify the nature of event by moving the pointer over the plotted event.• Reset the axes.• Add one or more counters.• Add one or more thresholds.• Save as an image file.• Print the chart.

If a chart contains managed objects and unmanaged objects, the tree view of the objects that appearin that particular chart shows only the managed objects. The tree view is blank if any chart containsunmanaged objects.

You can maximize or minimize the chart by double-clicking its title. Any view that contains only onechart is always maximized.

In a custom view, when you stop the data collection, the historical view chart displays a blank spacefrom the point of time the data collection was stopped. When the data collection starts again, thepolyline graph connects the point at which the data collection was stopped to the point where the datacollection is restarted. To identify the break in the data collection, you should go to the respectivescatter graph to know whether there is any disconnection.


Changing the time period in a chartYou can move the time period backward and forward to view the performance details for the selectedportion of time.

About this task

The screen resolution determines the amount of data that appears (as opposed to the entire range ofdata you can scroll to). Therefore, resizing a chart does not increase or decrease the amount of datayou can view; it only increases or decreases the number of pixels per data point. The zoom buttonshelp you compress or lengthen the time scale and reduce or enlarge the period of time in the chart.

Step

1. Click the buttons located next to the Show Threshold box to view the performance data.

• Click l< to move to the beginning of data collection.• Click > to move forward in the time scale.• Click < to move backward in the time scale.• Click >l to move to the ending of the time scale.

Resetting both axesYou can reset the axes in any chart of a view. The “x” or horizontal axis represents time in line chartsand objects in bar charts. The units on the “y” or vertical axis vary, depending on the countersincluded in the chart.

Step

1. Click Action > Reset Both Axes.

Alternatively, right-click anywhere in the chart and click Reset Both Axes.

Synchronizing time axesYou can synchronize the time axis in charts by using Performance Settings and activatingSynchronize in a view.

Steps

1. Click the Performance Settings icon in the menu bar.

2. Select one of the following options to accept or modify the settings:

• Synchronize all charts in a view• Synchronize all charts across all views


3. Click View and select a view from Available views.

4. Click View Actions > Synchronize.

Undoing synchronization of time axesYou can undo the synchronization of the time axis in charts by using Undo synchronize in a view.

Steps


2. Click View Actions > Undo synchronize.

How to use the legend to interpret performance dataYou can interpret performance data using legends. Each performance view provides a legend.

The legend for a canned view contains the following information:

• A sample of the graph color• The name of the counter• Unit of x-axis• Unit of y-axis

In addition to the above, the legend for a custom view includes the following information:

• The name of the storage system or vFiler unit, subsystem, or group of storage system or vFilerunits from which the data is being collected

To access the legend, you can select the Show Legend check box located below the time scale in thechart. To hide the legend, clear the Show Legend check box.

Using the legend, you can select the counters which you want to view in the performance chart.

How to view thresholdsYou can view thresholds along with the performance data for a counter using views.

You can select the Show thresholds check box located next to the Show legends check box to viewthresholds. The dotted lines that appear in the chart indicate the threshold values.

Creating advanced viewsYou can create advanced views by using the View window for any storage system.

Steps



2. Select the storage system.

3. Ensure that you are viewing Storage System Summary View.

4. Click View Actions > Metrics.

5. Click Metric Settings.

6. Select metric and computation types with period settings.

7. Click Add metric and then click OK.

8. Click Filter Settings.

9. Specify the time and day of the week and click OK.

Adding thresholds using a viewYou can use a view to add thresholds for the counters displayed on the chart.

Steps


2. Right-click the chart.

3. Move the pointer over Add Threshold and click a counter.

Changing the chart typeYou can identify the breaks in data collection by changing the type of chart.

Steps

1. Display the performance view.


3. Move the pointer over Chart Type and click Scatter Chart.

Playing or pausing data collectionYou can play or pause the data collection if you want to examine a graph in detail. This is useful ifthe graph is updating frequently.

About this task

Note: The play/pause functionality affects every chart in the current view.


Steps

1. Select a view from View > Available views.

2. Click View Actions.

3. Click Pause once to pause data collection.

However, the DataFabric Manager server does not stop collecting data.

4. Click Play again to restart the data collection.

Printing a chart from a viewYou can print a chart by clicking in the view.

Steps


2. Right-click in the chart.

3. Click Print.

4. Specify the name of the printer and click OK.

Enabling charts to scale the y axisYou can enable charts to scale the vertical axis automatically for best view during various periods ofdata collection.

Steps


2. Right-click in the chart.

3. Click Auto Scale Y-Axis.

Diagnosis of performance issuesYou can use Performance Advisor to diagnose problems that cause performance bottlenecks on astorage system.

During the diagnosis, Performance Advisor investigates the following categories for a storagesystem:

• Data ONTAP• Network• NAS


• SAN

After the diagnosis, Performance Advisor provides observations and recommendations for eachcategory.

In addition, events are plotted in performance views to mark values that are indicative ofperformance degradation. You can correlate events generated for an object and its related objectsduring any period of time.

The DataFabric Manager server indicates the impact of performance on a storage system bygenerating events such as disks:some-failed. Such events, which are not triggered due to a breach ofuser-defined thresholds, are also plotted in the performance views.

Diagnosing a performance issueYou can diagnose a performance issue by checking the health of Data ONTAP, network, NAS, andSAN on a storage system.

Steps


2. Open a performance view of a storage system.

3. Click View Actions.

4. Click Diagnose.

5. Enter the period of data collection.

6. Click Diagnose.

After you finish

You must check the details of each category and act on recommendations, if necessary.

Performance-monitoring with custom viewsCustom views expand the ways that the performance client can display data. Custom views enableyou to present data in combinations or styles that are not possible with canned views.

For example, you might want to monitor a single counter across several storage systems or vFilerunits. You might want to use canned views to view counter combinations, but in a different type ofchart.

Next topics

When you must create a custom view on page 74

How flexible the custom views are on page 74

Counter data across views on page 74


Counters with different units in a single chart on page 74

Viewing real-time data on page 75

Creating a custom view on page 75

Opening a custom view on page 76

Editing a custom view on page 76

Deleting a custom view on page 76

CPU utilization issues on storage systems or vFiler hosts on page 77

When you must create a custom viewYou must create a custom view if you want to access real-time data, display counters in combinationsthat are not offered by canned views, use chart types different from that of canned views, and usedifferent views to compare the performance of the same counter on different objects.

Note: You can temporarily select a different chart type in a canned view through the right-clickmenu. However, the view returns to the default chart type each time you open it from thePerformance Views panel.

How flexible the custom views areIf you use custom views, there are no restrictions on which counters you can combine in a singleview. You can include multiple counters in a single chart and multiple charts in a single view.

Additionally, you have the option to apply a custom view for all objects of the selected type. Forexample, you can create a view with several charts and apply that view to all storage systems in agroup.

Counter data across viewsCounter data is shared across views. Any data collected for the counter in other performance viewsappears in the new view.

When you add a counter to a performance view, if data collection is enabled for this counter inanother view, then the new view carries forward the collection.

Counters with different units in a single chartAlthough you can combine any counters in a single chart, combining counters with different unitsresults in a chart with multiple vertical (“y”) axes.

For example, combining the disk_data_read counter (KB per second) with the processor_busycounter (percentage) results in a chart with multiple vertical (“y”) axes, where each axis representsthe respective unit labels of the counters.


Viewing real-time dataYou can access either the historical data or the real-time data by selecting the respective data fromthe list. All performance views initially display historical data when they appear in PerformanceAdvisor.

About this task

Note: Real-time data collection can increase storage system CPU utilization. This mightnegatively impact the storage system.

Steps

1. Select a view from the Available Views list.

2. Click View Actions > Real Time.

3. Select Real Time.

4. Edit Sample Rate and Retention Period.

5. Click OK.

6. Click Yes to Performance Warning.

Creating a custom viewYou can create a custom view with selected counters in bar and line charts in Performance Advisor.

Before you begin

Ensure that you are assigned the DataFabric Manager GlobalWrite and GlobalDelete roles.

Steps

1. Click Set Up > Custom Views.

2. Click Add.

3. In the View name field, type the name of the custom view that you want to create.

The custom view names should be unique.

4. Select where you want the view to be applied.

5. Click Add Chart to add charts to the view.

You can add more than one chart to the view.

6. Select Show events to display the top five events for each object that is associated with the view.

7. Click Add.


Opening a custom viewYou can open a custom view in the performance view window.

Steps


2. Select the custom view and click Open.

Alternatively, right-click the custom view and select Open.

Editing a custom viewYou can edit the view name, the chart name, and the counter details in a custom view.

Before you begin

Ensure that you are assigned the GlobalWrite and GlobalDelete roles.

Steps


2. Select the custom view that you want to modify and click Edit.

You can edit a custom view in either of the following ways:

• Right-click the selected custom view and click Edit.• Double-click the custom view that you want to edit.

3. In the properties sheet for the selected custom view, click the fields you want to modify, and enterthe new values.

4. Click Apply.

Note: You can edit only custom views (not canned views).

Deleting a custom viewYou can delete a custom view from the Custom Views window.

Before you begin

Ensure that you are assigned the GlobalWrite and GlobalDelete roles that enable you to delete acustom view.


Steps


2. Select one or more custom views in the Custom Views window and click Delete.

Alternatively, you can select one or more custom views in the following way:

a. Press the Ctrl key.

b. Right-click and click Delete.

3. Click OK.

Note: You can delete only custom views (not canned views).

CPU utilization issues on storage systems or vFiler hostsThe performance-monitoring settings might increase the storage system CPU utilization beyonddesirable levels.

You must check the current CPU utilization level on the storage system or vFiler host before usingthe following settings:

• Enable SecureAdmin on a storage system, which requires DataFabric Manager to use HTTPStransport to monitor performance data

• Switch the data source in views to real time• Set the sample rate in views to a value less than 60 seconds

Analysis of real-time dataYou must either save or export the real-time data to analyze it. The real-time data is not archived andexists only when the performance view is displayed.

Next topics

Methods for preserving performance data on page 77

Saving a chart on page 78

Exporting a view to a .csv file on page 78

Limitations to exporting performance data on page 79

Methods for preserving performance dataYou can preserve the performance data of a storage system or a vFiler unit for analysis in eitherPerformance Advisor or in an external application.

The following two methods help you preserve the data:

• Saving a chart


• Exporting a view

Saving a chartYou can save the chart of the performance view with details in your directory.

About this task

Note: To get a good result for a .png file with proper details, you should maximize the chart viewby double-clicking the title bar of the chart and then save the chart.

Steps



3. Click Save as.

4. Select the directory in which you want to save the chart.

5. Specify the name of the file.

6. Click Save.

Exporting a view to a .csv fileYou can export the data from a view to a .csv (comma-separated value) file for analysis.

Steps

1. From the navigation pane, click the group or the object that you want to view and for which youwant to export data.

2. Click the name of the view you want to display.

3. From the File menu, select Export.

4. Select and specify the date range, the counters, and the date format.

For advanced views, you can specify the metric data that you want to export.

5. Click Browse.

6. Specify the location and the name of the file.

7. Select CSV in the Files of Type field to export the data to CSV format.

8. Click Save.

9. Click OK.


Result

The application saves a performance view file with a .csv extension to a location that you specify.

Note: You can view the .csv output file format in Microsoft Excel. You can also view this fileusing common text editors, because the data is in plain-text format.

Limitations to exporting performance dataYou can export performance data; however, there are some limitations.

• Exporting data is disabled for top performance charts.• If a view has top performance charts and detailed charts, then the Export dialog box displays only

the names of the detailed charts.

Canned views at various levelsYou can see canned views at various levels in the logical and physical hierarchy to interpret theperformance data.

Canned views are available at the following levels in the hierarchy:

• Global• Storage system• vFiler unit• Processor• Network Interface• Dataset• Resource pool• Volume• Qtree• LUN• Aggregate• Disk• Target

Next topics

Canned views at the global or group level on page 80

Canned views at the storage system level on page 82

Canned views at the vFiler level on page 87

Canned views at the processor level on page 88

Canned views at the network interface level on page 88

Canned views at the dataset level on page 89


Canned views at the resource pool level on page 89

Canned views at the volume level on page 90

Canned views at the qtree level on page 92

Canned views at the LUN level on page 93

Canned views at the aggregate level on page 94

Canned views at the disk level on page 94

Canned view at the target level on page 95

Canned views at the global or group levelYou can see the views at the global or group level to interpret the performance data.

Group Summary View

This view gives a summary for the entire system or group. This view is displayed in either of thefollowing cases:

• When the user first logs in• When any group is selected

The following panels are available under this view:

• Top Performance EventsCounters are not applicable for this panel.

• Top Storage Systems by Network ThroughputThis panel displays the load_total_mbps counter, which is calculated using the following formula:load_total_mbps = load_inbound_mbps + load_outbound_mbpsload_inbound_mbps = system:system:net_data_recv + fcp:fcp:fcp_write_data +iscsi:iscsi:iscsi_write_dataload_outbound_mbps = system:system:net_data_sent + fcp:fcp:fcp_read_data +iscsi:iscsi:iscsi_read_data

• Top Storage Systems by Total OpsThis panel displays the pa_total_ops counter, which is calculated using the following formula:system:pa_total_ops = system:system:nfs_ops + system:system:cifs_ops +system:system:fcp_ops + system:system:iscsi_ops + system:system:http_ops

• Top Storage Systems by CPU UtilizationThis panel displays the cpu_busy counter.

Group - Top Objects

This view shows the top five volumes and aggregates in a group. The following panels are availableunder this view:

• Top Volumes by Latency


This panel displays the avg_latency counter.• Top Volumes by Throughput

This panel displays the throughput counter, which is calculated using the following formula:throughput = volume:<instance>:read_data + volume:<instance>:write_data

• Top Volumes by Total OpsThis panel displays the total_ops counter.

• Top Aggregates by Total OpsThis panel displays the total_transfers counter.

Group -Top vFilers

This view shows the top five vFiler units in a group. The following panels are available under thisview:

• Top vFiler Units by Disk Data TransferredThis panel displays the vfiler_data_transferred counter, which is calculated using the followingformula:vfiler_disk_transferred = vfiler:<instance>:vfiler_read_bytes +vfiler:<instance>:vfiler_write_bytes

• Top vFiler Units by Network ThroughputThis panel displays the vfiler_nw_throughput counter, which is calculated using the followingformula:vfiler_nw_throughput = vfiler:<instance>:vfiler_net_data_recv +vfiler:<instance>:vfiler_net_data_sent

• Top vFiler Units by Total OpsThis panel displays the vfiler_total_ops counter, which is calculated using the following formula:vfiler_total_ops = vfiler:<instance>:vfiler_read_ops + vfiler:<instance>:vfiler_write_ops +vfiler:<instance>:vfiler_misc_ops

Top CPUs

This view shows the top five CPUs in a group. The following panels are available under this view:

• cpu_busyThis panel displays the cpu_busy counter.

Top Network Interfaces

This view shows the top five network interfaces in a group. The following panels are available underthis view:

• recv_dataThis panel displays the recv_data counter.

• recv_errorsThis panel displays the recv_errors counter.


• recv_packetsThis panel displays the recv_packets counter.

• send_dataThis panel displays the send_data counter.

• send_errorsThis panel displays the send_errors counter.

• send_packetsThis panel displays the send_packets counter.

Canned views at the storage system levelYou can see the views at the storage system level to interpret the performance data.

Storage system Summary View

This view gives a summary for a specific storage system. This is the default view that is displayedwhen a storage system is selected or navigated to.



• Network ThroughputThis panel displays the load_total_mbps, load_inbound_mbps, and load_outbound_mbpscounters, which are calculated using the following formulae:load_total_mbps = load_inbound_mbps + load_outbound_mbpsload_inbound_mbps = system:system:net_data_recv + fcp:fcp:fcp_write_data +iscsi:iscsi:iscsi_write_dataload_outbound_mbps = system:system:net_data_sent + fcp:fcp:fcp_read_data +iscsi:iscsi:iscsi_read_data

• Average Latency per ProtocolThis panel displays the nfsv3_avg_op_latency, nfsv4_op_latency, and cifs_latency counters.

• All Protocol OpsThis panel displays the following counters:

• system:system:nfs_ops• system:system:cifs_ops• system:system:http_ops• system:system:fcp_ops• system:system:iscsi_ops

• CPU UtilizationThis panel displays the cpu_busy counter.


Storage system - Top Logical Objects

This view shows the top five logical objects, such as volumes, qtrees, and LUNs, based on differentcounters.


• Top Volumes by OpsThis panel displays the volume:<instance>:total_ops counter.

• Top Volumes by LatencyThis panel displays the volume:<instance>:avg_latency counter.

• Top LUNs by OpsThis panel displays the lun:<instance>:total_ops counter.

• Top LUNs by LatencyThis panel displays the lun:<instance>:avg_latency counter.

• Top Qtrees by Total OpsThis panel displays the total_ops counter, which is a sum of qtree:<instance>:nfs_ops andqtree:<instance>:cifs_ops counters.

Storage system - Top Physical Objects

This view shows the top five aggregates and disks in a storage system.


• Top Aggregates by Total OpsThis panel displays the aggregate:<instance>:total_transfers counter.

• Top Disks by Disk Busy PercentageThis panel displays the disk_busy counter.

Storage system - Per Protocol Latencies

This view shows the details of the latencies for different protocols at the storage system level.


• Average Latency per ProtocolThis panel displays the nfsv3_avg_op_latency, nfsv4_op_latency, and cifs_latency counters.

• NFSv3 Latency by OptypeThis panel displays the nfsv3_avg_op_latency, nfsv3_read_latency, and nfsv3_write_latencycounters.

• NFSv4 Latency by OptypeThis panel displays the nfsv4_avg_latency counter.

• CIFS Latency by OptypeThis panel displays the cifs_latency counter.

• FCP Latency by Optype


This panel displays the fcp_read_latency and fcp_write_latency counters.• iSCSI Latency by Optype

This panel displays the iscsi_read_latency and iscsi_write_latency counters.

Storage system - Per Protocol Sizes

This view shows the details of the read and write sizes for different protocols at the storage systemlevel.


• NFSv3 Read SizesThis panel displays the nfsv3_read_size_histo counter.

• NFSv4 Read SizesThis panel displays the nfsv4_read_size_histo counter.

• NFSv3 Write SizesThis panel displays the nfsv3_write_size_histo counter.

• NFSv4 Write SizesThis panel displays the nfsv4_write_size_histo counter.

Storage system - Per Protocol Ops

This view shows the details of the total operations for different protocols at the storage system level.


• NFSv3 Ops by OptypeThis panel displays the nfsv3_read_ops, nfsv3_write_ops, and nfsv3_ops counters.

• NFSv4 Ops by OptypeThis panel displays the nfsv4_read_ops, nfsv4_write_ops, and nfsv4_ops counters.

• CIFS Ops by OptypeThis panel displays the system:cifs_ops counter.

• FCP Ops by OptypeThis panel displays the fcp_read_ops, fcp_write_ops, and fcp_ops counters.

• iSCSI Ops by OptypeThis panel displays the iscsi_read_ops, iscsi_write_ops, and iscsi_ops counters.

• All Protocol OpsThis panel displays the nfs_ops, cifs_ops, http_ops, fcp_ops, and iscsi_ops counters.

Storage system - Network Throughput

This view shows the details of the network throughput at the storage system level.


• Total Throughput


This panel displays the san_throughput and nas_throughput counters.This panel also displays the load_total_mbps, load_inbound_mbps, and load_outbound_mbpscounters, which are calculated using the following formulae:load_total_mbps = load_inbound_mbps + load_outbound_mbpsload_inbound_mbps = system:system:net_data_recv + fcp:fcp:fcp_write_data +iscsi:iscsi:iscsi_write_dataload_outbound_mbps = system:system:net_data_sent + fcp:fcp:fcp_read_data +iscsi:iscsi:iscsi_read_data

• NAS ThroughputThis panel displays the net_data_recv, net_data_sent, and nas_throughput counters.The nas_throughput counter is calculated using the following formula:nas_throughput = net_data_recv + net_data_sent

• FCP ThroughputThis panel displays the fcp_write_data, fcp_read_data, and fcp_throughput counters.The fcp_throughput counter is calculated using the following formula:fcp_throughput = fcp_write_data + fcp_read_data

• iSCSI ThroughputThis panel displays the iscsi_write_data, iscsi_read_data, and iscsi_throughput counters.The iscsi_throughput counter is calculated using the following formula:iscsi_throughput = iscsi_write_data + iscsi_read_data

• Disk Traffic to User Traffic ratioThis panel displays the load_write_bytes_ratio and load_read_bytes_ratio counters, which arecalculated using the following formulae:load_write_bytes_ratio = disk_data_written/load_ inbound_mbpsload_read_bytes_ratio = disk_data_read/load_out bound_mbps

Storage system - Disk Throughput

This view shows the details of the disk throughput. The data is averaged over all the disks in thestorage system.

The following panel is available under this view:

• Total disk Throughput in Kb/SecThis panel displays the disk_data_read, disk_data_written, and disk_throughput counters. Thedisk_throughput counter is calculated using the following formula:disk_throughput = disk_data_read + disk_data_written

Storage system - Top vFilers

This view shows the details of the top five vFiler units in the storage system.


• Top vFiler Units by CPU Utilization


This panel displays the vfiler_cpu_busy counter.• Top vFiler Units by Network Throughput

This panel displays the vfiler_nw_throughput counter.• Top vFiler Units by Disk Data Transferred

This panel displays the vfiler_data_transferred counter.• Top vFiler Units by Total Ops

This panel displays the vfiler_total_ops counter.

Top Aggregates

This view shows the details of the top five aggregates in the storage system.


• total_transfersThis panel displays the aggregate:total_transfers counter.

• user_readsThis panel displays the aggregate:user_reads counter.

• cp_readsThis panel displays the aggregate:cp_reads counter.

• user_writesThis panel displays the aggregate:user_writes counter.

Top LUNs

This view shows the details of the top five LUNs in the storage system.


• read_opsThis panel displays the lun:read_ops counter.

• other_opsThis panel displays the lun:other_ops counter.

• write_opsThis panel displays the lun:write_ops counter.

Top Volumes

This view shows the details of the top five volumes in the storage system.


• read_opsThis panel displays the volume:read_ops counter.

• other_opsThis panel displays the volume:total_ops counter.


• write_opsThis panel displays the volume:write_ops counter.

Canned views at the vFiler levelYou can see the views at the vFiler level to interpret the performance data.

vFiler Summary View

This view gives a summary information of the vFiler unit.


• Recent eventsCounters are not applicable for recent events.


• Top Volumes by OpsThis panel displays the volume:<instance>:total_ops counter.

• Top Qtrees by OpsThis panel displays the total_ops counter, which is calculated using the following formula:total_ops = qtree: <instance>:nfs_ops + qtree: <instance>:cifs_ops

• Top LUNs by OpsThis panel displays the lun:<instance>:total_ops counters.

• Top LUNs by LatencyThis panel displays the lun:<instance>:avg_latency counters.

vFiler Basic View

This view shows the details of different metrics for the vFiler unit.


• vfiler_read_opsThis panel displays the vfiler_read_ops counter.

• vfiler_write_opsThis panel displays the vfiler_write_ops counter.

• vfiler_misc_opsThis panel displays the vfiler_misc_ops counter.

• vfiler_cpu_busyThis panel displays the vfiler_cpu_busy counter.

• vfiler_read_bytesThis panel displays the vfiler_read_bytes counter.

• vfiler_write_bytesThis panel displays the vfiler_write_bytes counter.


• vfiler_net_data_recvThis panel displays the vfiler_net_data_recv counter.

• vfiler_net_data_sentThis panel displays the vfiler_net_data_sent counter.

Canned views at the processor levelYou can see the view at the processor level to interpret the performance data.

Processor Summary View

This view provides the summary detail of a processor.


• CPU UtilizationThis panel displays the processor_busy counter.

• Average CPU Busy per DomainThis panel displays the following counters:

• processor:domain_busy:idle• processor:domain_busy:kahuna• processor:domain_busy:storage• processor:domain_busy:exempt• processor:domain_busy:raid• processor:domain_busy:target• processor:domain_busy:netcache• processor:domain_busy:netcache2• processor:domain_busy:cifs• processor:domain_busy:wafl_exempt• processor:domain_busy:network

Canned views at the network interface levelYou can see the views at the network interface level to interpret the performance data.

Network Interface Summary View

This view provides the summary detail of a network interface card.



• collisionsThis panel displays the collisions counter.


• Throughput in Pkts/SecThis panel displays the ifnet:recv_packets, ifnet:send_packets, and ifnet:total_packets counters.The ifnet:total_packets counter is calculated using the following formula:ifnet:total_packets = ifnet:recv_packets + ifnet:send_packets

• Throughput in Bytes/SecThis panel displays the ifnet:recv_data, ifnet:send_data, and ifnet:total_data counters.The ifnet:total_data counter is calculated using the following formula:ifnet:total_data = ifnet:recv_data + ifnet:send_data

• Packet ErrorsThis panel displays the ifnet:recv_errors, ifnet:send_errors, and ifnet:total_errors counters.The ifnet:recv_errors counter is calculated using the following formula:ifnet:total_errors = ifnet:recv_errors + ifnet:send_errors

• Multicast PacketsThis panel displays the ifnet:send_mcasts, ifnet:recv_mcasts, and ifnet:total_mcasts counters.The ifnet:total_mcasts counter is calculated using the following formula:ifnet:total_mcasts = ifnet:send_mcasts + ifnet:recv_mcasts

Canned views at the dataset levelYou can see the views at dataset levels to interpret the performance data.

Dataset Summary View

This view shows the top five components in a dataset.



• Top Volumes by opsThis panel displays thevolume:<instance>:total_ops counter.

• Top Qtrees by OpsThis panel displays a counter, which is a sum of the qtree:<instance>:nfs_ops andqtree:<instance>:cifs_ops counters.

Canned views at the resource pool levelYou can see the views at the resource pool level to interpret the performance data.

Resource Pool Summary View

This view shows the top five components in a resource pool.



• Top Aggregates by Total TransfersThis panel displays the aggregate:<instance>:total_transfer counter.

• Top Disks by Disk BusyThis panel displays the disk:<instance>:disk_busy counter.

• Top Storage Systems by OpsThis panel displays the pa_total_ops counters.

Canned views at the volume levelYou can see the views at the volume level to interpret the performance data.

Volume Summary View

This view provides the summary detail of the volume.



• Top LUNs by OpsThis panel displays the lun:<instance>:total_ops counter.

• Top LUNs by Avg LatencyThis panel displays the lun:<instance>:avg_latency counter.

• Top Qtrees by OpsThis panel displays the total_ops counter, which is a sum of qtree:<instance>:nfs_ops andqtree:<instance>:cifs_ops.

• Overall Latency by OptypeThis panel displays the following counters:

• volume:<instance>:avg_latency• volume:<instance>:read_latency• volume:<instance>:write_latency• volume:<instance>:other_latency

• Overall Iops by OptypeThis panel displays the following counters:

• volume:<volume_name>:read_ops• volume:<volume_name>:write_ops• volume:<volume_name>:other_ops• volume:<volume_name>:total_ops

Volume Latency View

This view shows the details of the latencies for different protocols for the volume.



• Overall Latency by OptypeThis panel displays the following counters:

• volume:<instance>:avg_latency• volume:<instance>:read_latency• volume:<instance>:write_latency• volume:<instance>:other_latency

• NFS Latency by OptypeThis panel displays the following counters:

• volume:<instance>:nfs_read_latency• volume:<instance>:nfs_write_latency• volume:<instance>:nfs_other_latency• volume:<instance>:nfs_latency

• CIFS Latency by OptypeThis panel displays the following counters:

• volume:<instance>:cifs_read_latency• volume:<instance>:cifs_write_latency• volume:<instance>:cifs_other_latency• volume:<instance>:cifs_latency

• SAN Latency by OptypeThis panel displays the following counters:

• volume:<instance>:san_read_latency• volume:<instance>:san_write_latency• volume:<instance>:san_other_latency• volume:<instance>:san_latency

Volume IOPs View

This view shows the details of the iops for different protocols at the storage system level.


• Overall Iops by OptypeThis panel displays the following counters:

• volume:<volume_name>:read_ops• volume:<volume_name>:write_ops• volume:<volume_name>:other_ops• volume:<volume_name>:total_ops

• NFS Iops by OptypeThis panel displays the following counters:


• volume:<volume_name>:nfs_read_ops• volume:<volume_name>:nfs_write_ops• volume:<volume_name>:nfs_other_ops• volume:<volume_name>:nfs_ops

• CIFS Iops by OptypeThis panel displays the following counters:

• volume:<volume_name>:cifs_read_ops• volume:<volume_name>:cifs_write_ops• volume:<volume_name>:cifs_other_ops• volume:<volume_name>:cifs_ops

• SAN Iops by OptypeThis panel displays the following counters:

• volume:<volume_name>:san_read_ops• volume:<volume_name>:san_write_ops• volume:<volume_name>:san_other_ops• volume:<volume_name>:san_ops

• User Ops v/s Sys OpsThis panel displays the following counters:

• volume_user_ops• volume_sys_ops

The above counters are calculated using the following formula:

• volume_user_ops = sum(volume:<instance>:nfs_read_ops + nfs_write_ops + nfs_other_ops +cifs_read_ops + cifs_write_ops + cifs_other_ops + san_read_ops + san_write_ops +san_other_ops)

• volume_sys_ops = volume:<instance>:total_ops - volume_user_ops• Flexcache Iops by Optype

The above counters are calculated using the following formula:

• volume:<volume_name>:flexcache_read_ops• volume:<volume_name>:flexcache_write_ ops• volume:<volume_name>:flexcache_other_ ops

Canned views at the qtree levelYou can see the views at the qtree level to interpret the performance data.

Qtree Summary View

This view provides the summary details of the qtree.




• Throughput in OpsThis panel displays the following counters:

• qtree:<instance>:nfs_ops• qtree:<instance>:cifs_ops• qtree:<instance>:internal_ops

Canned views at the LUN levelYou can see the views at the LUN level to interpret the performance data.

LUN Summary View

This view provides the summary details of the LUN.



• Throughput in OpsThis panel displays the following counters:

• lun:<instance>:read_ops• lun:<instance>:write_ops• lun:<instance>:other_ops• lun:<instance>:total_ops

• Throughput in BpsThis panel displays the following counters:

• lun:<instance>:read_data• lun:<instance>:write_data• throughput

• Average LatencyThis panel displays the lun:<instance>:avg_latency counter.

LUN Alignment View

This view provides the alignment details of the LUN.


• Aligned WAFL OpsThis panel displays the following counters:

• read_align_histo


• write_align_histo• Unaligned WAFL Ops

This panel displays the following counters:

• unaligned_reads• unaligned_writes

• Partial blocksThis panel displays the following counters:

• read_partial_blocks• write_partial_blocks

Canned views at the aggregate levelYou can view the details of the aggregate to interpret the performance data.

Aggregate Summary View

This view provides the summary details of the aggregate.



• Top Disks by disk_busyThis panel displays the disk:<instance>:disk_busy counter.

• Throughput in Blocks/SecThis panel displays the following counters:

• aggregate:<instance>:user_read_blocks• aggregate:<instance>:user_write_blocks• aggregate:<instance>:cp_read_blocks• total_transfers

Canned views at the disk levelYou can see the views at the disk level to interpret the performance data.

Disk Summary view

This view provides the summary details of the disk.



• Throughput in Ops


This panel displays the following counters:

• disk:<instance>:total_transfers• disk:<instance>:user_reads• disk:<instance>:user_writes• disk:<instance>:cp_reads

• Throughput in BlocksThis panel displays the following counters:

• disk:<instance>:user_read_blocks• disk:<instance>:user_write_blocks• disk:<instance>:cp_read_blocks• throughput

• Disk LatencyThis panel displays the following counters:

• disk:<instance>:user_read_latency• disk:<instance>:user_write_latency• disk:<instance>:cp_read_latency

• Disk Chain InfoThis panel displays the following counters:

• disk:<instance>:user_read_chain• disk:<instance>:user_write_chain• disk:<instance>:cp_read_chain

• Disk UtilThis panel displays the disk:<instance>:disk_busy counter.

• Guarenteed OpsThis panel displays the guarenteed_reads and guarenteed_writes counters.

Canned view at the target levelYou can see a view at the target level to interpret the performance data.

Target Basic

This view provides the basic details at the target level.


• read_opsThis panel displays the read_ops counter.

• write_opsThis panel displays the write_ops counter.

• other_ops


This panel displays the other_ops counter.• queue_full

This panel displays the queue_full counter.


Alarms, hosts, and thresholds

You can use alarms, list of hosts, and thresholds in Performance Advisor to monitor datasets andresources, edit login credentials on hosts, and identify potential performance issues of storagesystems.

Next topics

What alarms are on page 97

Overview of alarm management tasks on page 98

How hosts become visible to Performance Advisor on page 103

What thresholds are on page 106

What alarms areAlarms are configured notifications that are sent whenever a specific event or an event of a specificseverity type occurs, not necessarily related to a specific user. Alarms are used to monitor andmanage datasets and resources, as a whole.

Alarms are not the events themselves, only the notification of events. Alarms are not the same as useralerts. For a complete description of user alerts, see the Operations Manager Administration Guide.

You can create alarms for any storage system for which you want automatic notification of events.You can use the Set Up Alarms window to add, monitor, and edit alarms.

List of alarm management tasksYou can respond to, view, edit, test, disable, enable, add, or delete data performance alarms.

Alarm managementtasks

This list describes the alarm management tasks that you perform mostoften.

Responding toalarms

When you receive an alarm, you should acknowledge the event and resolvethe condition that triggered the alarm. If the repeat notification feature isenabled and the alarm condition persists, you continue to receivenotifications until you acknowledge the event.

Viewing the list ofalarms

View the list of all currently configured data protection alarms when youwant an overview of what is configured for any selected group. You cancontrol the sort order of any column in the list. You can also control whichalarms are displayed by filtering the information in any column in the list.

Alarms, hosts, and thresholds | 97

Viewing theconfiguration valuesfor any selectedalarm

View the configuration details of any selected alarm when you want toknow specific information or you are considering modifying theconfiguration.

Editing the alarm Modify the configuration of an alarm when you need to accommodatechanges in the environment. For example, you might need to update arecipient e-mail address, add another recipient pager address, add arecipient script name, or change the hours during which the alarm is active.

Testing an alarm Test a new or modified alarm by having Protection Manager send a testmessage to all the recipients configured for the selected alarm.

Disabling orenabling an alarm

Disable an alarm when you need to stop its functioning for a while but wantto retain the alarm configuration. For example, if you have an alarmconfigured for the SnapVault Backup Failed event and you scheduleddowntime for maintenance on that host, you might disable the alarm duringthe planned downtime. When the host is up again, you can enable thealarm.

Adding an alarm Configure a new alarm when you want notification of a specific event orany event of a specific severity type for a group.

Deleting an alarm Delete an alarm only when you are sure that you no longer need notificationof the event or severity type occurring for the group.

Overview of alarm management tasksYou can use the Alarms window to perform all alarm monitoring and managing tasks. Whenresponding to an alarm, you can use the Events window to view details about the event.

Next topics

Monitoring alarms on page 99

Editing alarms on page 99

Testing alarms on page 100

Enabling or disabling alarms on page 100

Adding alarms on page 101

Deleting alarms on page 102

Responding to alarms on page 102


Monitoring alarmsYou can view a list of all the alarms configured for your application and the configuration values foreach one.

Before you begin


Steps

1. For licensed applications, from the navigation pane, click Notifications > Alarms. ForPerformance Advisor, from the navigation pane, click Set Up > Alarms.

2. (Optional) You can customize the Alarms window in any of the following ways:

• Select an alarm to see the configured property values for that alarm.• Click in a column header to control which alarm entries you want displayed. The color of

the column header changes to indicate that you are filtering the entries in that column.Some column filters display a drop-down list to select from and other column filters display asearch field in which you can enter text to select.

• Click the sort arrows in a column header to change the sort order of the entries in that column.•

Click in the upper-right corner of the list to select which columns are displayed.• Drag the bottom of the alarms list area up or down to resize that area.

Editing alarmsYou can change the configuration of an existing alarm.

Before you begin


Steps


2. Select an alarm in the Alarms window and click Edit.

3. Use the tabs in the Properties sheet (Group, Event Type, Recipient, and Details) to change theconfigured property values.

4. Click Apply to commit your changes.


5. Click OK to close the Properties sheet.

The new configuration is immediately activated and displayed in the Alarms list.

6. Verify your changes by viewing the results that are displayed in the Alarms window.

You can select an alarm to view the current property values for that alarm.

Testing alarmsYou can test an alarm to check its configuration, after creating or editing the alarm.

Before you begin


Steps


2. Select an alarm and click Test.

You can select any alarm regardless of whether it is enabled or disabled.

3. Click OK in the Confirm Alarm(s) Test window to begin the test.

A test event notification is sent to each configured recipient. If a script is configured, the testnotification runs the script.

4. Verify the alarm recipient configuration by checking that each recipient received a test message.If a script recipient is configured, check that the script was successfully started.

After you finish

You can modify any of the alarm properties from the Alarms window.

Enabling or disabling alarmsYou can disable an alarm when you need to temporarily stop its functioning and enable it when youwant it to start functioning again.

Before you begin


About this task

Disabling an alarm enables you to retain the alarm configuration while its functioning is temporarilystopped. For example, if you have an alarm that is configured for the SnapVault Backup Failed event


and you scheduled down time for maintenance on a host that is in the resource pool for the groupassociated with that alarm, you might disable the alarm during the planned down time. When the hostis up again, you can enable the alarm to resume its functioning.

Steps


2. In the alarms list, locate the entry for the alarm you want to enable or disable.

3. Clear the Enabled check box to disable the associated alarm, or select the Enabled check box toenable it.

The alarm is immediately disabled or enabled.

4. Verify your change by viewing the results that are displayed in the Alarms window.

Adding alarmsYou can add an alarm when you want immediate notification that a specified event or event class orevent of a specified severity level occurred.

Before you begin


Have the following information available to complete this task:

• The resource group with which you want the alarm associated.• The event name, event class, or severity type that triggers the alarm.• Who and what you want the event notification sent to.• The time period during which the alarm is active.• Whether you want the event notification repeated until the event is acknowledged and how often

the notification should be repeated.

Steps

1. For licensed applications, from the navigation pane, click Notifications > Alarms. ForPerformance Advisor, from the navigation pane, click the Set Up Alarms window.

2. Click Add to start the Add Alarm wizard.

3. On each page of the wizard, enter the appropriate information.

4. Click Finish to commit your choices and close the Add Alarm wizard.

5. Verify the creation and configuration of the alarm by viewing the results that are displayed in theAlarms window.


After you finish

You can edit the alarm properties from the Alarms window.

Deleting alarmsYou can delete an alarm when you no longer need immediate notification of an event or severitytype.

Before you begin


Steps


2. Select one or more alarms in the Alarms window and click Delete.

3. Click OK in the Confirm Alarm(s) Delete dialog box to delete the selected alarms.

The alarm configurations are deleted and the alarms are removed from the Alarms list.

4. Verify your deletion by viewing the results that are displayed in the Alarms window.

Responding to alarmsWhen you receive an alarm, you can acknowledge the event and resolve the condition that triggeredthe alarm.

Before you begin


About this task

If the repeat notification feature is not enabled, you receive the event notification only once. If therepeat notification feature is enabled and the alarm condition persists, you continue to receivenotifications until you acknowledge the event.

Steps


2. Select an event whose details you want to view.


3. Click Acknowledge to indicate that you are taking responsibility for managing that event.

4. Click Yes in the Acknowledge Events dialog box to acknowledge the selected event.

Your user name and the time are entered in the Events list for the selected event.

5. Find the cause of the event and take corrective action.

How hosts become visible to Performance AdvisorYou can use Performance Advisor to view a list of hosts discovered by DataFabric Manager. Youcan set, modify, and check the login credentials on hosts using the application.

From the navigation pane, you can click Set Up > Hosts to display the Set Up Hosts window. Bydefault, the first row is selected.

The DataFabric Manager server can start collecting data from the storage system discovered by it,only if the following conditions are set:

• The login credentials for the host are properly set.

• The transport protocol is properly set for the storage system discovered by DataFabric Manager.• The performance data collection option is set.

Prior to NetApp Management Console 2.1, users had to use DataFabric Manager to set, modify, andcheck the login credentials on hosts, and use Performance Advisor to monitor performance data.Using NetApp Management Console 2.1 or later, you can edit the credentials on hosts usingPerformance Advisor.

Next topics

List of fields in the hosts table on page 103

Editing host credentials on page 104

Configuration of data collection on page 104

Copying the data collection configuration on page 105

Effect of increasing sample rates on page 105

Setting retention periods and sample rates on page 105

List of fields in the hosts tableYou can select a host to display the host name, IP address, host status, user name, password, hosttransport, and other Performance Advisor details.

Field Description

Name The name of the host

Host Transport The protocol used to communicate


Field Description

Data Collection Enabled Check box to enable or disable data collection fromthe host

Login Credential Describes whether the user was authenticated on thehost

Status The status of the host

Editing host credentialsYou can set, modify, or check the host credentials using Performance Advisor.

Steps

1. From the navigation pane, click Set Up > Hosts.

2. Select a host, and then click Edit.

Alternatively, you can right-click the selected host, and then click Edit to display the HostProperties window.

The Host Properties window is displayed.

3. Enter the user name and password values in the Login Credentials section.

4. Select the Use hosts.equiv check box to use the hosts.equiv option.

For more information about the hosts.equiv option, see the Operations Manager AdministrationGuide.

5. Select the Enable Data Collection check box if you want to collect data from this specific host.

6. Select the transport protocol for data collection from the Host Transport list.

The transport protocol can be one of httpsOk, httpOnly, or Global Default.

7. Click Apply, and then click OK.

The credentials are updated in the DataFabric Manager database.

Configuration of data collectionYou can configure data collection by enabling or disabling counters through the Data CollectionConfiguration Wizard in the Set Up Hosts window.

You can also change the sample rate and retention period for counter groups using the wizard.


Copying the data collection configurationYou can copy the configuration relating to the performance data collection of one storage system andapply it to one or more than one storage system.

Before you begin

Before copying the configuration of a storage system, ensure that the Data Collection Status icon inthe hosts list for the storage system is green.

Steps

1. Click Set Up > Hosts.

2. Click Data Collection > Copy, and then click Next.

3. Select one or more available storage systems.

Note: If you select more than 10 storage systems, NetApp Management Console might hang.

4. Click Next, and then click Finish.

Effect of increasing sample ratesYou can lose the historical data if sample rates for data collection are increased. Because the purge inthe data samples collected is even, some space is reclaimed on the DataFabric Manager server.

For example, If you have 100 data samples collected for a week, when you double the samplinginterval, the number of data samples collected are halved to 50. Every second sample from thecollection is purged.

Setting retention periods and sample ratesYou can set retention periods and sample rates for data collection from various objects by using theData Collection Configuration Wizard. You can invoke the wizard from the Set Up Hosts window.

Steps

1. Click Set Up > Hosts.

2. Select the storage system.

3. Click Data Collection > Edit.

4. Set the retention period and sample rate in the Data Collection Configuration Wizard.

Note: If you decrease the retention period immediately after increasing the sample rate, theremight be a data loss.


What thresholds areThresholds are limits that are set on counters and associated with a set of objects. A threshold breachfor a specified interval of time raises events and alarms. They are used to identify potential blockagesin the performance of storage systems.

You can use thresholds to determine the point at which Performance Advisor should generate anevent. This point depends on the threshold interval and the counter value you set. If you set athreshold on the counter such as the average latency counter on any volume, then an event occurswhen the threshold breaches a particular counter value for an interval.

Any threshold applied to counters on a parent object is automatically applied to counters on any childobjects. If the parent-child object relationship is broken, the thresholds inherited from the parent nolonger apply to the child.

You can apply thresholds for two or more counters on one or more objects. The threshold interval forall thresholds in a combination remains same. For example, you can set thresholds for theavg_latency and total_ops counters within a volume. Events are generated only when all thethresholds are breached.

If you apply thresholds to objects with a parent-child relationship, such as the aggregate and diskobjects, any event is generated at the child object level. For example, you can set the threshold valuesfor the cp_read_blocks counter on aggregate and the disk_busy counter on disk. However, the eventis generated on the object lowest in the hierarchy.

You cannot apply thresholds in combination on the following pairs of objects:

• Network interface and targets• Aggregates and vFiler units• vFiler units and disks• Disks and LUNs• Disks and qtrees• Disks and volumes

You can set thresholds on counters specific to objects such as targets, interfaces, and disks. However,you cannot set thresholds on processors, RAID groups, and plexes.

Next topics

What resource properties are on page 107

Viewing threshold values on page 107

When to use thresholds in combination on page 107

What the Details tab is on page 107

Adding thresholds on page 108

Editing thresholds using the Summary tab on page 108

Editing thresholds using charts on page 109


Editing thresholds using the Details tab on page 109

Deleting thresholds on page 110

Creating alarms on page 110

What a threshold template is on page 110

What resource properties areResource properties are conditions such as disk type, disk roles, system models, and system versionsthat should be met while applying thresholds on the objects. By applying resource properties, you cangenerate events only when a threshold breach occurs on a particular model and a version of thestorage system.

For example, you can set a threshold only on FAS3050 storage systems in a group by setting thefilter details. However, the threshold must be rule-based.

You can use the Add Threshold Wizard to set the filter details for a threshold. You can display thefilter property and the corresponding value by selecting a value in the list.

Viewing threshold valuesYou can view existing threshold values by using the Summary tab.

Steps

1. Click Set Up > Thresholds.

2. Click the Summary tab.

When to use thresholds in combinationYou might not want to receive alerts for every small activity on an object. In such cases, you shouldconsider applying thresholds in combination by creating a rule on more than one counter for objectsin storage systems.

For example, if you set thresholds on the avg_latency counter for all LUNs in a storage system, youmight receive alarms from LUNs which are hardly used. To stop generating such events, you mustapply thresholds on more than one counter. In this case, you can create a rule to generate alerts onlywhen the total_ops counter and the avg_latency counter breach the threshold values.

What the Details tab isThe Details tab displays a page with the object tree on the left pane and the list of counters with thethreshold values on the right pane.

The objects in the object tree are of two types:

• Logical• Physical


Next topics

What logical objects are on page 108

What physical objects are on page 108

What logical objects are

Object types that represent storage containers such as volume, qtree, LUNs, vFiler units, and datasetsare known as logical objects.

From the navigation pane, you can click View to view the logical object types. The default isLogical.

What physical objects are

Object types that represent the physical resources in a storage system such as disks, aggregates,memory, network interfaces, resource pools, and RAID groups are known as physical objects.

From the navigation pane, you can click View to view the object types. You can click Physical toview the physical objects. The default is Logical.

Adding thresholdsYou can add thresholds for the performance counters on storage systems using the Set UpThresholds window. Alternatively, you can right-click the object in the logical or physical hierarchyto add thresholds.

Steps



3. Click Add.

You cannot add thresholds on the objects whose storage systems do not have valid credentials.

4. Complete the Add Thresholds wizard by entering the threshold values you want to set for yourperformance counters.

Editing thresholds using the Summary tabYou can edit the type, value and interval of thresholds and event name using the Summary tab in theSet Up Thresholds window. However, you cannot edit thresholds based on threshold templates.

Steps




3. Perform either of the following actions:

• Select the threshold that you want to modify, and click Edit.• Right-click the threshold that you want to modify, and click Edit.

4. Edit the threshold details.

5. Click Apply.

6. Click OK.

Editing thresholds using chartsYou can modify the threshold value and the threshold interval on charts.

Steps

1. Select Show Threshold on the chart.

2. Drag the dotted line.

However, you cannot edit thresholds if they are based on threshold templates or are set on parentobjects. In such cases, you cannot drag the line.

3. Edit the threshold details.

4. Click Apply.

5. Click OK.

Editing thresholds using the Details tabYou can modify the threshold value and the threshold interval for the selected threshold using theDetails tab in the Set Up Thresholds window.

About this task

Note: You cannot edit thresholds based on threshold templates or thresholds set on parent objects.

Steps


2. Click the Details tab.

3. Select the object.

4. Click the threshold in the Existing threshold box.

5. Edit the threshold value.

6. Edit the threshold interval.


7. Click Apply.

8. Click OK.

Deleting thresholdsYou can delete thresholds by using the Summary tab in the Set Up Thresholds window.

Steps



3. Perform either of the following actions:

• Select one or more thresholds that you want to delete, and click Delete.• Select one or more thresholds that you want to delete, and then right-click and click Delete.

4. Click OK.

Creating alarmsYou can create alarms by using the Add Alarm wizard.

Steps



3. Select the threshold.

4. Click Add Alarm.

5. Complete the Add Alarm wizard.

What a threshold template isA threshold template is a set of thresholds. You can apply a threshold template to one or moreapplicable objects in a storage system.

An object can be assigned to a template only if there is at least one applicable threshold in thetemplate. For example, you can assign a volume to a template with two thresholds: one set onsystem:cpu_busy and other set on volume:total_ops. You cannot assign a volume object if volumethreshold was not present in the template.

Threshold templates simplify threshold management. Using the templates, you can group multiplethresholds into a single unit and associate multiple objects to these thresholds.

You might want to generate a set of events for an object when it exceeds its counter thresholds. Insuch cases, you should consider creating a threshold template, which defines a set of thresholds, and


apply to the object. For example, you can create a threshold template that contains a threshold for thecp_writes counter greater than 10 per second for 5 minutes. You can apply these values in a templateto objects to generate events.

You can create and view templates using the Set Up Threshold Templates window, which isdivided into two halves. The top half of the Set Up Threshold Templates window displays thetemplate table, and the bottom half lists the events, counters, and resource properties. The object tabin the bottom-half of the page displays the objects that are applied to the selected template.

Next topics

When to use threshold templates on page 111

Creating threshold templates on page 111

Editing threshold templates on page 112

When to use threshold templates

You might want to use threshold templates to modify a threshold applied on more than one object.

For example, if you want to change the threshold limit from 90% to 80% for a threshold applied onsystem:cpu_busy counter for ten storage systems, you can modify all the ten thresholds by usingthreshold templates. Threshold templates allow you to modify the threshold on all the storagesystems with just one change in the template.

Using threshold templates, you can easily add or remove thresholds on objects. For example, insteadof creating ten identical thresholds for ten storage systems, you could just apply a template to all thestorage systems to create as many thresholds. Similarly, you can remove the objects from an existingthreshold using threshold templates.

Creating threshold templates

You can add threshold templates by using the Add Threshold Template wizard.

Steps

1. Click Set Up > Threshold Templates.

2. Click Add.

Note: You can add a threshold template by using the Set Up Thresholds window.

3. Complete the wizard to create a new threshold template.


Editing threshold templates

You can modify existing threshold templates by using the Set Up Threshold Templates window.

Steps

1. Click Set Up > Threshold Templates.

2. Click Edit.

3. Click Threshold Details to add, edit, or delete the thresholds in the template.

4. Click OK.


Glossary

associated view A performance view available for an object.

An associated view can be a canned view or a custom view. Canned views areassociated to an object based on the object type. Custom views can beassociated to an object by the user.

chart A graphical display of the data provided by one or more counters.

counter A statistical measurement of activity on a storage system or storage subsystemthat is provided by Data ONTAP.

custom views Custom views are user-defined views that includes metrics from specificinstances.

dataset A collection of storage sets along with configuration information associatedwith data. The storage sets associated with a dataset include a primary storageset used to export data to clients, and the set of replicas and archives that existon other storage sets. Datasets represent exportable user data.

hierarchicalgroups

The customizable grouping of storage systems and vFiler units whoseperformance can be monitored and displayed through Performance Advisor.All systems that are monitored by the DataFabric Manager server and whoseperformance can be monitored by Performance Advisor are displayed in theGroups browser panel of Performance Advisor.

historical data Data that is archived by the performance-monitoring server on the DataFabricManager server. All the data that is included in the Performance Advisordefault views is also archived as historical data. The historical data is accessibleto any Performance Advisor that can connect to the workstation. Historical datais collected on an ongoing basis, independent of whether a client has theassociated performance view open or not. Historical data can be used fordiagnosing past performance problems or for short-term trend analysis.

hosting storagesystem

The physical storage system on which one or more vFiler units are configured.Some counters that Performance Advisor tracks apply to both storage systemsand vFiler units. Other counters (for example, CPU usage) apply only tostorage systems and the associated host of a vFiler unit. The hosting storagesystem is also referred to as the “vFiler host.”

logical objects Object types that represent storage containers such as volume, qtree, LUNs,vFiler units, and datasets are known as Logical Objects.

logical hierarchy The hierarchy that displays the logical objects and instances.

Glossary | 113

lower threshold The type of threshold set for an event generation when the counter value fallsand remains below the lower threshold value for longer than the ThresholdInterval specified.

managed object A managed object represents any object that has an identity and a name in theDataFabric Manager object table. A managed object is an object that iscontained within a DataFabric Manager group. Volumes, aggregates, qtrees,and LUNs are examples of managed objects.

NetAppManagementConsole

NetApp Management Console is the client platform for the Java-based NetAppManageability Software applications. NetApp Management Console runs on aWindows or Linux workstation, separate from the system on which theDataFabric Manager server is installed.

object Typically there is an object associated with each hardware or softwaresubsystem within Data ONTAP. Examples of hardware objects are processor,disk, NVRAM, and networking card objects. FCP, iSCSI, CIFS, and NFS arethe examples of software protocol objects. WAFL, RAID and target are theexamples of internal objects specific to Data ONTAP. Virtual objects like thesystem object capture key statistics across all the other objects in one singleplace. Examples of system objects are avg_processor_busy, nfs_ops, cifs_ops,and net_data_recv.

PerformanceAdvisor

The Performance Advisor component installed on the NetApp ManagementConsole platform enables you to monitor the performance of storage systemsand vFiler units.

The user interface of Performance Advisor contains only performance-monitoring information. This Performance Advisor interface is distinct fromOperations Manager, which contains other DataFabric Manager information.

performance-monitoringserver

The Performance Advisor component that is enabled on the DataFabricManager server to collect and archive performance data at regular intervalsfrom the monitored storage systems and vFiler units. Performance Advisorgathers sets of data that the performance-monitoring server has collected togenerate its graphical charts and views.

performanceview

A collection of one or more counters accessible through Performance Advisorfor a group of storage systems or subsystems through one or more charts.Performance Advisor displays canned views that are built in to the system, andcustom views.

physical objects Object types that represent the physical resources in a storage system such asdisk, aggregates, memory, network interfaces, resource pools, and RAIDgroups are known as Physical Objects.

physicalhierarchy

The hierarchy that displays the physical objects and instances.


real-time data Data that is passed through Performance Advisor for display but is not stored.Real-time data is collected only as long as the performance view windowdisplaying it is open on Performance Advisor. Real-time data is accessible onlyto the client by which it is retrieved.

Real-time data is suitable for diagnosing immediate performance issues.

resource pool A managed object in DataFabric Manager, containing storage provisioningresources like storage systems, aggregates, and spare disks.

retention period The amount of previously collected data for a counter that is accessible to theuser. You can set the retention period based on units of a minute, hour, day,week, month, or year.

sample rate The rate at which data is collected for a counter. You can set the sample ratebased on units of a minute.

storage set Containers that are used for delegation, replication, and in some cases,substorage provisioning. The only container of merit in a storage set is avolume (flexible or traditional). A storage set contains a group of volumeswhereas a volume should be in at most one storage set.

storage system An appliance that is attached to a computer network and is used for datastorage. FAS appliances and NearStore systems are examples of storagesystems.

templates andcanned templates

A template is a view definition that applies to object instances known toPerformance Advisor.

A set of pre-configured templates in Performance Advisor is known as CannedTemplate.

thresholdinterval

The amount of time in seconds for which an event generation is suppressedbefore Performance Advisor decides that a counter has crossed a specifiedthreshold and an event needs to be generated. The same interval is also be usedto generate a normal event.

unmanagedobject

Objects apart from the managed objects belong to the class of unmanagedobjects. An unmanaged object does not have a unique identity in theDataFabric Manager table.

upper threshold The type of threshold set for an event generation when the counter valueexceeds and remains above the higher threshold value for longer than theThreshold Interval specified.

vFiler unit One or more virtual storage systems that can be configured on a single physicalstorage system licensed for the MultiStore feature. DataFabric Manager 3.4 andlater enables monitoring and management of vFiler units.

view A collection of related panels represented together and displayed by thePerformance Advisor client.

Glossary | 115

IndexA

access permissionsoverview 45

Add Alarm wizardtask 101

administrator rolesrequirements for Performance Advisor tasks 25

advanced viewscreating 70

alarmsadding 101deleting 102disabling 100enabling 100list of monitoring and managing tasks for 98modifying 99monitoring 99pausing 100responding to 102testing 100viewing 99

applications in NetApp Management Console 27

C

canned views 64charts

bar chart 68changing the time period 69changing the type 71polyline 68resetting axes 69saving 78scatter 68

client platforms 27console

adjusting column widths 30applications supported 27applications, toggling 29filtering data display 31Get Started feature 29Help, accessing 29logging out 29moving backward and forward 29rearranging column order 30

selecting columns to display 30Tasks bar, hiding and redisplaying 30

countersabout 34shared across views 74

Custom viewmultiple axis 74purpose 73when to create 74

custom viewsabout 64creating 75deleting 76editing 76flexible 74opening 76

D

dashboardsdescribed 59panels, descriptions of 59

dashboards, performancedescriptions of panels 59

dashboards, protectiondescriptions of 59

dashboards, provisioningdescriptions of 59

data collectionconfiguration 104copying configuration 105

DataFabric Managerclient platform for 27

E

eventsdefinition of 60deleting from event log 62determining when they occur 60responding when an event occurs 61viewing the list of 61

F

filtering data display 31

Index | 117

G

Get Started feature 29

H

Helpaccessing 29

hierarchical groupssupport for 24

host credentialsediting 104

L

licensesNetApp Management Console applications 27, 28

limitations, exporting data 79limitations, Performance Advisor 24logging out 29logs

for viewing the list of events 61

M

monitoringdashboard panels, overview of 59dashboards, overview of 59events 61

N

NetApp Management Consoleabout 27applications supported 27defined 27licenses 28

See also console

P

Performance Advisoraccessing 55application described 27hosts in 103location in the network topology 23log in to 24

overview 23requirement for starting 51secure connections for 52starting 55stopping 56

performance counterslist of 35, 37–42

Performance dashboard 59performance data

exporting 78methods for preserving 77viewing 67

performance issuesdiagnosing 73diagnosis 72

performance monitoring serverlocation in the network topology 23

performance objectsabout 33list of 33

performance viewsworking with data in 68

performance-monitoring serverenabling 51

R

real-time dataviewing 75

regular expressions, filtering examples 31

S

sample buffer, setting 105sample rates

Increasing 105Secure connections

support for 25SSL

certificate for 25storage systems

CPU utilization 77CPU utilization limitations 24display unavailable 56refreshing the list 57

synchronizecharts across all views 69charts in a view 69undo 70


T

Tasks bar, hiding and redisplaying 30thresholds

about 106add 71adding 108deleting 110editing 108logical hierarchy 108physical hierarchy 108view 70

Thresholdsadding 111adding alarms 110Details tab 107editing 109, 112

transport protocols 53

U

unavailable menu options 56

V

vFiler unitsdisplay unavailable 56refreshing the list 57support for 25

viewsadvanced 65aggregate 94blank charts 57canned 79chart axes 74dataset 89disk 94displaying 66event correlation 72global 80, 81group 80, 81how to select 67LUN 93navigation 66network interface 88processor 88qtree 92resource pool 89storage system 82–86target 95vFiler unit 87volume 90, 91

Z

zero-value datadisplay of 56

Index | 119

Advisor

Documents

performance advisor

performance advisor

performance advisor

performancemonitoring

list of performance

list of performance

netapp management console

support telephone