vfiler

Data ONTAP® 7.3MultiStore Management Guide

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.com

Part number 210-04163_A0June 2008

Contents

Copyright information...................................................................................9Trademark information...............................................................................13About this guide............................................................................................15

Audience......................................................................................................................15

Accessing Data ONTAP man pages............................................................................15

Terminology.................................................................................................................16

FilerView as an alternative to the command-line interface.........................................18

Command, keyboard, and typographic conventions....................................................18

Special messages.........................................................................................................19

What MultiStore is.......................................................................................21Reasons for using MultiStore......................................................................................21

MultiStore for consolidating servers...........................................................................21

MultiStore for service providers and enterprises.........................................................22

MultiStore for data migration......................................................................................23

What the host storage system is...................................................................................23

What the default vFiler unit is.....................................................................................23

Some limitations of MultiStore...................................................................................24

Host storage system's access to vFiler unit data..........................................................24

Types of system administration tasks..........................................................................25

Host storage system tasks................................................................................25

MultiStore management tasks.........................................................................25

Tasks related to using Data ONTAP on vFiler units........................................26

MultiStore Management..............................................................................27How to create a vFiler unit .........................................................................................28

Enabling the MultiStore license......................................................................28

Disabling the MultiStore license.....................................................................29

Storage guidelines ...........................................................................................29

Language guidelines .......................................................................................30

Quota guidelines .............................................................................................30

Active/active considerations ...........................................................................30

SAN considerations ........................................................................................31

The vFiler commands .....................................................................................31

Table of Contents | 3

Ensuring that the network interface is ready ..................................................32

Assigning and configuring vFiler unit resources.............................................33

Manually setting up a vFiler unit.....................................................................34

How to manage vFiler unit storage from the storage system......................................35

Effects of adding, removing, and moving vFiler unit resources......................36

Adding resources to a vFiler unit....................................................................36

Requirements for moving and removing resources.........................................36

Removing resources from a vFiler unit...........................................................37

Moving resources between vFiler units ..........................................................37

Destroying a vFiler unit...................................................................................38

Restoring a vFiler unit.....................................................................................39

Default vFiler unit limits.............................................................................................39

Maximum vFiler units allowed....................................................................................39

Viewing the current vFiler unit limit...........................................................................41

Increasing the vFiler unit limit....................................................................................41

Decreasing the vFiler unit limit...................................................................................42

Renaming a vFiler unit................................................................................................42

Stopping a vFiler unit..................................................................................................43

Starting a vFiler unit....................................................................................................44

Allowing a protocol for a vFiler unit...........................................................................44

Disallowing a protocol for a vFiler unit......................................................................45

Displaying vFiler unit status........................................................................................46

Viewing commands that can be executed from a vFiler unit.......................................46

Executing commands from a vFiler unit.....................................................................47

Executing commands for a vFiler unit from the host storage system..........................47

Executing commands for a vFiler unit using RSH......................................................48

RSH commands...............................................................................................48

Executing commands for a vFiler unit using SSH.......................................................49

SSH commands................................................................................................49

Effects of storage system reboot on a vFiler unit........................................................50

Volumes and qtrees on a vFiler unit............................................................................50

Effects of taking a vFiler unit volume offline..................................................51

Changes required after volumes are renamed..................................................51

Who can change qtree security styles and oplock settings..............................51

Differences in qtree command output..............................................................52

Viewing all qtrees and the owner vFiler units ................................................52

4 | Data ONTAP 7.3 MultiStore Management Guide

Backups for vFiler units .............................................................................................52

NDMP support.................................................................................................53

LUNs on a vFiler unit..................................................................................................55

What is supported on a vFiler unit...................................................................55

iSCSI LUNs and igroups on a vFiler unit........................................................55

The iSCSI service on a vFiler unit...................................................................56

Networking considerations..........................................................................................57

Effects of a routed daemon..............................................................................57

Command for changing the routing table in the default IPspace.....................58

The /etc/dgateways file....................................................................................58

IPsec on a vFiler unit.......................................................................................58

SnapMirror on the host storage system.......................................................................58

Considerations for using the snapmirror command.........................................59

Determining the status of SnapMirror relationships........................................60

SnapVault on the host storage system..........................................................................60

Where to enter SnapVault commands..............................................................60

Considerations for using the snapvault command...........................................60

Determining the status of SnapVault relationships..........................................61

SNMP used with a vFiler unit.....................................................................................61

vFiler unit data from MIBs..............................................................................61

What is not supported on a vFiler unit........................................................................62

Monitoring performance and statistics .......................................................................62

Displaying storage system statistics................................................................62

Displaying uptime statistics.............................................................................63

Displaying NFS statistics.................................................................................63

Displaying CIFS statistics...............................................................................63

What IPspaces are .......................................................................................65What to consider for a vFiler unit participation in an IPspace....................................65

Application of IPspaces...............................................................................................66

How interfaces participate in an IPspace.....................................................................67

Routing in an IPspace..................................................................................................68

Loopback interfaces in IPspaces..................................................................................68

Advantages of using VLAN tagging for IPspaces.......................................................68

Traffic separation.............................................................................................68

More IPspaces than interfaces are allowed......................................................69

Secure delivery of packets to a vFiler unit in an IPspace................................69


Active/active configuration and IPspaces....................................................................69

What the IPspace naming requirement is........................................................69

IPspace assignment requirement.....................................................................69

What an asymmetric active/active setup is......................................................70

Specifying partners in an asymmetric active/active setup...............................70

Creating an IPspace.....................................................................................................71

Listing IPspaces on a storage system..........................................................................72

Removing an IP address from an interface..................................................................72

Assigning an interface to an IPspace...........................................................................73

Destroying IPspaces.....................................................................................................73

Creating a vFiler unit in a nondefault IPspace.............................................................74

File Access Using NFS and CIFS.................................................................77How to access the vFiler unit's file system with NFS and CIFS.................................77

How to access a file system with CIFS...........................................................77

How to access a file system with NFS.............................................................77

How to specify path names for NFS exports or CIFS shares......................................77

How to prepare the vFiler unit for NFS.......................................................................78

Starting the NFS protocol................................................................................78

Exporting all file systems in /etc/exports.........................................................79

How to prepare the vFiler unit for CIFS......................................................................79

Tasks that can be performed only on the host storage system.........................80

Per vFiler unit limit with CIFS........................................................................80

Local user accounts for vFiler units................................................................80

Disk Space Management Using Quotas......................................................81How to manage quotas ................................................................................................81

Allowing or disallowing quotas for a volume..................................................81

Quota specification management.....................................................................82

Turning quotas on or off from a vFiler unit.....................................................83

When quota thresholds and soft quotas are exceeded.................................................83

How you can resize quotas..........................................................................................84

How the quotas file works...........................................................................................84

Displaying quota status................................................................................................84

Displaying a quota report............................................................................................85

Disaster recovery and data migration.........................................................87How to prepare for a disaster.......................................................................................87

How to prepare the destination storage system...............................................87


Creating a disaster-recovery vFiler unit...........................................................95

What the vFiler dr configure command does..................................................97

Responding to a disaster..............................................................................................98

Activating the disaster-recovery vFiler unit.....................................................98

What activating the disaster-recovery vFiler unit does....................................99

How to reactivate the original vFiler unit after damage repair........................99

Deleting the disaster-recovery vFiler unit.................................................................109

How to move (migrate) a vFiler unit.........................................................................109

Prerequisites for migrating a vFiler unit by copying data.............................110

How migrating a vFiler unit affects clients...................................................110

What the vFiler unit migrate commands do..................................................111

Migrating thevFiler unit by copying data......................................................112

Adjusting client and network configurations if migrating

to a different subnet.................................................................................113

What SnapMover vFiler unit migration is.....................................................114

Requirements for vFiler unit migration between

active/active nodes...................................................................................115

Guidelines for setting up volumes to support vFiler unit migration..............115

What SnapMover vFiler unit migration does................................................116

How to migrate the vFiler unit between active/active

nodes with SnapMover ............................................................................116

Disabling SnapMover vFiler unit migration..................................................118

Virus protection for CIFS..........................................................................121Who can configure virus scanning............................................................................121

Storage systems with which virus scanners can be registered...................................121

Virus scanning for vfiler0..........................................................................................121

Requirements for virus scanning on a vFiler unit other than vfiler0.........................122

Effect of virus scanner availability on CIFS access...................................................122

Configuring virus scanning for a vFiler unit.............................................................122

Index.............................................................................................................125


Copyright information

Copyright © 1994–2008 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 an electronic retrievalsystem—without prior written permission of the copyright owner.

Portions of this product are derived from the Berkeley Net2 release and the 4.4-Lite-2 release, whichare copyrighted and publicly distributed by The Regents of the University of California.

Copyright © 1980–1995 The Regents of the University of California. All rights reserved.

Portions of this product are derived from NetBSD, copyright © Carnegie Mellon University.

Copyright © 1994, 1995 Carnegie Mellon University. All rights reserved. Author Chris G. Demetriou.

Permission to use, copy, modify, and distribute this software and its documentation is hereby granted,provided that both the copyright notice and its permission notice appear in all copies of the software,derivative works or modified versions, and any portions thereof, and that both notices appear insupporting documentation.

CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS “AS IS” CONDITION.CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND FOR ANY DAMAGESWHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.

Software derived from copyrighted material of The Regents of the University of California and CarnegieMellon University is subject to the following license and disclaimer:

Redistribution and use in source and binary forms, with or without modification, are permitted providedthat the following conditions are met:

Redistributions of source code must retain the above copyright notices, this list of conditions, and thefollowing disclaimer.

Redistributions in binary form must reproduce the above copyright notices, this list of conditions, andthe following disclaimer in the documentation and/or other materials provided with the distribution.

All advertising materials mentioning features or use of this software must display this text:

This product includes software developed by the University of California, Berkeley and its contributors.

Neither the name of the University nor the names of its contributors may be used to endorse or promoteproducts derived from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS “AS IS” AND ANYEXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIEDWARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE AREDISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FORANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL

Copyright information | 9

DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODSOR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVERCAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICTLIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAYOUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCHDAMAGE.

This software contains materials from third parties licensed to NetApp Inc. which is sublicensed, andnot sold, and title to such material is not passed to the end user. All rights reserved by the licensors.You shall not sublicense or permit timesharing, rental, facility management or service bureau usage ofthe Software.

Portions developed by the Apache Software Foundation (http://www.apache.org/). Copyright © 1999The Apache Software Foundation.

Portions Copyright © 1995–1998, Jean-loup Gailly and Mark Adler

Portions Copyright © 2001, Sitraka Inc.

Portions Copyright © 2001, iAnywhere Solutions

Portions Copyright © 2001, i-net software GmbH

Portions Copyright © 1995 University of Southern California. All rights reserved.

Redistribution and use in source and binary forms are permitted provided that the above copyrightnotice and this paragraph are duplicated in all such forms and that any documentation, advertisingmaterials, and other materials related to such distribution and use acknowledge that the software wasdeveloped by the University of Southern California, Information Sciences Institute. The name of theUniversity may not be used to endorse or promote products derived from this software without specificprior written permission.

Portions of this product are derived from version 2.4.11 of the libxml2 library, which is copyrightedby the World Wide Web Consortium.

NetApp modified the libxml2 software on December 6, 2001, to enable it to compile cleanly on Windows,Solaris, and Linux. The changes have been sent to the maintainers of libxml2. The unmodified libxml2software can be downloaded from http://www.xmlsoft.org/.

Copyright © 1994–2002 World Wide Web Consortium, (Massachusetts Institute of Technology, InstitutNational de Recherche en Informatique et en Automatique, Keio University). All Rights Reserved.http://www.w3.org/Consortium/Legal/

Software derived from copyrighted material of the World Wide Web Consortium is subject to thefollowing license and disclaimer:

Permission to use, copy, modify, and distribute this software and its documentation, with or withoutmodification, for any purpose and without fee or royalty is hereby granted, provided that you includethe following on ALL copies of the software and documentation or portions thereof, includingmodifications, that you make:

The full text of this NOTICE in a location viewable to users of the redistributed or derivative work.


Any pre-existing intellectual property disclaimers, notices, or terms and conditions. If none exist, ashort notice of the following form (hypertext is preferred, text is permitted) should be used within thebody of any redistributed or derivative code: “Copyright © [$date-of-software] World Wide WebConsortium, (Massachusetts Institute of Technology, Institut National de Recherche en Informatiqueet en Automatique, Keio University). All Rights Reserved. http://www.w3.org/Consortium/Legal/”

Notice of any changes or modifications to the W3C files, including the date changes were made.

THIS SOFTWARE AND DOCUMENTATION IS PROVIDED “AS IS,” AND COPYRIGHTHOLDERS MAKE NO REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED,INCLUDING BUT NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY OR FITNESSFOR ANY PARTICULAR PURPOSE OR THAT THE USE OF THE SOFTWARE ORDOCUMENTATION WILL NOT INFRINGE ANY THIRD PARTY PATENTS, COPYRIGHTS,TRADEMARKS OR OTHER RIGHTS.

COPYRIGHT HOLDERS WILL NOT BE LIABLE FOR ANY DIRECT, INDIRECT, SPECIAL ORCONSEQUENTIAL DAMAGES ARISING OUT OF ANY USE OF THE SOFTWARE ORDOCUMENTATION.

The name and trademarks of copyright holders may NOT be used in advertising or publicity pertainingto the software without specific, written prior permission. Title to copyright in this software and anyassociated documentation will at all times remain with copyright holders.

Software derived from copyrighted material of NetApp, Inc. is subject to the following license anddisclaimer:

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, exceptas expressly agreed to in writing by NetApp. The use or purchase of this product does not convey alicense under any patent rights, trademark rights, or any other intellectual property rights of NetApp.

The product described in this manual may be protected by one or more U.S.A. patents, foreign patents,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 Computer Softwareclause at DFARS 252.277-7103 (October 1988) and FAR 52-227-19 (June 1987).

Copyright information | 11

Trademark information

All applicable trademark attribution is listed here.

NetApp, the Network Appliance logo, the bolt design, NetApp-the Network Appliance Company,Cryptainer, Cryptoshred, DataFabric, DataFort, Data ONTAP, Decru, FAServer, FilerView, FlexClone,FlexVol, Manage ONTAP, MultiStore, NearStore, NetCache, NOW NetApp on the Web, SANscreen,SecureShare, SnapDrive, SnapLock, SnapManager, SnapMirror, SnapMover, SnapRestore,SnapValidator, SnapVault, Spinnaker Networks, SpinCluster, SpinFS, SpinHA, SpinMove, SpinServer,StoreVault, SyncMirror, Topio, VFM, and WAFL are registered trademarks of NetApp, Inc. in theU.S.A. and/or other countries. gFiler, Network Appliance, SnapCopy, Snapshot, and The evolution ofstorage are trademarks of NetApp, Inc. in the U.S.A. and/or other countries and registered trademarksin some other countries. The NetApp arch logo; the StoreVault logo; ApplianceWatch; BareMetal;Camera-to-Viewer; ComplianceClock; ComplianceJournal; ContentDirector; ContentFabric; EdgeFiler;FlexShare; FPolicy; Go Further, Faster; HyperSAN; InfoFabric; Lifetime Key Management, LockVault;NOW; ONTAPI; OpenKey, RAID-DP; ReplicatorX; RoboCache; RoboFiler; SecureAdmin; ServingData by Design; SharedStorage; Simplicore; Simulate ONTAP; Smart SAN; SnapCache; SnapDirector;SnapFilter; SnapMigrator; SnapSuite; SohoFiler; SpinMirror; SpinRestore; SpinShot; SpinStor; vFiler;VFM Virtual File Manager; VPolicy; and Web Filer are trademarks of NetApp, Inc. in the U.S.A. andother countries. NetApp Availability Assurance and NetApp ProTech Expert are service marks ofNetApp, Inc. in the U.S.A.

IBM, the IBM logo, AIX, and System Storage are trademarks and/or registered trademarks ofInternational Business Machines Corporation.

Apple is a registered trademark and QuickTime is a trademark of Apple, Inc. in the U.S.A. and/or othercountries. 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, and SureStream aretrademarks 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 iscertified RealSystem compatible.

Trademark information | 13

About this guide

This guide describes how to administer vFiler units with the optional MultiStore software available bylicense with Data ONTAP.

Next topics

Audience on page 15

Accessing Data ONTAP man pages on page 15

Terminology on page 16

FilerView as an alternative to the command-line interface on page 18

Command, keyboard, and typographic conventions on page 18

Special messages on page 19

AudienceThis guide is for system administrators who are familiar with operating systems that run on the storagesystem’s clients, such as UNIX, Windows 95, Windows NT, and Windows 2000.

You should be familiar with how to configure the storage system and how the NFS, CIFS, and HTTPprotocols are used for file sharing or transfers. This guide does not cover basic system or networkadministration topics, such as IP addressing, routing, and network topology; it emphasizes thecharacteristics of the storage system.

Accessing Data ONTAP man pagesYou can use the Data ONTAP manual (man) pages to access technical information.

Considerations

Data ONTAP manual pages are available for the following types of information. They are grouped intosections according to standard UNIX naming conventions.

Man page sectionTypes of information

1Commands

4Special files

About this guide | 15

Man page sectionTypes of information

5File formats and conventions

8System management and services

Step

1. View man pages in the following ways:

• By entering the following command at the storage system command line:

man command_or_file_name

• By clicking the manual pages button on the main Data ONTAP navigational page in theFilerView® user interface

• By using the Commands: Manual Page Reference, Volumes 1 and 2 (which can be downloadedor ordered through the NOW NetApp on the Web® site)

Note: All Data ONTAP man pages are stored in the storage system in files whose names areprefixed with the string "na_" to distinguish them from client man pages. The prefixed names areused to distinguish storage system man pages from other man pages and sometimes appear in theNAME field of the man page, but the prefixes are not part of the command, file, or services.

TerminologyTo understand the concepts in this document, you might need to know the terms defined here.

General storage system terminology

• Storage systems that run Data ONTAP are sometimes referred to as filers, appliances, storageappliances, or systems. The name of the FilerView graphical user interface for Data ONTAP reflectsone of these common usages.

• Controller or storage controller refers to the component of a storage system that runs the DataONTAP operating system and controls its disk subsystem. Controllers or storage controllers arealso sometimes called storage appliances, appliances, storage engines, heads, CPU modules, orcontroller modules.

Active/active configuration terminology

• An active/active configuration is a pair of storage systems configured to serve data for each otherif one of the two systems becomes impaired. In Data ONTAP documentation and other information


resources, active/active configurations are sometimes also referred to as clusters or active/activepairs.

• When in an active/active configuration, systems are often called nodes. One node is sometimescalled the local node, and the other node is called the partner node or remote node.

• Standard active/active configuration refers to a configuration set up so that one node automaticallytakes over for its partner when the partner node becomes impaired.

• Mirrored active/active configuration is similar to the standard active/active configuration, exceptthat there are two copies, or plexes, of the data. This is also called data mirroring.

• Fabric-attached MetroCluster refers to an active/active configuration running the syncmirror_localand cluster_remote licenses, where the nodes are attached to two pairs of Fibre Channel switches,and they are separated by more than 500 meters.

• Stretch MetroCluster refers to an active/active configuration running the syncmirror_local andcluster_remote licenses, where the nodes are separated by up to 500 meters, and no switches areused between the nodes. This configuration is also sometimes called a nonswitched MetroCluster.

• Controller failover, also referred to as cluster failover or CFO, refers to the technology that enablestwo storage systems to take over each other's data, thus improving data availability.

• Remote storage refers to the storage that is accessible to the local node, but is at the location of theremote node.

Storage hardware terminology

• FC HBA for Disk or FC HBA refers to the Fibre Channel host bus adapter that connects the nodeto the switch or to the disks.

• Disk shelf refers to a unit of the disk subsystem component of the storage system.

• LRC (Loop Resiliency Circuit) disk shelf module refers to a component that keeps the FibreChannel-Arbitrated Loop (FC-AL) intact during the addition and removal of disks within a diskshelf. It also contains the enclosure services processor, which communicates the environmental dataof the disk shelf.

• ESH (Embedded Switching Hub) disk shelf module refers to a component that provides a means ofmanaging an FC-AL loop in an intelligent manner, such that a single drive failure does not takedown the loop. It also contains the enclosure services processor, which communicates theenvironmental data of the disk shelf.

• ESH2 disk shelf module refers to a second-generation ESH module.

• ESH4 disk shelf module refers to a third-generation ESH module.

• AT-FCX refers to an enhanced FC-AL to Serial ATA (SATA) bridge used in some disk shelves.

General terms

• The term type means pressing one or more keys on the keyboard.

• The term enter mean pressing one or more keys on the keyboard and then pressing the Enter key,or clicking in a field in a graphical interface and typing information into it.


FilerView as an alternative to the command-line interface

Use the FilerView graphical user interface to perform many common tasks, as well as to view andmanage a storage system from a Web browser.

Your tasks as a Data ONTAP administrator can be performed by entering commands at the storagesystem console, in configuration files, or through a Telnet session or Remote Shell connection.

Another method of performing many common tasks is to use FilerView. FilerView comes with everystorage system, is easy to use, and includes Help that explains Data ONTAP features and how to workwith them in FilerView.

For more information about accessing a storage system with FilerView, and about FilerView Help, seethe System Administration Guide.

Command, keyboard, and typographic conventionsThis document uses command, keyboard, and typographic conventions that help you enter commands.

Command conventions

In examples that illustrate commands executed on a UNIX workstation, the command syntax and outputmight differ, depending on your version of UNIX.

Keyboard conventions

• When describing key combinations, this document uses the hyphen (-) to separate individual keys.For example, "Ctrl-D" means pressing the "Control" and "D" keys simultaneously.

• This document uses the term "Enter" to refer to the key that generates a carriage return, althoughthe key is named "Return" on some keyboards.

Typographic conventions

The following table describes typographic conventions used in this document.


Type of informationConvention

Words or characters that require special attention.

Placeholders for information you must supply. For example, if the guide says to enterthe arp -d hostname command, you enter the characters "arp -d" followed bythe actual name of the host.

Book titles in cross-references.

Italic font

Command names, option names, keywords, and daemon names.

Information displayed on the system console or other computer monitors.

The contents of files.

Monospaced font

Words or characters you type. What you type is always shown in lowercase letters,unless you must type it in uppercase letters.

Bold monospaced

font

Special messagesThis document might contain the following types of messages to alert you to conditions you need tobe aware of. Danger notices and caution notices only appear in hardware documentation, whereapplicable.

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

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

Danger: A danger notice warns you of conditions or procedures that can result in death or severepersonal injury.

Caution: A caution notice warns you of conditions or procedures that can cause personal injury thatis neither lethal nor extremely hazardous.


What MultiStore is

MultiStore is a software product you can use optionally. MultiStore enables you to partition the storageand network resources of a single storage system so that it appears as multiple storage systems on thenetwork. Each "storage system" created as a result of the partitioning is called a vFiler unit. A vFilerunit, using the resources assigned, delivers file services to its clients as a storage system does.

The storage resource assigned to a vFiler unit can be one or more qtrees or volumes. The networkresource assigned can be one or more base IP addresses or IP aliases associated with network interfaces.You can add or remove resources at any time.

Next topics

Reasons for using MultiStore on page 21

MultiStore for consolidating servers on page 21

MultiStore for service providers and enterprises on page 22

MultiStore for data migration on page 23

What the host storage system is on page 23

What the default vFiler unit is on page 23

Some limitations of MultiStore on page 24

Host storage system's access to vFiler unit data on page 24

Types of system administration tasks on page 25

Reasons for using MultiStoreMultiStore has many features that you can use for your provisioning needs.

• To consolidate multiple servers to one storage system.

• To use the storage system to host data for multiple customers, such as clients of a service provideror different organizations within an enterprise.

• To use the SnapMirror technology for migrating data from one storage system to anothertransparently.

MultiStore for consolidating serversIf you manage multiple servers, you might want to store all data on one storage system for easieradministration.

What MultiStore is | 21

To consolidate the servers, you can partition the storage system into vFiler units and then copy the datafrom the servers to the vFiler units. If the vFiler units are for CIFS users, you can set up the vFiler unitsto use the same computer names as the servers. This enables CIFS clients to share resources withouthaving to remap their drives or search for the new server in Network Neighborhood. If the vFiler unitsare for NFS users, the NFS clients might need to remount the file systems, or they can use theautomounter to automatically mount the file systems from the new locations.

A storage system without MultiStore can participate in only one security domain, so if your environmentrequires that different groups of CIFS users be in different domains, you must use multiple storagesystems. MultiStore, however, enables you to install each vFiler unit in the appropriate domain whilekeeping all of the data on the same physical storage system.

Because you can set up NIS and DNS servers for individual vFiler units, after you consolidate theservers on one storage system, network clients of the vFiler units can continue to use the same NIS andDNS servers as before.

MultiStore for service providers and enterprisesService providers, such as ISPs and SSPs, can partition the resources of a storage system to create manyvFiler units for client companies. Similarly, the information technology (IT) department of an enterprisecan create vFiler units for various organizations, or customers, within the enterprise.

The administrator for each customer can manage and view files only on the assigned vFiler unit, noton other vFiler units that reside on the same storage system. In addition, there is no data flow betweenvFiler units. A customer using a vFiler unit is assured that no sensitive information is exposed to othercustomers that store data on the same storage system.

For example, an SSP can create the following vFiler units on a storage system:

• A vFiler unit named vFilerAIt uses the /vol/vol1 voume and the e0 interface on the storage system.It is leased to CompanyA.

• A vFiler unit named vFilerBIt uses the /vol/vol2 volume and the e1 interface on the storage system.It is leased to CompanyB.

Although both CompanyA and CompanyB store data on the same storage system, network traffic foreach company is restricted to the specified interface. The administrator at CompanyA (that uses NFSto access data) cannot use the showmount command on a UNIX client to view directories on the storagesystem that are outside the /vol/vol1 volume. Similarly, the administrator at CompanyB (that uses CIFSto access data) cannot browse the shares that are outside the /vol/vol2 volume.

Related concepts

What IPspaces are on page 65


MultiStore for data migrationMultiStore enables you to migrate data from one storage system to another without extensivereconfiguration on the destination storage system.

Without MultiStore, you can use the SnapMirror technology to copy data from one storage system toanother; however you then need to edit the files’ACLs, local user group definitions, user mappinginformation, and so on, before users can access the data. With MultiStore, the data being copied isstored on a vFiler unit that encapsulates all user, group, and ACL information. You re-create a vFilerunit on the destination storage system with the encapsulated information, and data can be served fromthe destination storage system without further reconfiguration. UNIX clients do not experience anydisruption in service when the vFiler unit on the destination storage system starts serving data insteadof the vFiler unit on the source storage system.

Related concepts


What the host storage system isThe storage system on which you create vFiler units is called the host storage system. The storage andnetwork resources used by the vFiler units exist on the host storage system.

What the default vFiler unit isWhen you license MultiStore, Data ONTAP automatically creates a default vFiler unit on the hoststorage system. The name of the default vFiler unit is vfiler0.

Initially, vfiler0 owns all the resources on the storage system. After you create additional vFiler units,vfiler0 owns the resources that are not owned by the vFiler units you create.

The default vFiler unit exists as long as the MultiStore license is enabled. On a storage system with theMultiStore license enabled, you cannot destroy vfiler0.

All information provided in this guide about the vFiler units is applicable to vfiler0, unless notedotherwise.


Some limitations of MultiStoreMultiStore has limitations to the maximum number of vFiler units, supported interfaces, supportedprotocols, and supports most Data ONTAP commands.

• You can have a maximum of 65 vFiler units on a storage system. However, the maximum limit doesdepend on the memory capacity of the host storage system.You can create 64 vFiler units on a storage system. The 65th vFiler unit is vfiler0, which is createdautomatically when MultiStore is licensed on the storage system. vfiler0 continues to exist as longas MultiStore is licensed.In an active/active configuration, you can create up to 64 vFiler units on each node of the active/activeconfiguration, for a maximum of 130 vFilers in the active/active configuration.

• Only Ethernet interfaces are supported for vFiler units.

• Only the NFS, CIFS, iSCSI, HTTP, and FTP protocols are supported for the vFiler units. However,protocols not supported on vFiler units are supported on vfiler0, the default vFiler unit.

• Most Data ONTAP commands are supported on the vFiler units. The specific commands that arenot supported are listed on a feature-by-feature basis later in this guide.

Note: These limitations can be exceeded only during a takeover scenario, when one storage systemtakes over the vFiler resources of another storage system.

A takeover scenario happens during either of the following.

• A maintainance activity.

• A high-availability event.

Related tasks

Viewing the current vFiler unit limit on page 41

Host storage system's access to vFiler unit dataAs the host storage system administrator, you do not have administrative, CIFS, or NFS access to thedata owned by the vFiler units (other than data owned by the default vFiler unit , vfiler0). After youassign a qtree or volume to a vFiler unit , you lose access to the data in that qtree or volume.

Example:

Before you create a vFiler unit with the /vol/vol1 volume, you can configure the/etc/exports file so that you can mount the /vol/vol1 volume. After you create the vFilerunit , an attempt to mount the /vol/vol1 volume from the host storage system results in the


error message ".../vol/vol1 belongs to vFiler unit A, can't mount from vfiler0."

Related concepts

How to manage vFiler unit storage from the storage system on page 35

Related tasks

Executing commands for a vFiler unit from the host storage system on page 47

Types of system administration tasksYou can perform host storage system tasks, which are the same as those for a storage system that is notsupporting vFiler units, MultiStore management tasks, which are related to establishing vFiler unitsand managing resources on vFiler units, and tasks related to using DAta ONTAP features that aresupported by vFiler units.

Next topics

Host storage system tasks on page 25

MultiStore management tasks on page 25

Tasks related to using Data ONTAP on vFiler units on page 26

Host storage system tasksYou can perform tasks related to managing the resources on the host storage system in the same waythat you perform them on a storage system without a MultiStore license.

You can use either the command line or FilerView to perform these tasks.

• Manage volumes, disks, and RAID groups.

• Increase data availability through Snapshot management, SnapMirror management, and volumecopy.

• Back up and recover data

These tasks are covered in detail in the Data ONTAP File Access and Protocols Management Guide,the Data ONTAP Storage Management Guide, and the Data ONTAP Data Protection Online Backupand Recovery Guide.

MultiStore management tasksYou can manage MultiStore from the host storage system using the command line or FilerView.

You can perform the following MultiStore management tasks


• Enable and disable the MultiStore license

• Allow and disallow protocols to be run a vFiler unit

• Create a vFiler unit

• Set up a vFiler unit

• Start and stop a vFiler unit

• Destroy a vFiler unit

• Moving resources to and from a vFiler unit

• Monitor the status of a vFiler unit

From a destination storage system, you can perform the following additional tasks:

• Move a vFiler unit from one host storage system to another

• Create a disaster-recovery vFiler unit

Related concepts

MultiStore Management on page 27

Disaster recovery and data migration on page 87

Tasks related to using Data ONTAP on vFiler unitsYou can use Data ONTAP features on individual vFiler units the same way that you would use themon a storage system.

FilerView cannot be used to configure or use these features on vFiler units; you must use the commandline.

The tasks are as follows.

• Manage users

• Manage iSCSI, LUNs and initiator groups

• Manage NFS exports and CIFS shares

• Manage quotas

• Prepare vFiler units for virus protection


MultiStore Management

You run MultiStore management tasks from the host storage system.

Next topics

How to create a vFiler unit on page 28

How to manage vFiler unit storage from the storage system on page 35

Default vFiler unit limits on page 39

Maximum vFiler units allowed on page 39


Increasing the vFiler unit limit on page 41

Decreasing the vFiler unit limit on page 42

Renaming a vFiler unit on page 42

Stopping a vFiler unit on page 43

Starting a vFiler unit on page 44

Allowing a protocol for a vFiler unit on page 44

Disallowing a protocol for a vFiler unit on page 45

Displaying vFiler unit status on page 46

Viewing commands that can be executed from a vFiler unit on page 46

Executing commands from a vFiler unit on page 47

Executing commands for a vFiler unit from the host storage system on page 47

Executing commands for a vFiler unit using RSH on page 48

Executing commands for a vFiler unit using SSH on page 49

Effects of storage system reboot on a vFiler unit on page 50

Volumes and qtrees on a vFiler unit on page 50

Backups for vFiler units on page 52

LUNs on a vFiler unit on page 55

Networking considerations on page 57

SnapMirror on the host storage system on page 58

SnapVault on the host storage system on page 60

SNMP used with a vFiler unit on page 61

What is not supported on a vFiler unit on page 62

Monitoring performance and statistics on page 62

MultiStore Management | 27

How to create a vFiler unitBefore you can start protocol servers on a vFiler unit , you must create a vFiler unit on your storagesystem.

The following conditions must be met before you can create a vFiler unit :

• You must create at least one unit of storage (qtrees, volumes - traditional or flexible) before creatingthe vFiler unit .

• The storage unit that contains vFiler unit configuration information must be writable. It must notbe a read-only file system, such as the destination volume or qtree in a SnapMirror relationship.

• The IP address used by the vFiler unit must not be configured when you create the vFiler unit .

Next topics

Enabling the MultiStore license on page 28

Disabling the MultiStore license on page 29

Storage guidelines on page 29

Language guidelines on page 30

Quota guidelines on page 30

Active/active considerations on page 30

SAN considerations on page 31

The vFiler commands on page 31

Ensuring that the network interface is ready on page 32

Assigning and configuring vFiler unit resources on page 33

Manually setting up a vFiler unit on page 34

Related concepts


Enabling the MultiStore licenseTo begin using MultiStore, you must enable the MultiStore license key.

Step

1. Enter the following command:

license add license_key

Enabling the MultiStore license has the following effects on the storage system:

• You can use the vfiler and ipspace commands.

• Data ONTAP starts logging vFiler status and sends the information using the AutoSupport feature.


• The routed daemon is disabled.

• The ip.match_any_ifaddr option is set to Off.

• The vFiler unit limit (the number of vFiler units you can create on this storage system, includingvfiler0) is set to a default value between 3 and 11, depending on the memory capacity of the hoststorage system.

Related tasks


Disabling the MultiStore licenseIf you are not using vFiler units , you can disable the MultiStore license.

Before You Begin

You can disable the MultiStore license only when no vFiler units other than vfiler0 are on the storagesystem. If there are other vFiler units on the storage system, you must destroy them before disablingthe MultiStore license.

Step


license delete multistore

Disabling the MultiStore license has the following effects:

• MultiStore becomes unavailable immediately. You can no longer use the vfiler and ipspacecommands.

• The routed daemon is enabled after the next reboot.

Related tasks

Destroying a vFiler unit on page 38

Storage guidelinesYou need to remember certain guidelines when deciding which storage units to assign to vFiler units.

• The first storage unit assigned to the vFiler unit contains the vFiler unit configuration information.It is called the primary storage unit. Although you can remove storage units from a vFiler unit atany time after the vFiler unit is created, the primary storage unit must remain for as long as thevFiler unit exists.

• The primary storage unit has the same security characteristics it had before it was transferred to thevFiler unit . When you create a new vFiler unit , C$ share-level permissions are restricted toadministrators only, but file-level security is not modified. The vFiler unit administrator can setmore restrictive file-level permissions.


• If the qtree or volume to be used as the primary storage unit contains an /etc directory, the data inthe directory is lost after you add the qtree or volume to a vFiler unit . Data in qtrees and volumesused as non-primary storage units is preserved.

• A volume assigned to a vFiler unit must not be the storage system’s root volume. You can, however,assign qtrees in the root volume to a vFiler unit.

• A volume assigned to a vFiler unit can be a traditional volume or a FlexVol volume. You cannotassign aggregates to a vFiler unit.For information about traditional volumes, FlexVol volumes, and aggregates, see the Data ONTAPStorage Management Guide.

• FlexCache volumes can be created on the default vFiler unit , vfiler0, but cannot be assigned ormoved to any other vFiler unit

• A qtree is assigned to a vFiler unit , owned by a vFiler unit or associated with a vFiler unit only ifthat qtree is added as a resource to a vFiler unit . If a volume containing a qtree is added as a resourceto a vFiler unit , then the qtree implicitly becomes a resource of that vFiler unit.

• If the vFiler unit administrator needs to create qtrees on the vFiler unit , assign volumes instead ofqtrees to the vFiler unit when creating the vFiler unit . This is because qtrees can be created only atthe root of a volume.

• If you anticipate that you might have to move the disks that are used for the vFiler unit's storagefrom one storage system to another, assign volumes instead of qtrees, to the vFiler unit.

• When managing NFS exports, CIFS shares, quotas, and options, vFiler unit administrators need toenter the complete path names of the storage resources used by the vFiler units in commands andconfiguration files. Therefore, choose volume and qtree names appropriately, so that you can sharewith the vFiler unit administrators the complete path names beginning withfiler_name:/vol/vol_name.

Language guidelinesvFiler unit administrators need to edit the /etc/quotas and /etc/usermap.cfg files for their vFiler units .These files support Unicode and root volume UNIX encoding. To ensure that vFiler unit administratorscan edit these files without requiring Unicode-capable editors, create vFiler units on a storage systemwhose root volume language can be used for editing.

Quota guidelinesWhen you create a vFiler unit the ownership of a volume or qtree is changed from the host storagesystem to a specified vFiler unit . This change requires that quotas be turned off for the affected volume.You can turn the quotas back on for the volume after the vFiler unit is created.

Active/active considerationsThere are some considerations to be taken into account for establishing vFiler units on an active/activeconfiguration:


• You can create up to 64 vFiler units on each member of an active/active configuration, dependingon the memory capacity of the host storage systems.

• Each member of an active/active configuration must have a MultiStore license to take over its partnerwith a MultiStore license.

• The vFiler units hosted by the storage systems of the active/active configuration are created andconfigured independently. This means each storage system can host a different number of vFilerunits , and the vFiler unit configurations on the storage systems can be different from each other.

• In takeover mode, the functioning storage system takes over all vFiler units created on the failedstorage system. These vFiler units include the vFiler units you create and the unit called vfiler0.Therefore, for vFiler units on the failed storage system to work correctly after the takeover, eachnetwork interface used by a vFiler unit in an active/active configuration must have a partner interface.

Related tasks


SAN considerationsWhen you create vFiler units on a SAN, you must keep some LUN and iSCSI considerations in mind.

• FCP LUNs and igroups are supported only on the host storage system, not on vFiler units.

• iSCSI LUNs and igroups are supported on all vFiler units and managed separately for each vFilerunit .

• When you create a vFiler unit on a storage system on which iSCSI is licensed, the iSCSI service isautomatically started on the vFiler unit .

Note: If iSCSI is licensed on the host storage system and the storage to be allocated to the vFilerunit contains LUNs, unmap the LUNs.

• Starting a vFiler unit starts iSCSI packet processing for that vFiler unit .

• Stopping a vFiler unit stops iSCSI packet processing for that vFiler unit .

The vFiler commandsThe vfiler commands, which are supported only on the host storage system, enable the host storagesystem administrator to set up vFiler units , manage vFiler unit resources, and manage Data ONTAPfeatures on individual vFiler units .

Eachvfiler command has syntax different from the others. The general vfiler command syntaxis as follows:

vfiler command vfilertemplate options...

Some vfiler commands support the vfilertemplate option. vfilertemplate can be any of thefollowing.

• A vFiler unit name

• A comma-seperated list of vFiler unit names


• An IPspace declaration

• An asterisk used as a wild card

If you use the asterisk, the command takes effect on all vFiler units , including vfiler0 (the host storagesystem), unless the command is inapplicable to vfiler0. See the na_vfiler man page for more information.

Some vfiler commands include a complete path name for the qtree or volume assigned to the specifiedvFiler unit.

Related concepts

LUNs on a vFiler unit on page 55

File Access Using NFS and CIFS on page 77

Disk Space Management Using Quotas on page 81

Virus protection for CIFS on page 121

Ensuring that the network interface is readyBefore you create a vFiler unit, you must ensure that the network interface is ready.

Considerations

To ensure that the network interface to be configured with a vFiler unit IP address is ready for vFilerunit creation, complete the following steps.

• If the IP address for the vFiler unit is a base IP address for an interface, enter the following commandto change the state of the interface for the IP address to down.

ifconfig interface down

ExampleThe following example changes the state of the e0 interface to down.

ifconfig e0 down

• If the IP address for the vFiler unit is an IP alias for an interface, enter the following command toremove the alias:

ifconfig interface -alias address

ExampleThe following example removes the IP alias from the e0 interface:

ifconfig e0 -alias 123.123.123.123

• If the IP alias is currently assigned to an interface, enter the following command to remove the alias:


• If the IP alias is currently unassigned, the network interface is ready.• If the base IP address for the vFiler unit is assigned to an interface in the up state, enter the following

command to change the state of the interface to down :

ifconfig interface down


• If the base IP address for the vFiler unit is assigned to an interface in the down state, the networkinterface is ready

Assigning and configuring vFiler unit resourcesTo create a vFiler unit, and assign and configure its resources, you can use the CLI or FilerView.

Before You Begin

The following procedure assumes that you want to set up the vFiler unit for networking (and CIFS ifnecessary) immediately after creating the vFiler unit. If you want to defer the setup, use the -n optionof the vfiler create command. For more information about the vfiler create command, seethe na_vfiler man page.

Steps


vfiler create vfiler_name -i ip_address [ -i ip_address ] ... path [ path

] ...

vfiler_name is the name of the vFiler unit.

ip_address is an IP address.

path is the complete path name to an existing volume or qtree. The first path name is the storageunit that contains the /etc directory. The etc directory contains the configuration information aboutthe vFiler unit.

ExampleThe following example creates a vFiler unit using two IP addresses, one volume, and one qtree:

vfiler create vfiler1 -i 123.123.123.123 -i 123.123.123.124 /vol/vol1/vol/vol2/qtree2

2. Respond to the prompts to set up the storage system, and to set up CIFS if necessary.

The setup process does the following on the new vFiler unit:

• Starts NFS (if NFS is licensed on the host storage system) and configures the vFiler unit’s primarystorage (root volume) to be exported to the vFiler unit’s administrative host (using an entry in the/etc/exports file)

• Configures the vFiler unit’s IP addresses and adds the appropriate entries to /etc/rc

• Creates a “pseudo-root” that allows CIFS clients to see all the storage that has been assigned to thevFiler unit as a single tree

• Starts the iSCSI service (if iSCSI is licensed on the host storage system)


After You Finish

As the vFiler unit administrator, you can now mount the root volume of the vFiler unit, edit/etc/exports to suit your needs and rerun the exportfs command.

Related concepts


Related tasks


Manually setting up a vFiler unitIf you need to change the network configuration of your vFiler unit, you can manually set up the vFilerunit.

Considerations

If you use the -n option of the vfiler create command, no automatic setup is done, and no protocolservers will run on the vFiler unit until you set it up manually.

The procedure for setting up a vFiler unit manually is similar to that for setting up a storage system.You run the setup command, which enables you to do the following tasks:

• Specify the administration host for the vFiler unit

• Specify the password for the root user on the vFiler unit

• Specify the bindings of IP addresses to network interfaces

• Enable DNS and specify the DNS domain and servers

• Enable NIS client service and specify the NIS domain and servers

Note: The setup command does not prompt you for the time zone. All vFiler units are in the sametime zone as the host storage system.

If the vFiler unit is licensed to deliver CIFS service, you must run cifs setup, as you would for astorage system, in addition to running setup.

Steps


vfiler run vfiler_name setup

The setup command displays prompts for you to configure the vFiler unit. After you respond toall the prompts, configuration files, such as the /etc/exports file, are created in the /etc directoryfor the vFiler unit.

Note: Unlike the setup command for the storage system, the setup command for a vFiler unitdoes not cause NFS to start running.


If the vFiler unit runs the CIFS protocol, go to Step 2. Otherwise the setup is complete.


vfiler run vfiler_name cifs setup

The cifs setup command displays prompts for you to configure CIFS on the vFiler unit. Afteryou respond to all the prompts, CIFS starts running.

Related concepts


How to manage vFiler unit storage from the storage systemAs the physical storage system administrator, if you need to manage storage resources that belong to avFiler unit, but you do not have administrative access to the vFiler unit, you can temporarily movethevFiler unit's resources, or temporarily destroy the vFiler unit.

Note: Before taking either of the following actions, unmap any LUNs that have been created in theaffected storage resources. For instructions, see the Data ONTAP Block Access Management Guidefor iSCSI and FCP.

• Temporarily move the resources to the host storage system; however you cannot move the vFilerunit's primary /etc volume.

• Temporarily destroy the vFiler unit.This returns ownership of all resources to the host storage system. No user data is modified whenyou destroy a vFiler unit.

Depending on which action you take, you can later move the storage resources back to the vFiler unit,or restore the vFiler unit.

Next topics

Effects of adding, removing, and moving vFiler unit resources on page 36

Adding resources to a vFiler unit on page 36

Requirements for moving and removing resources on page 36

Removing resources from a vFiler unit on page 37

Moving resources between vFiler units on page 37


Restoring a vFiler unit on page 39


Effects of adding, removing, and moving vFiler unit resourcesAdding, removing, or moving vFiler unit resources affects only the association between the vFiler unitand the resources. It does not have any effect on user data in the vFiler unit.

• After you add storage resources to a vFiler unit, the resources are moved from the host storagesystem to a vFiler unit.

• After you remove storage resources from a vFiler unit, the resources are removed from the vFilerunit to the host storage system.

• After you add an IP address to a vFiler unit, you can assign the address as an IP alias of an interfaceor assign the address to a network interface that has not been configured.

• After you remove an IP address from a vFiler unit, the IP address becomes an unassigned IP address.

• After you move resources from one vFiler unit to another, the resources are moved from the resourcevFiler unit and added to the destination vFiler unit.

Adding resources to a vFiler unitTo partition the storage and network resources of a single storage system so that it appears as multiplestorage systems on the network, you must add resources to a vFiler unit.

Step


vfiler add vfiler_name [ -f ] [ -i ip_address [ -i ip_address ] ...] [

path [ path ... ] ]

Use the -f option to skip warning messages.

ExampleThe following example adds an IP address and a volume to an existing vFiler unit:

vfiler add vfiler1 -i 123.123.123.125 /vol/vol3

Requirements for moving and removing resourcesWhen you move or remove vFiler unit resources, you must consider some requirements.

• The storage unit to be moved or removed must not contain the vFiler unit’s /etc directory.

• If a storage unit is to be moved or removed, or contains any CIFS shares, home directories, or openfiles and directories, you must remove the CIFS shares, remove the home directories from the listof home directories, or close open files and directories.

• Both the source vFiler unit, and destination vFiler units must be in the same IPspace.

• If the address is an IP alias, the alias must be removed. If the address is not an IP alias, the networkinterface associated with the address must not be configured.


Removing resources from a vFiler unitYou can remove resources you have added to the vFiler unit. An example of such a requirement is whenan SSP wants to reduce the amount of storage used, to bring down the operating costs.

Before You Begin

If a storage unit you want to remove contains LUNs, you must unmap the LUNs first. For instructions,see the Data ONTAP Block Access Management Guide for iSCSI and FCP

Step


vfiler remove vfiler_name [ -f ] [ -i ip_address [ -i ip_address ] ...] [

path [ path ... ] ]


ExampleThe following example removes an IP address and a volume from an existing vFiler unit:

vfiler remove vfiler1 -i 123.123.123.125 /vol/vol3

Moving resources between vFiler unitsYou can move storage resources from one vFiler unit to another. For example, if the data on a qtree orvolume changes ownership, and that data can be moved to another vFiler unit that is owned by anothercompany or business unit.

Before You Begin

If a storage unit that you want to move from one vFiler unit to another contains LUNs, you must unmapthe LUNs first. See the Data ONTAP Block Access Management Guide for iSCSI and FCP.

Considerations

• After you move a storage unit from one vFiler unit to another, the security information associatedwith the files in the storage unit is retained. As a result , users might be unable to access files properly.

• If you reassign a volume from one vFiler unit to another, Data ONTAP turns off quotas for thevolume. After the volume is moved, you can turn quotas on again for the volume from the destinationvFiler unit.

• If you reassign a qtree from one vFiler unit to another, Data ONTAP turns off quotas for the volumecontaining the qtree on both the source vFiler unit and the destination vFiler unit. After the qtree ismoved, you can turn quotas on again for the volume.

• When resources are being moved, all network connections to the vFiler units are terminated


Step


vfiler move source_vfiler destination_vfiler [ -f ] [ -i ip_address [ -i

ip_address ] ...] [ path [ path ... ] ]


ExampleThe following example moves an IP address and a volume from one vFiler unit to another:

vfiler move vfiler1 vfiler2 -i 123.123.123.125 /vol/vol3

Destroying a vFiler unitIf you want to return storage resources back to the host storage system (and the storage administrator'sdomain), destroy the vFiler unit that contains the storage resources.

Before You Begin

Before you destroy the vFiler unit, you must unmap any LUNs that are mapped to the vFiler unit'sstorage, and stop the vFiler unit.

Steps

1. Unmap any LUNS that are mapped to the vFiler unit's storage.

2. To stop the vFiler unit, enter the following command:

vfiler stop vfiler_name

3. To destroy the vFiler unit, enter the following command:

vfiler destroy [ - f ] vfiler_name

Without the -f option, the command displays a confirmation prompt. With the -f option, thecommand destroys the vFiler unit immediately.

• All resources associated with the vFiler unit are released to the host storage system.

• No data is destroyed.

• All the vFiler unit's IP addresses are not configured and the corresponding entries in the storagesystem's /etc/rc file are removed.

• If the vFiler unit was not in the same IPspace as the host storage system, the IP addresses previouslyowned by the vFiler unit are not available for use after you destroy the vFiler unit.

• The effects on quotas are the same as when you move resources from a vFiler unit to the host storagesystem.

• Clients using LUNs experience an interruption of service.


Related tasks

Moving resources between vFiler units on page 37

Restoring a vFiler unitIf you need to reallocate storage resources from a storage system to a vFiler unit, you can restore anyvFiler unit you have destroyed.

Step


vfiler create oldname -r oldfirstpath

oldname is the name of the original vFiler unit.

oldfirstpath is the first path that was specified in the original vfiler create command thatcreated the vFiler unit.

For more information about the vfiler create command, see the na_vfiler man page.

Related tasks


Default vFiler unit limitsSee the following default vFiler unit limits for systems on which you have activated a MultiStore license.

• Three vFiler units for storage systems with less than 1 GB (1,024 MB) of memory

• Five vFiler units for storage systems with 1 GB to less than 2 GB of memory

• Eleven vFiler units for storage systems with 2 GB or more of memory

The default vFiler unit limit is 11 for systems on which a MultiStore license has been in effect since arelease of Data ONTAP earlier than 6.4.

These limits include vfiler0, therefore a limit of 11 means that you can create a maximum of 10 vFilerunits on each active/active configuration node.

Maximum vFiler units allowedThere are limits to the number of vFiler units that can be created, depending on the available memory.


Number of vFiler unitsMemory

11Less than 1 GB

261 GB or more

652 GB or more

Use the sysconfig -v command to verify the memory size of your system.

Note: The numbers in the preceding table include the default vFiler unit (vfiler0). There is anexception to the maximum limits; the memory on a FAS270 storage system is 1,022 MB (less than1,024 MB); however it can still support 26 vFiler units.

Example:

To verify the memory size on your system, enter the following command on your storage systemconsole:

toaster> sysconfig -v

..........

..........

..........

slot 0: System Board 650 MHz (TSANTSA C0)

Model Name: FAS270

Part Number: 110-00046

Revision: C0

Serial Number: 287200

Firmware release: CFE 1.2.0

Processors: 2

Processor revison: B2

Processor type: 1250

Memory Size: 1022 MB

NVMEM Size: 128 MB of Main Memory Used

In this example, the memory size is 1,022 MB, or less than 1 GB (1,024 MB), so the maximumnumber of supported vFiler units should be 11. However, the Model Name field in the preceding


example indicates that it is a FAS270 storage system; therefore, the maximum number of vFilerunits in this case is 26.

Viewing the current vFiler unit limitTo determine whether you want to increase or decrease the current vFiler unit limit, you can view thecurrent limit.

Considerations

The vFiler unit limit specifies the maximum number of vFiler units that can exist on the host storagesystem. Because the limit includes the host storage system, vfiler0 (which always exists if the MultiStorelicense is enabled), the number of vFiler units you can create on a storage system is one less than thevFiler unit limit set on a storage system.

In an active/active configuration, the same limit applies to each active/active node.

Step


vfiler limit

Increasing the vFiler unit limitIf you need more partitions on your host storage system, you can increase the vFiler unit limit. Themaximum number of vFiler units you can have on a storage system depends on the memory capacityof the host storage system.

Step


vfiler limit number

ExampleTo raise the number of vFiler units you can create to 15, enter the following command on the hoststorage system:

vfiler limit 16

In an active/active configuration, this sets a limit of 16 vFiler units on each active/active node.


After You Finish

For the change to take effect, you must reboot the storage system (or each storage system in anactive/active configuration).

Decreasing the vFiler unit limitIf you need fewer partitions on your host storage system, you can decrease the vFiler unit limit.Whenyou decrease the limit, the change is effective immediately and does not require a reboot of the storagesystem.

Step


vfiler unit number

ExampleTo reduce the number of vFiler units you can create from 15 to 10, enter the following commandon the host storage system:

vfiler limit 11

Renaming a vFiler unitUsing the vfiler rename command, you can rename a vFiler unit to give it a unique name.

Considerations

• The new name should not exist on the storage system, or on the partner storage system in anactive/active configuration. Although Data ONTAP allows the storage system and its partner tohave vFiler units with identical names, it is easier to administer the systems if you create a uniquename for each vFiler unit.

• The vfiler rename command does not change the CIFS node name of the vFiler unit or affectany client connections that are active. The vfiler rename command changes the name of thevFiler unit only within Data ONTAP; it does not re-broadcast the new name to the CIFS domaincontrollers or the NetBIOS nameservers because these protocols might be using a different namefor the vFiler unit than Data ONTAP uses. To change the name mapping in the CIFS domaincontrollers, run CIFS setup for each of these protocols.

• Do not rename a vFiler unit while it is being migrated. Doing so will cause the migrate commandon the remote system to fail.


Step


vfiler rename old_vfiler_name new_vfiler_name

ExampleIn the following example, a vFiler unit, vfiler1, is renamed to vfiler2:

vfiler rename vfiler1 vfiler2

Stopping a vFiler unitYou can stop a vFiler unit if you need to troubleshoot vFiler unit problems or destroy a vFiler unit.

Considerations

Note: You cannot stop vfiler0.

After you stop a vFiler unit, the vFiler unit can no longer receive packets from clients.

The stopped state is not persistent across reboots; after you reboot the storage system the vFiler unitresumes automatically.

If iSCSI is licensed on the storage system, stopping a vFiler unit stops iSCSI packet processing for thatvFiler unit.

Step


vfiler stop vfilertemplate

ExampleThe storage system supports two vFiler units : vfiler1 and vfiler2. The following command stopsall vFiler units , except vfiler0:

vfiler stop *

The following messages appear after you enter the command:

stoppedvfiler1

stoppedvfiler2


Starting a vFiler unitUse the vfiler start command to start a vFiler unit.

Considerations

Note: You cannot start vfiler0.

You can start a vFiler unit that is in the stopped state; after a vFiler unit starts, it is in running state andcan receive packets from clients.

If iSCSI is licensed on the storage system, starting or stopping a vFiler unit starts or stops iSCSI packetprocessing for that vFiler unit.

Step


vfiler start vfilertemplate

ExampleThe storage system supports two vFiler units named vfiler1 and vfiler2. The following commandstarts all vFiler units , except vfiler0, which is already running:

vfiler start *

The following messages appear after you enter the command:

runningvfiler1

runningvfiler2

Allowing a protocol for a vFiler unitBy default, a vFiler unit can use the protocols for the host storage system. You can select the protocolsthat you want to allow on the vFiler units.

Considerations

The maximum number of FTP connections to a vFiler unit is determined by theftpd.max_connections option set on the host storage system. The value set for this option is sharedamong the vFiler units on a storage system.

You can select the protocols that you allow on the vFiler units using the vfiler allow command.vFiler units support the following protocols:


• CIFS

• NFS

• RSH

• SSH

• iSCSI

• FTP

• HTTP

To allow CIFS, NFS, FTP, or HTTP on a vFiler unit, each of the allowed protocols must have an activelicense on the host storage system.

Step

1. To allow a protocol, enter the following command:

vfiler allow vfilertemplate proto=protocol ...

protocol is nfs, cifs, iscsi, rsh, ssh, ftp, or http.

ExampleThe following example allows the NFS and RSH protocols on the vFiler unit named vfiler1:

vfiler allow vfiler1 proto=nfs proto=rsh

Disallowing a protocol for a vFiler unitBy default, a vFiler unit can use protocols for the host storage system. You can select the protocols thatyou want to restrict on the vFiler units.

Considerations

If a CIFS, NFS, iSCSI, FTP, or HTTP protocol is running and you restrict it, the protocol continues torun until the storage system reboots, but packets destined for the vFiler unit are ignored.

Step

1. To restrict a protocol, enter the following command:

vfiler disallow vfilertemplate proto=protocol ...

protocol is nfs, cifs, iscsi, rsh, ssh, ftp, or http.

ExampleThe following command restricts the NFS and RSH protocols on the vFiler unit named vfiler1:

vfiler disallow vfiler1 proto=nfs proto=rsh


Effects of restricting iSCSI: After iSCSI is restricted on a vFiler unit, the following conditions applyon that vFiler unit:

• You cannot start iSCSI.

• No new iSCSI sessions are allowed.

• iSCSI commands on existing sessions are rejected.

Effects of restricting FTP: After FTP is restricted on a vFiler unit, no new FTP connections are allowedto the vFiler unit; however transfers that started before FTP was restricted will finish.

Effects of restricting HTTP: After HTTP is restricted on a vFiler unit, no new HTTP connections areallowed on the vFiler unit. Each new request receives a "503 HTTP server is disabled" message. Anyexisting connections remain alive.

Effects of restricting RSH and SSH: Although the Data ONTAP rsh.enable and ssh.enable optionvalues (On or Off) determine whether the RSH or SSH server is enabled or disabled on a storage system,restricting RSH or SSH on a vFiler unit is independent of the value for that option. A vFiler unit canbe configured to restrict RSH or SSH even when the corresponding enable option is set to On.

Note: To allow RSH on a vFiler unit, you must have the rsh.enable option set to On . To allowSSH on a vFiler unit, you must have the ssh.enable option set to On .

Displaying vFiler unit statusYou can use the vfiler status command to see whether the vFiler unit is stopped or started, IPaddresses that have been assigned are configured, which interfaces they are bound to, which protocolsare allowed or disallowed, and so on. See the na_vfiler(1) man page for more information.

Step


vfiler status

Viewing commands that can be executed from a vFiler unitYou can use the vfiler run vfilertemplate ? command to view what commands you can executefrom a vFiler unit.


Before You Begin

Not all Data ONTAP commands can be executed for a vFiler unit, and special considerations orrestrictions apply to some of those that can be executed. For a description of the special considerationsand restrictions, see the vFiler considerations section of each command's man pages.

Step


vfiler run vfilertemplate ?

Executing commands from a vFiler unitIf you want to work strictly with the data on a vFiler unit, you can execute commands directly from it.These commands are run as if they are being run on the vFiler unit's console, and are subject to theconstraints of the vFiler unit.

Steps


vfiler context vfiler_name

Examplevfiler context vfiler1

2. Enter the command you want to run:

Example?

This shows all the commands available to you from the context of the vFiler unit.

3. To return to the context of the host storage system, enter the following command:

vfiler context vfiler0

Executing commands for a vFiler unit from the host storagesystem

You can execute commands for a vFiler unit from the host storage system.


Step


vfiler run vfilertemplate command

Examplevfiler run vfiler1 useradmin

Related tasks


Executing commands for a vFiler unit using RSHYou can execute commands for a vFiler unit using RSH.

Before You Begin

The RSH protocol must be allowed for the vFiler unit; by default, RSH is allowed.

The RSH protocol must be enabled for the vFiler unit; the rsh.enable option for the vFiler unit mustbe set to On. By default, RSH is enabled.

You must enter the command on a client that is permitted to have RSH access to the vFiler unit; theclient must be one of the hosts specified by the rsh.access option for the vFiler unit.

Step


rsh vfiler_IP_address command

ExampleThe following command displays all options on the vFiler unit whose IP address is 123.123.123.1:

rsh 123.123.123.1 options

RSH commandsYou can enter the following RSH commands on a vFiler unit.

snapmirrornfshostname?

snapvaultnfsstatigroupcifs

useradminnisipsecconfig

voloptionsiscsidf


vscanpasswdkeymgrdns

wccqtreelockecho

ypcatquotalunexportfs

ypgrouproutenbtstatfilestats

ypmatchsecureadminndmpcopyfpolicy

ypwhichsnapndmpdfsecurity

netstathelp

Note: The hostname command is only for displaying, not changing, the name of the vFiler unit.

Executing commands for a vFiler unit using SSHYou can execute commands for a vFiler unit using SSH.

Before You Begin

The SSH protocol must be allowed for the vFiler unit; by default, RSH is allowed.

The SSH protocol must be enabled for the vFiler unit; that is the ssh.enable option for the vFilerunit must be set to On; by default, RSH is enabled.

You must enter the command on a client that is permitted to have SSH access to the vFiler unit; theclient must be one of the hosts specified by the ssh.access option for the vFiler unit.

Step


ssh vfiler_IP_address command

ExampleThe following command displays all options on the vFiler unit whose IP address is 123.123.123.1:

ssh 123.123.123.1 options

SSH commandsYou can enter the following SSH commands on a vFiler unit.

snapmirrornfshostname?

snapvaultnfsstatigroupcifs

useradminnisipsecconfig


voloptionsiscsidf

vscanpasswdkeymgrdns

wccqtreelockecho

ypcatquotalunexportfs

ypgrouproutenbtstatfilestats

ypmatchsecureadminndmpcopyfpolicy

ypwhichsnapndmpdfsecurity

netstathelp

Note: You cannot launch SSH as an interactive shell or issue a vFiler command that requires userinteraction through SSH.

Effects of storage system reboot on a vFiler unitWhen you reboot the storage sytem on which a vFiler unit is, the reboot can affect the state of a protocolor the state of the vFiler unit.

If you allow or disallow a protocol on a vFiler unit, this state persists across reboots. For example, ifyou disallow NFS for a vFiler unit and then reboot the storage system, NFS remains disallowed for thevFiler unit after the reboot.

If you disallow CIFS, NFS, or iSCSI for a vFiler unit when the protocol is enabled, rebooting the storagesystem disables that protocol for the vFiler unit. If you allow a protocol again after a reboot, you mustreenable the protocol.

Rebooting the storage system causes all stopped vFiler units to start running. For example, if you stopa vFiler unit and then reboot the storage system, the vFiler unit starts running again after the reboot.

Volumes and qtrees on a vFiler unitA volume assigned to a vFiler unit can be a traditional volume or a flexible (FlexVol) volume. Anaggregate cannot be assigned to a vFiler unit and you can create qtrees only if your vFiler unit ownsthe volume containing the qtree.

For information about traditional volumes, FlexVol volumes, and aggregates, see the Data ONTAPStorage Management Guide.


Example of creating a qtree on vfiler0

The /vol/vol0 volume is owned by vfiler0. You can use the following command to create aqtree in the /vol/vol0 volume:

qtree create /vol/vol0/qtree1

Example of creating a qtree on a vFiler unit

The /vol/vol1 volume is owned by vfiler1. The administrator for vfiler1 can use the followingcommand to create a qtree in the /vol/vol1 volume:

rsh vfiler1 qtree create /vol/vol1/qtree2

Next topics

Effects of taking a vFiler unit volume offline on page 51

Changes required after volumes are renamed on page 51

Who can change qtree security styles and oplock settings on page 51

Differences in qtree command output on page 52

Viewing all qtrees and the owner vFiler units on page 52

Effects of taking a vFiler unit volume offlineTaking a volume used for vFiler unit storage offline affects protocols used by the vFiler unit and LUNsin the volume.

• All CIFS shares and NFS exports in the volume are deactivated.

• If the volume contains the /etc directory for a vFiler unit, the vFiler unit stops running.After you put the volume back online, the vFiler unit starts running again.

• All LUNs become unavailable.

Changes required after volumes are renamedAfter you rename a volume used for vFiler unit storage, the vFiler unit administrator should change thepath names used in the vFiler unit’s /etc/exports file accordingly. Additionally, the vFiler unitadministrator should verify that CIFS shares and quotas are configured properly.

Who can change qtree security styles and oplock settingsYou can change the security style and oplock setting for a qtree or volume only if you are the ownerof that qtree or volume.

• If a vFiler unit owns a volume, you can change the security styles or oplock settings for the volumeand all qtrees on the volume from the vFiler unit.


• If a vFiler unit owns qtrees on a volume owned by the host storage system, you can change thesecurity styles or oplock settings from the vFiler unit only for the qtrees the vFiler unit owns.

• If the host storage system owns a volume that contains qtrees assigned to vFiler units , you canchange the security styles or oplock settings from the host storage system only for the qtrees thehost storage system owns.

Differences in qtree command outputThe qtree command output changes, depending on where you enter the command.

• If you enter the qtree command from the host storage system, the command displays informationabout all qtrees on the storage system, whether the qtrees are owned by the host storage system orvFiler units.

• If you enter the qtree command from a vFiler unit, the command displays information about qtreeson that vFiler unit only.

• If you enter the qtree command without arguments from a vFiler unit, a qtree that is the destinationqtree for SnapMirror is shown as read_only in the Status column.If you enter the qtree commandwithout arguments from a vFiler unit, a qtree that is the destination qtree for SnapMirror is shownas read_only in the Status column.

Viewing all qtrees and the owner vFiler unitsIf you want to view a list of qtrees grouped by the vFiler units that own the qtrees, you must run aspecific command from the host storage system.

Step


vfiler run * qtree status

Backups for vFiler unitsYou can back up vFiler unit data from the host storage system or from a vFiler unit’s client.

Remember the following points when you plan vFiler unit backups:

• From the host storage system, you can back up the storage units owned by vFiler units just as youwould if the vFiler units did not exist–—using the dump command, for example.You can back up all vFiler units at once. This method does not separate the data by vFiler unit, soit is not suitable if each vFiler unit’s data must be archived separately.

• From a client of a vFiler unit, you can back up all of that vFiler unit’s data, but no other vFiler unit’sdata.


A CIFS client can mount “/” from a vFiler unit and see a virtual tree comprising all of that vFilerunit’s storage units.

•

• A CIFS client can capture all of a vFiler unit’s data, both CIFS and NFS.

• An NFS client cannot see a virtual tree for the vFiler unit.

• An NFS client can back up all of the vFiler unit’s NFS data, but not its CIFS data.

If you want to back up an individual vFiler unit’s data separately, a good way is to back up from aclient (particularly a CIFS client). This backup method does not allow you to back up all vFilerunits at once.

NDMP supportNDMP supports vFiler units as well as physical storage system.

NDMP vFiler unit support is identical to storage system support except in the following areas:

• Local tape backup and restore commands are not supported in the individual vFilers. Commandsthat access physical tape drive resources must be executed in the default vFiler (vfiler0) context.

• NDMP SAN management commands are not supported in the individual vFiler unit context. Thesecommands must be executed in the default vFiler (vfiler0) context.

• VERITAS NDMP management commands are not supported in the individual vFiler unit context.These commands must be executed on the storage system.

• There is a hard limit of 160 concurrent NDMP sessions per storage system; therefore, an NDMPserver running on a vFiler unit could return “All sessions used up” even when there are no activesessions running on the vFiler.

Because each vFiler unit has its own NDMP server, NDMP enables you to back up or restore eachvFiler unit independently, and you can set NDMP options on each vFiler unit as well.

Next topics

Available NDMP options on page 53

ndmpcopy support on page 54

NDMP command support on page 54

NDMP password support on page 54

Available NDMP options

All the NDMP options from the options command are available on the default vFiler unit (vfiler0).

The following are the available NDMP options on the default vFiler (vfiler0).

• ndmpd.access

• ndmpd.authtype

• ndmpd.connectlog.enabled

• ndmpd.enable


• ndmpd.ignore_ctime.enabled

• ndmpd.offset_map.enable

• ndmpd.password.length

• ndmpd.perferred_interface

These NDMP options are also available on the nondefault vFiler units exceptndmp.perferred_interface.

ndmpcopy support

You can use the ndmpcopy command to copy data from one location to another using the NDMPprotocol.

In Data ONTAP 7.2 and later, you can use the ndmpcopy command to copy data from one vFiler unitto another vFiler unit, or between different locations on the same vFiler. The ndmpcopy commanduses the NDMP protocol over the external IP interfaces; therefore you must first ensure that you havenetwork connectivity, name resolution, and NDMP services configured properly at the source anddestination locations before attempting to use the command.

NDMP command support

You can use the ndmpd command to manage and monitor the NDMP service for individual vFiler unit.

The ndmpd command is used to manage and monitor the NDMP service for the individual vFiler unit.Enabling the NDMP service in a vFiler unit only enables the service for that vFiler unit; it does notenable the service for all vFiler units . Also, using the ndmpd command to monitor NDMP services andsessions in an individual nondefault vFiler unit context only displays information about the vFiler unityou are currently monitoring.

NDMP password support

When using the NDMP commands on the storage system, use the storage system’s root user’s passwordin the ndmpcopy command.

For enhanced security, the NDMP root user for individual nondefault vFiler units has a separate accountand password.

To view a nondefault vFiler unit’s root user or nonroot password on any vFiler unit, use the ndmpdpassword command with that user's user name. This command prints the NDMP user password requiredby the ndmpdcopy .


LUNs on a vFiler unitLUNs are portions instorage system that, when exported, look and act to the importing host like localdisks. Data on a LUN can be managed at the block level (for example, by a database manager) as wellas at the file level. A LUN is the basic unit of storage in a SAN.

Next topics

What is supported on a vFiler unit on page 55

iSCSI LUNs and igroups on a vFiler unit on page 55

The iSCSI service on a vFiler unit on page 56

What is supported on a vFiler unitData ONTAP allows you to create and manage a separate set of iSCSI LUNs and igroups on each vFilerunit. For detailed information about iSCSI LUNs, see the Data ONTAP Block Access ManagementGuide for iSCSI and FCP.

iSCSI LUNs and igroups on a vFiler unitFrom the point of view of a host importing LUNs, a vFiler unit looks and acts just like a storage system.Administrators of those hosts do not need to be aware that the LUN resides on a storage unit owned bya vFiler unit. However, as the vFiler unit or host storage system administrator, you need to be awareof some considerations.

Remember the following points when you manage LUNs on a storage system on which a MultiStorelicense is enabled:

• You must create and manage LUNs from the vFiler unit that owns the storage unit that contains theLUNs.

• A vFiler unit is aware only of those LUNs in the storage unit that it owns. When executed on avFiler unit, the lun show command displays only that vFiler unit’s LUNs.

• Ownership of LUNs changes with the ownership of the storage unit that contains the LUNS.

• LUNs must be unmapped before the storage unit containing them can be moved. This means youmust unmap all affected LUNs before doing any of the following:

• Assigning storage that contains LUNs to a vFiler unit, either when you create the vFiler unit orlater

• Destroying a vFiler unit that owns storage containing LUNs

• Moving storage that contains LUNs from one vFiler unit to another, or between a vFiler unitand the host storage system


If you try to move a storage unit without unmapping the LUNs it contains, the operation fails. Seethe Data ONTAP Block Access Management Guide for iSCSI and FCP for instructions for unmappingLUNs.

Note: You do not need to unmap LUNs when you migrate a vFiler unit or replace it fordisaster-recovery purposes.

• igroups are owned by the vFiler unit on which they are created.

• Like LUNs, a vFiler unit is aware only of those igroups that it owns. When executed on a vFilerunit, the igroup show command displays only that vFiler unit’s igroups.

• LUNs must be mapped to igroups owned by the vFiler unit that owns the LUNs.

• Each vFiler unit has its own name space for LUNs and igroups.

• igroups on different vFiler units can have the same initiator.

• LUNs on different vFiler units can have the same LUN ID.

• When you migrate a vFiler unit or replicate it for disaster-recovery purposes, LUNs owned by thevFiler unit are also migrated or replicated, along with their maps, igroups, and iSCSI configuration(the node names and the state of the iSCSI service). However, iSCSI authentication is not migratedor replicated.

Related concepts

Disaster recovery and data migration on page 87

Related tasks


The iSCSI service on a vFiler unitIn general, the iSCSI service operates on individual vFiler units , treating them as though each were aphysical storage system. But the iSCSI software adapter itself (iSCSI software target ), and the commandsthat manage and report on it, and the underlying NICs, operate on the host storage system. This isbecause an iSCSI adapter on a storage system has only one identity (there are no vFiler unit-specificadapter names) and so there is only one set of iSCSI sessions and statistics.

This section lists the dependants of the storage system and the vFiler unit. For more information on theiSCSI service, see the Data ONTAP Block Access Management Guide for iSCSI and FCP.

The following depend on the host storage system:

• The iswt command, which you use to manage the iSCSI service on the storage system’s NICs,operates on the host storage system, not on individual vFiler units . The iSCSI software target driverallows the storage system to be accessed as an iSCSI target device over the storage system’s standardnetwork interfaces.

• If iSCSI is licensed on the host storage system, the iSCSI service is started by default during vFilerunit setup.


• Although most iSCSI subcommands operate specifically on each vFiler unit on which they areexecuted, the iscsi stats command displays statistics by host storage system and iSCSI adapter.

The following considerations are dependant on the individual vFiler unit:

• The iSCSI protocol can be allowed and disallowed, and the iSCSI service can be started and stopped,for each vFiler unit.

• The adapter is online or offline for each vFiler unit, depending on whether the iSCSI service isrunning or stopped on that vFiler unit.

• Each vFiler unit has its own iSCSI node name, which includes the vFiler unit’s UUID.

• Portal groups are defined for each vFiler unit.

• iSCSI subcommands operate specifically on each vFiler unit on which they are executed, exceptfor the iscsi stats command.

• You configure iSCSI security separately for each vFiler unit.This includes setting the default authentication mode: none, deny, or CHAP. For more informationabout CHAP see the Data ONTAP Block Access Management Guide for iSCSI and FCP.

• In the case of CHAP, there is a separate list of initiators and passwords for each vFiler unit.

• You can configure iSNS separately for each vFiler unit.You can use iscsi isns subcommands on each vFiler unit to do the following:

• Configure which iSNS server to use

• Turn iSNS registration on or off

When created, a vFiler unit’s iSNS configuration is in the not configured state, regardless of its stateis on the host storage system.For more information, see the na_iscsi man page.

Networking considerationsTo understand how vFiler units function, you must know how vFiler units operate with routing tables,gateways, and IPsec.

Next topics

Effects of a routed daemon on page 57

Command for changing the routing table in the default IPspace on page 58

The /etc/dgateways file on page 58

IPsec on a vFiler unit on page 58

Effects of a routed daemonEnabling the MultiStore license enables the routed daemon, but only in vfiler0.


When vFiler units are licensed and the routed daemon is on, or when vFiler units are already licensedand routed daemon is turned on, the console displays the following message.

anumita3> routed on Fri Nov 4 22:42:10 GMT[ip.drd.vfiler.info:info]: Although vFiler units are licensed, therouting daemon runs in the default IP space only.

Command for changing the routing table in the default IPspaceBecause vFiler units in the same IPspace share one routing table, you can manipulate the routing tableby entering the route command from the host storage system.

The route command has the following syntax:

route [-fn] add|delete [ host|net ] destination [ gateway metric ]

You can include the command in the storage system /etc/rc file. In this way, the routes are addedautomatically after each storage system reboot.

Related concepts


The /etc/dgateways fileOnly the host storage system has the /etc/dgateways file. vFiler units do not maintain their own/etc/dgateways file.

IPsec on a vFiler unitYou can configure a vFiler unit as an IPsec end point.

You can configure each vFiler unit separately with unique security configuration. When you migratethe vFiler unit from one host storage system to another (or replicate it for disaster-recovery purposes),the IPsec configuration is preserved so long as the vFiler unit’s IP address does not change.

For more information, see the Data ONTAP Network Management Guide and the na_ipsec man page

SnapMirror on the host storage systemThe SnapMirror product for mirroring volumes and qtrees has been integrated to work with vFilertechnology after SnapMirror is licensed on the source and destination storage systems. You can enterSnapMirror commands from the default storage system (vfiler0) or from a specific nondefault vFilerunit. SnapMirror commands entered from the default storage system can be used to affect or displayinformation about all the nondefault vFiler units on the host storage system. SnapMirror commandsentered from a nondefault vFiler unit only affect or display information about that specific unit.


For backward compatibility, the default storage system (vfiler0) can operate on all volumes and qtrees,even if they are owned by other vFiler units.

All volumes and qtrees should be mirrored by either vfiler0 or the host vFiler unit, but not by both.When vFiler unit storage volumes and qtrees are mirrored by vfiler0, make sure SnapMirror is off onthe vFiler unit. If vfiler unit storage volumes and qtrees are mirrored by vfiler0, the SnapMirrorrelationship will be reflected only on vfiler0.

Next topics

Considerations for using the snapmirror command on page 59

Determining the status of SnapMirror relationships on page 60

Considerations for using the snapmirror commandThe procedure for using the snapmirror command for data on a nondefault vFiler unit is the sameas that for data on a storage system; however there are some exceptions.

• Qtree SnapMirror is only supported for qtrees inside volumes that a vFiler unit owns.

• Qtree SnapMirror is only supported if a vFiler unit is rooted on a volume.

• Tape devices are not supported.

• SnapMirror sources and destinations cannot be qtrees in shared volumes.

• Synchronous SnapMirror is not supported.

Additionally, SnapMirror within a MultiStore context has the following features:

• SnapMirror can be turned on and off independently on each vFiler unit.

• The options snapmirror.access, snapmirror.checkip.enable, and snapmirror.enablecan be manipulated independently on each vFiler unit.

• Each vFiler unit has its own snapmirror.conf file in the /etc directory.

• A nondefault vFiler unit can only operate on the volumes and qtrees it owns.

• vFiler units do not require additional SnapMirror licenses. (They use the same license as physicalfilers.)

• SnapMirror relationships established between vFiler units are maintained across vFiler unit migration.

• SnapMirror destination updates are supported on both the host storage system and the vFiler unit.

Note: SnapMirror relationships and all Snapshot copies between vFilers will be destroyed before arevert.

When specifying a path name in the /etc/snapmirror.conf file, be sure to use the storage systemname, not the vFiler unit name. For more information, see the na_snapmirror.conf(5) manual page.


Determining the status of SnapMirror relationshipsOn a vFiler unit, you can display entries related only to that vFiler unit. On the physical storage unit,you can display active transfer entries from all vFiler units . Inactive transfers are only displayed onthe relevant vFiler unit.

Step

1. To display a comprehensive and readable list of SnapMirror transfers, run the following command:

vfiler run * snapmirror status

This command iterates through all vFiler units and lists their transfers.

SnapVault on the host storage systemAfter SnapVault is licensed on the source and destination storage units, the SnapVault product forbacking up volumes and qtrees is integrated to work with vFiler technology.

Next topics

Where to enter SnapVault commands on page 60

Considerations for using the snapvault command on page 60

Determining the status of SnapVault relationships on page 61

Where to enter SnapVault commandsYou can enter SnapVault commands either from the default storage system (vfiler0) or from anynondefault vFiler unit.

Any commands entered on the default unit affect or display information from all the vFiler units on thehost storage system. Commands entered on a nondefault vFiler unit affect or display information aboutonly that specific vFiler unit.

For backward compatibility, the default storage system (vfiler0) can operate on all volumes and qtrees,even if they are owned by vFiler units .

All storage units (volumes and qtrees) should be mirrored from either vfiler0 or the host vFiler unit,not both. When you mirror vFiler storage units from vfiler0, make sure SnapVault is off on the hostvFiler unit.

Considerations for using the snapvault commandWhen the snapvault command is used within the MultiStore context, it shows some limitations andadditional features.


Following are the snapvault command limitations when used in a MultiStore context:

• SnapVault cannot be used in nondefault vFiler units that are rooted in a qtree.

• The SnapVault SnapLock feature is not supported in nondefault vFiler units.

• SnapVault can be turned on and off independently on each vFiler unit.

• The options snapvault.access and snapvault.enable can be manipulated independently oneach vFiler unit.

• Each vFiler unit has its own SnapVault.conf file in the /etc directory.

• Nondefault vFiler units can only operate on the volumes and qtrees they individually own.

• Additional SnapVault licenses are not required; vFiler units use the same source and destinationlicenses as physical storage systems.

• SnapVault relationships established between vFiler units are maintained across vFiler unit migration.

Determining the status of SnapVault relationshipsOn a vFiler unit, the status command displays entries related only to that vFiler unit. On the storagesystem, the status command displays active transfer entries from all vFiler units . Inactive transfersare only displayed on the relevant vFiler unit.

Step

1. To display a comprehensive and readable list of SnapVault transfers, you can run the followingcommand:

vfiler run * snapvault status

This command iterates through all vFiler units and lists their transfers.

SNMP used with a vFiler unitSNMP is not supported on individual vFiler units ; it is supported only on the host storage system. Youcan enable SNMP on the host storage system to collect data about vFiler units.

vFiler unit data from MIBsData about vFiler units can be collected from the standard MIB and from the Data ONTAP customMIB.

In the standard MIB, all vFiler unit data is global. It pertains to the sum of data from all vFiler units onthe storage system, with the following exceptions:

• Statistics related to network interfaces are for the interfaces in the default IPspace.


• TCP statistics include data only from the connections and listen sockets in the default vFiler unit.

• UDP statistics include data only from sockets in the default vFiler unit.

• Quota information is gathered for each volume. If the host storage system or a vFiler unit owns avolume with quotas, quota information is provided for the host storage system or vFiler unit owningthe volume. If a vFiler unit owns qtrees in a volume that it does not own, no quota information isprovided for the vFiler unit.

In the Data ONTAP custom MIB, a group named vFiler is included. It is for information about eachvFiler unit, such as the MultiStore license, IP address, protocols allowed, and so on.

What is not supported on a vFiler unitFCP LUNs are supported only on the host storage system, not on a vFiler unit.

Remember the following limitations when you create LUNs.

• You can create FCP igroups only on the host storage system.

• FCP-connected hosts can access only those LUNs that are owned by the host storage system.

• The fcp command does not recognize vFiler units.

• You can create only iSCSI igroups on vFiler units ; you cannot create FCP igroups on vFiler units.

For detailed information about FCP LUNs, see the Data ONTAP Block Access Management Guide foriSCSI and FCP.

Monitoring performance and statisticsYou can display storage system statistics, NFS statistics, and CIFS statistics to determine how wellyour vFiler units are performing.

Next topics

Displaying storage system statistics on page 62

Displaying uptime statistics on page 63

Displaying NFS statistics on page 63

Displaying CIFS statistics on page 63

Displaying storage system statisticsYou can only display storage system statistics for the sum of statistics generated by all vFiler units ,including vfiler0. You cannot display the statistics for a particular vFiler unit.


Step


sysstat

Displaying uptime statisticsYou can only display the uptime for the storage system. You cannot display the uptime for particularvFiler units.

Step


uptime

Displaying NFS statisticsYou can display NFS statistics for the entire storage system, for specified vFiler units , or for the hoststorage system.

Step

1. The NFS statistics displayed depend on the command format.

Enter the following command from the host storagesystem...

If you want to display statistics from...

nfsstatAll vFiler units together

vfiler run vfilertemplate nfsstatSpecified vFiler units

vfiler run vfiler0 nfsstatThe host storage system

Displaying CIFS statisticsYou can display CIFS statistics for the entire storage system, for specified vFiler units , or for the hoststorage system.

Step

1. The CIFS statistics displayed depend on the command format.


Enter the following command from the host storagesystem...

If you want to display statistics from...

cifs statAll vFiler units together

vfiler run vfilertemplate cifs statSpecified vFiler units

vfiler run vfiler0 cifs statThe host storage system


What IPspaces are

An IPspace defines a distinct IP address space in which vFiler units can participate. IP addresses definedfor an IPspace are applicable only within that IPspace. A distinct routing table is maintained for eachIPspace. No cross-IPspace traffic is routed

Next topics

What to consider for a vFiler unit participation in an IPspace on page 65

Application of IPspaces on page 66

How interfaces participate in an IPspace on page 67

Routing in an IPspace on page 68

Loopback interfaces in IPspaces on page 68

Advantages of using VLAN tagging for IPspaces on page 68

Active/active configuration and IPspaces on page 69

Creating an IPspace on page 71

Listing IPspaces on a storage system on page 72

Removing an IP address from an interface on page 72

Assigning an interface to an IPspace on page 73

Destroying IPspaces on page 73

Creating a vFiler unit in a nondefault IPspace on page 74

What to consider for a vFiler unit participation in an IPspaceThere are some guidelines to remember when assigning an IPspace to a vFiler unit .

• An IPspace can contain multiple vFiler units, but a vFiler unit can belong only to one IPspace.

• Each vFiler unit in an IPspace must have an IP address that is unique within that IPspace, but avFiler unit in one IPspace can have the same IP address as a vFiler unit in a different IPspace.

• Ensure that you assign an IPspace correctly because once you assign an IPspace to a vFiler unit ,you cannot change the assignment without destroying the vFiler unit .

• Each vFiler unit must have one IP address on the interface that leads to the default gateway of theassigned IPspace. This requirement ensures that the vFiler unit is reachable from within the IPspace.

What IPspaces are | 65

Application of IPspacesA typical application for IPspaces is an SSP that needs to connect customers of company A and companyB to a storage system on its premises.

The SSP creates two vFiler units on the physical storage system—one per customer—and provides adedicated network path from one vFiler unit to company A’s network and from the other vFiler unit tocompany B’s network.

This deployment should work if both companies are using non-private IP address ranges; however, thefollowing illustration shows otherwise.

Figure 1: IP interfaces in vFiler units

Both companies use the private IP address subnet 10.0.0.0, causing the following problems:

• The two vFiler units on the storage system at the SSP location have conflicting IP addresses if bothcompanies decide to use the same IP address for their respective vFiler units.

• If the two companies agree on using different IP addresses for their vFiler units, problems still arise:if any client in Company A’s network has the same IP address as a client in Company B’s network,packets destined for a client in A’s address space might get routed to a client in B’s address space,and vice versa.


• Suppose the two companies decide to use mutually exclusive address spaces (for example, CompanyA uses 10.0.0.0 with a network mask of 255.128.0.0 and Company B uses 10.128.0.0 with a networkmask of 255.128.0.0). The SSP needs to configure static routes on the storage system to route trafficappropriately to A’s and B’s networks. This solution is neither scalable (because of static routes)nor secure (broadcast traffic is sent to all interfaces of the storage system).

To overcome these problems, two IPspaces are defined on the storage system—one per vFiler unit .Because a distinct routing table is maintained for each IPspace and no cross-IPspace traffic is routed,the data for each company is securely routed to its respective network even if the two vFiler units areconfigured in the 10.0.0.0 address space, as shown in the following illustration:

Figure 2: IPspace in vFiler units

Additionally, the IP addresses referred to by the various configuration files, such as the /etc/hostsfile, /etc/hosts.equiv file, and /etc/rc file, are relative to that IPspace. Therefore, the IPspacesallow the SSP to configure the same IP address for the configuration and authentication data for bothvFiler units, without conflict.

How interfaces participate in an IPspaceIf the storage system is licensed for vFiler units, all its IP-addressable interfaces, including interfacessuch as vifs and VLAN, belong to the default IPspace. The default IPspace exists automatically andcannot be destroyed.


When you create a new IPspace, you assign interfaces to the new IPspace from the default IPspace. Aninterface can belong only to one IPspace.

Routing in an IPspaceA distinct routing table is maintained for each IPspace. All vFiler units participating in an IPspace shareits routing table.

All packets coming in through an interface are tagged with the IPspace identifier of the IPspace to whichthe interface belongs. The IP address of the interface and the IPspace identifier are used to identify thevFiler unit for which the packet is intended.

All outgoing traffic uses the IPspace identifier of the vFiler unit that is generating the traffic to determinethe routing table to use. Data ONTAP ensures that packets generated by the vFiler units of an IPspaceare transmitted through the interfaces that belong to that IPspace.

Note: Broadcast packets are restricted to the vFiler units within the destination IPspace.

Loopback interfaces in IPspacesEach IPspace has a unique loopback interface assigned to it. The loopback traffic of each IPspace iscompletely isolated from the other IPspaces.

Advantages of using VLAN tagging for IPspacesYou can use VLAN tagging for IPspaces to provide traffic separation for customers, to set up moreIPspaces than there are physical interfaces on the storage system, and to securely deliver packets to avFiler unit in an IPspace.

Next topics

Traffic separation on page 68

More IPspaces than interfaces are allowed on page 69

Secure delivery of packets to a vFiler unit in an IPspace on page 69

Traffic separationVLAN tagging for IPspaces provides traffic separation from each customer to the storage system withoutincurring the cost of additional network devices, such as switches.


Without VLANs, you must provide physically separate network connections to ensure that the trafficfrom each customer is forwarded securely to and from the storage system. This solution is neithercost-effective nor scalable.

With VLAN tagging, you can set up distinct VLANs for each customer on a single switch; thus, VLANtagging provides an alternative to physically separate networks.

More IPspaces than interfaces are allowedDedicating at least one physical interface per IPspace limits the number of IPspaces that can be set upon a storage system to the number of physical interfaces available on the storage system. VLAN taggingovercomes this limitation. VLAN tags can be used to forward traffic to appropriate IPspaces in caseswhere more than one IPspace shares the same physical interface.

Secure delivery of packets to a vFiler unit in an IPspaceVLANs inherently confine the broadcast domains. Therefore, only vFiler units belonging to a VLANreceive broadcasts intended for that VLAN, even if multiple vFiler units share a physical networkinterface.

Active/active configuration and IPspacesYou can use IPspaces in an active/active configuration; however, there are naming and IPspaceassignment requirements.

Next topics

What the IPspace naming requirement is on page 69

IPspace assignment requirement on page 69

What an asymmetric active/active setup is on page 70

Specifying partners in an asymmetric active/active setup on page 70

What the IPspace naming requirement isThe names of IPspaces to which the partner interfaces are assigned must be the same on both storagesystems.

For example, in an active/active configuration of storage system A and storage system B, if IPspaceAis created on storage system A , an IPspaceA must also exist on storage system B .

IPspace assignment requirementThe partner interfaces on both partners must be assigned to IPspaces with the same name on theirrespective storage systems.


For example, in an active/active configuration of storage system A and storage system B , interface e4of storage system B is the takeover interface of interface e0 of storage system A , and vice versa. Ifinterface e0 belongs to IPspaceA on storage system A , interface e4 must belong to IPspaceA on storagesystem B .

What an asymmetric active/active setup isIn an asymmetric active/active setup, vFiler unit-IPspace configuration on one active/active configurationis different from the other; for example, each partner might have a different number of vFiler unitsconfigured in a specific IPspace, or one partner might have no vFiler units.

A “standby” active/active configuration is an example of an asymmetric active/active setup in whichone of the hosts is connected to minimal storage of its own. This host takes over its partner’s storageand vFiler units if the partner fails. In such a configuration, the standby host might not have enoughstorage to support the number of vFiler units that the primary host has.

Specifying partners in an asymmetric active/active setupYou can use the interface name of the partner instead of the IP address to specify the partner interfacewhen setting up such an active/active configuration.

Considerations

For example, two storage systems, storage system1 and storage system2, are configured as shown inthe following table.

Interface nameIPspacevFiler unit (associated IPaddress)

Storage system

e0defaultvfiler0 (1.1.1.1)storage system1

e1ips1vFiler1 (2.1.1.1)

e2ips2vFiler2 (3.1.1.1)

e0defaultvfiler0 (1.1.1.2)storage system2

e4a

e4b


Steps

1. Specify partner interfaces on storage system1 by creating a host-partner relationship using thefollowing commands:

ifconfig e0 1.1.1.1 netmask 255.255.255.0 partner e0

ifconfig e1 2.1.1.1 netmask 255.255.255.0 partner e4a

ifconfig e2 3.1.1.1 netmask 255.255.255.0 partner e4b

2. Specify partner interfaces on storage system2 by completing the following steps:

a) Create two IPspaces: ips1 and ips2.b) Assign interface e4a to ips1 and interface e4b to ips2.c) Create a host-partner relationship on storage system2 in the following way:

ifconfig e0 1.1.1.2 netmask 255.255.255.0 partner e0

ifconfig e4a partner e1

ifconfig e4b partner e2

Creating an IPspaceIPspaces are distinct IP address spaces in which vFiler units reside. You create IPspaces when you needyour vFiler units to have their own secure storage, administration, and routing.

Considerations

You can have a maximum of 101 IPspaces per storage system. Of the 101 IPspaces, one is created bydefault when you install the MultiStore license on your storage system. You can create the remaining100 IPspaces on the storage system.

You can use an alphanumeric string, 1 to 31 characters long, as the IPspace name.

All IPspace names you create on a storage system must be unique. However, the IPspace names onactive/active configuration partners must be the same.

Step


ipspace create ipspacename

ipspacename is the IPspacename that you want to create.

ExampleTo create an IPspace1 on a storage system, enter the following command:


ipspace create ipspace1

Listing IPspaces on a storage systemYou can see which interfaces are assigned to each IPspace by using the ipspace list command.

Step


ipspace list

ExampleIf you enter the ipspace list command on a storage system that has three nondefault IPspaces,you see the following output:

Number of ipspaces configured: 4

default-ipspace (e3)

ipspace1 (e2d)

ipspace2 (e2c)

ipspace3 (e10 e2b sf_vif)

Removing an IP address from an interfaceYou must remove an IP address from an interface before you can assign it to an IPspace

Steps

1. To remove an IP address from an interface, complete one of the following steps:

Then...If the interface is...

Remove the IP address configured for the interface from the vFilerunit as you would when changing resources for a vFiler unit .

Currently configured with an IP addressthat belongs to a vFiler unit in anondefault IPspace

Go to step 2.In the default IPspace

Enter the following command to remove the IPv6 address:

ifconfig interface inet6 -alias IPv6_address

interface is the name of the interface.

IPv6_address is the IPv6 address configured on the interface.

Configured with an autoconfigured IPv6address


Then...If the interface is...

Enter the following command to remove the IP alias:


interface is the name of the interface.

address is the IP address configured for the alias.

Go to the next step.

Configured with an IP alias

Go to step 2.Not configured with an IP alias


ifconfig interface 0.0.0.0

interface is the name of the interface from which you want to remove the IP address.

Related concepts

Effects of adding, removing, and moving vFiler unit resources on page 36

Assigning an interface to an IPspaceTo assign an interface to an IPspace, you must make sure that the interface does not have a configuredIP address. If IPv6 is enabled on your storage system and you bring up an interface, an IPv6 addresswill be automatically configured on the interface.

Step


ipspace assign ipspacename interface-name

ipspacename is the IPspace name to which the interface will be assigned.

interface-name is the name of the interface to be assigned.

You can assign only one interface at a time.

Destroying IPspacesIf you no longer need an IPspace, you can destroy it.


Before You Begin

Ensure there are no network interfaces or vFiler units associated with the IPspace you want to destroy.

Step


ipspace destroy ipspacename

ipspacename is the IPspace name that you want to destroy.

Related tasks

Listing IPspaces on a storage system on page 72


Displaying vFiler unit status on page 46


Creating a vFiler unit in a nondefault IPspaceIf you are a SSP, you might have different clients that require complete network isolation. CreatingvFiler units in separate, or nondefault, IPspaces ensures that the data on these vFiler units is not routedto other vFiler units or the host storage system in other IPspaces.

Considerations

When creating a vFiler unit in a nondefault IPspace, you need to meet the same prerequisites and followthe same guidelines as you would when creating a vFiler unit in the default IPspace.

Steps

1. Create an IPspace, as described in the IPspace creation procedure.

2. Assign an interface to be used by the vFiler unit in the newly created IPspace, as described in theinterface assignment procedure.

3. Verify that each interface to be used by the vFiler unit is ready to be configured, as described in thenetwork interface readiness procedure.

4. To create the vFiler unit , enter the following command:

vfiler create vfiler_name -n -s ipspace -i ip_address [ -i ip_address ]

... path [ path ] ...

ipspace is the IPspace in which the vFiler unit IP addresses reside.

ip_address is an IP address.


path is the complete path name to an existing volume or qtree. The first path name is the storageunit that contains the /etc directory, which contains the configuration information about the vFilerunit.

Note: Ensure you use the -n option of the vfiler create command, and do not use the setupcommand to specify IP addresses for interfaces assigned to different IPspaces. The setupcommand (which runs automatically after the vfiler create command unless you use the -noption) does not allow duplicate IP addresses even if they are for interfaces in different IPspaces.

ExampleThe following command creates a vFiler unit named vFiler1 in the IPspace named ipspace1, usingthe IP address 123.123.123.123 and the /vol/vol1 volume as resources:

vfiler create vfiler1 -n -s ipspace1 -i 123.123.123.123 /vol/vol1

5. Perform one of the following steps:

Go to...For each IP address used by the vFiler that is...

Step 6.An IP alias

Step 7.A base IP address

6. Enter the following command to create the IP alias:

ifconfig interface alias address

If the interface to which you are assigning the alias is currently down, go to Step 7. Otherwise goto Step 8.

7. Use the ifconfig command to configure the interface as Up with the IP address you specified inStep 4.

8. To modify the routing table for the vFiler unit , enter the following command:

vfiler run vfiler_name route add [host | net] [prefixlen prefixlen]

destination gateway metric

ExampleThe following example adds a route to the routing table used by the vFiler unit named vFiler1:

vfiler run vfiler1 route add 1.2.3/24 1.2.3.1 1

9. For each interface used by the vFiler unit , add the following information to the /etc/rc file onthe host storage system:

• An ifconfig command to configure the interface as Up, with the address you specified in thefirst step

• The route command you entered in the previous step

This configures the interfaces as Up and enforces the route commands across reboots.


10. If the host storage system is part of an active/active configuration, edit the /etc/rc file on eachactive/active partner to ensure that each interface the vFiler unit uses has a partner interface definedfor it.

Exampleifconfig e10 partner e10

Note: If you use the storage system setup command to automatically define a partner interface,it will clear all existing vFiler configuration information.

11. The IPspace that the vFiler unit is assigned to must have a default gateway. If the IPspace alreadyhas a default gateway, skip this step. Enter the following command to establish the route to a defaultgateway in the IPspace that the vFiler unit can use:

vfiler run vfiler_name route add default gateway metric

ExampleThe following example adds a route to the default gateway for the IPspace used by the vFiler unitnamed vFiler1:

vfiler run vfiler1 route add default 1.2.3.1. 1

Related concepts

How to create a vFiler unit on page 28

Related tasks

Creating an IPspace on page 71


Ensuring that the network interface is ready on page 32


File Access Using NFS and CIFS

You can access the vFiler unit's file system using NFS or CIFS. To do this, you must prepare the vFilerunit for NFS and CIFS using the NFS and CIFS protocol respectively.

Next topics

How to access the vFiler unit's file system with NFS and CIFS on page 77

How to specify path names for NFS exports or CIFS shares on page 77

How to prepare the vFiler unit for NFS on page 78

How to prepare the vFiler unit for CIFS on page 79

How to access the vFiler unit's file system with NFS and CIFSTo access the vFiler unit's file system with NFS and CIFS, you must prepare the vFiler unit using theNFS and CIFS protocol respectively.

Next topics

How to access a file system with CIFS on page 77

How to access a file system with NFS on page 77

How to access a file system with CIFSFor CIFS clients, the root of the primary storage qtree is the root (“/”) of a vFiler unit’s file system. If“/” is shared, a CIFS client mapping to it can browse all of the vFiler unit’s storage in a single tree. Thismechanism is called the vFiler unit’s “pseudo-root.”

How to access a file system with NFSNFS clients must import discrete storage units as they are defined on the host storage systems.

Pseudo-root directories are not available to NFS clients.

How to specify path names for NFS exports or CIFS sharesWhen you specify a path to export to NFS clients or to share with CIFS clients, use the complete pathname.

File Access Using NFS and CIFS | 77

Example of a path for an NFS export

Suppose a vFiler unit named vfiler1 uses the /vol/vol1 volume for storage. To export the homedirectory at the root of this volume to the clients of vfiler1 , use /vol/vol1/home in the/etc/exports file or in the exportfs command.

Example of a path for a CIFS share

Suppose a vFiler unit named vfiler1 uses the host storage system’s /vol/vol1 volume as itsprimary storage. To share the entire volume, and all other storage owned by the vFiler unit, in asingle tree, specify / as the share path. To offer the home directory at the root of this volume asthe home share, specify /home as the path name for the home share. The vFiler unit mechanismthat makes this possible is known as “pseudo-root.”

Related concepts

How to access the vFiler unit's file system with NFS and CIFS on page 77

How to prepare the vFiler unit for NFSTo prepare the vFiler unit for NFS, you must start the NFS protocol and export discrete storage unitsto NFS clients in the same way they are defined on the host storage system.

However, you must do this only if you want to start NFS manually; if you used the default form of thevfiler create command, you will be prompted for NFS setup information as soon as you create thevFiler unit.

Starting the NFS protocol on page 781.Exporting all file systems in /etc/exports on page 792.

Related tasks


Starting the NFS protocolTo start the NFS protocol on the vFiler unit, use the nfs on command.

Step

1. Complete one of the following steps:

• From the vFiler unit, enter the following command:

nfs on


• From the host storage system, enter the following command:

vfiler run vfiler_name nfs on

The NFS protocol server starts running on the vFiler unit.

Exporting all file systems in /etc/exportsBecause pseudo-root directories are not available to NFS clients, to start using NFS on your vFiler unitNFS clients must import discrete storage units in the same way they are defined on the host storagesystem.

Step

1. Complete one of the following steps:

• From the vFiler unit, enter the following command:

exportfs -a

• From the host storage system, enter the following command:

vfiler run vfiler_name exportfs -a

• From a vFiler unit client that is allowed to connect to the vFiler unit through RSH, enter thefollowing command:

rsh vfiler_name exportfs -a

How to prepare the vFiler unit for CIFSFrom the host storage system, you can use the vfiler run command format to issue cifs commandsfor vFiler units.

From the vFiler units, you use User Manager or Server Manager to manage user accounts and shares.The cifs command is not available through RSH.

Next topics

Tasks that can be performed only on the host storage system on page 80

Per vFiler unit limit with CIFS on page 80

Local user accounts for vFiler units on page 80

Related tasks


File Access Using NFS and CIFS | 79

Tasks that can be performed only on the host storage systemServer Manager does not perform all the functions of the cifs shares -add and cifs shares-change commands. You can execute the commands that Server Manager cannot perform from thevFiler unit command line (by means of the vfiler context command) or from the host storagesystem (by means of the vfiler run command).

The Server Manager cannot perform the following functions:

• cifs shares -add -forcegroup group_name

• cifs shares -add share_name pathname -nosymlink_strict_security

• cifs shares -add -widelink

• cifs shares -add -novscan

• cifs shares -add -novscanread

• cifs shares -change share_name { -forcegroup group_name | -noforcegroup }

• cifs shares -change share_name { -symlink_strict_security |-nosymlink_strict_security }

• cifs shares -change share_name { -widelink | -nowidelink }

• cifs shares -change share_name { -vscan | -novscan }

• cifs shares -change share_name { -vscanread | -novscanread }

Per vFiler unit limit with CIFSData ONTAP does not limit the number of users, shares, open files, and locked files on a per vFilerunit basis.

Local user accounts for vFiler unitsFrom the host storage system, you can use the useradmin command to create local accounts for CIFSusers of each vFiler unit.

Each vFiler unit supports up to 96 local user accounts. The maximum number of vFiler unit user accountsper storage system is 96 times the maximum number of vFiler units for that storage system.

For additional information about using local accounts for CIFS authentication, see the Data ONTAPFile Access and Protocols Management Guide has

For detailed information about managing local user accounts using the useradmin command, see theData ONTAP Storage Management Guide.

Related tasks



Disk Space Management Using Quotas

You can apply user, group and tree quotas on a vFiler unit in the same way that you can on a storagesystem.

Next topics

How to manage quotas on page 81

When quota thresholds and soft quotas are exceeded on page 83

How you can resize quotas on page 84

How the quotas file works on page 84

Displaying quota status on page 84

Displaying a quota report on page 85

How to manage quotasWhen you create a vFiler unit, quotas are automatically turned off on both the host storage system andthe new vFiler unit, for all the volumes you assign to the new vFiler unit, and for all volumes fromwhich you assign qtrees to the new vFiler unit. Quotas are also turned off for volumes that you movefrom one vFiler unit to another. To activate quotas again, you must allow them and turn them on .

Considerations

On a host storage system licensed for vFiler units, the host storage system administrator must allowquotas for a volume before you can turn quotas on and off for the volume. By default, quotas are allowedon all volumes.

Only host storage system administrators can allow or disallow quotas for a volume.

Next topics

Allowing or disallowing quotas for a volume on page 81

Quota specification management on page 82

Turning quotas on or off from a vFiler unit on page 83

Allowing or disallowing quotas for a volumeYou must allow quotas for a volume on the host storage system before you can turn quotas on and offfor the volume. By default, quotas are allowed on all volumes.

Disk Space Management Using Quotas | 81

Step

1. To allow or disallow quotas for a volume, enter the following command:

quota allow | disallow volume

After you enter the quota allow command, you can turn on quotas for the specified volume froma vFiler unit.

After you enter the quota disallow command, vFiler units are prevented from turning quotas onfor the specified volume. If any vFiler units currently have quotas turned on for the volume, quotasare turned off immediately for those vFiler units.

Disallowing quotas on a volume has these effects on all vFiler units that have storage units in the volume:

• If quotas are currently turned off, you or the vFiler unit administrator are prevented from turningquotas on for that volume.

• If quotas are currently turned on, they are turned off immediately and are prevented from beingturned back on.

Quota specification managementThe vFiler unit administrator specifies the size of each quota in the vFiler unit’s /etc/quotas file.That is, the vFiler unit administrator tracks and limits the amount of disk space and the number of fileseach user, group, or qtree uses.

If a qtree owned by the vFiler unit resides on a volume owned by the host storage system, then the hoststorage system administrator, can also specify a quota for the qtree in the host storage system’s/etc/quotas file. The following example shows how the qtree quota on the host storage system affectsa vFiler unit qtree.

How the qtree quota on the host storage system affects a vFiler unit qtree

Suppose the /vol/vol1/qtree1 qtree is a storage unit of the vFiler unit, and the /vol/vol1volume is owned by the host storage system.

In the /etc/quotas file of the vFiler unit, the vFiler unit administrator specifies that this qtreeis limited to 20 GB of disk space. In the /etc/quotas file of the host storage system, you canspecify the disk space limit for the qtree to be 10 GB. This means that if quotas are turned onfrom the host storage system for the /vol/vol1 volume, the qtree cannot exceed the limit ineither of the /etc/quotas files, whichever is lower. In this example, the qtree cannot exceed10 GB.

The host storage system administrator controls the usage of quotas on each volume that the storagesystem owns using the quota allow and quota disallow commands. Once the host storagesystem administrator allows quotas, you can turn quotas on or off on the vFiler units using thequota on and quota off commands, respectively.

For more information on quotas, see the Data ONTAP Storage Management Guide.


Turning quotas on or off from a vFiler unitYou can turn on quotas using the quota on volume command, which activates quotas on the specifiedvolume based on the contents of /etc/quotas. Changes made to /etc/quotas do not take effect the nexttime the quota on or quota resize command is executed. Turning quotas off when using thequota off volume command deactivates the specified volume. You can turn quotas on and off ona per volume basis for a vFiler unit. After you turn quotas on for a particular volume, Data ONTAPinitializes quotas for the storage units residing on that volume, which are owned by the vFiler unit.Quota on or off states are persistent and stay set after reboots.

Step

1. To turn quotas on or off for a volume from a vFiler unit, complete one of the following steps:

Then...If you manage the vFiler unit from...

Enter the following command:

vfiler run vfilertemplate quota on | offvolume

The host storage system

Enter the following command through an RSH connection to thevFiler unit:

quota on | off volume

The vFiler unit

Note: Whenever a qtree is explicitly reassigned to a vFiler unit, you must re-enable the quotamanually if quotas are used. Qtrees are explicitly reassigned to vFiler units when creating vFilerunits (using the vfiler create command) or when you move qtrees between vFiler units (usingthe vfiler move, vfiler add, or vfiler remove commands).

When quota thresholds and soft quotas are exceededWhen a threshold or soft quota defined on a vFiler unit is exceeded, a warning message is logged onthe storage system console.

The following is an example of the warning message:

[[email protected]:notice]: Threshold exceeded for tree 3 on

volume vol1 for vfiler "vfiler1"


How you can resize quotasYou use the quota resize command to cause Data ONTAP to reread the quotas file for the specifiedvolume. Resizing only works for certain types of changes to the quotas file. For other changes, youneed to reinitialize quotas.

How the quotas file worksThe quotas file, found in the /etc directory, contains one or more entries specifying limit or trackingquotas for qtrees, groups, and users. The file can contain default (general) and specific entries.

Displaying quota statusYou can display quota status for any volume on which your vFiler unit owns storage space.

Step

1. To display quota status, complete the following step.

Then...If you manage the vFiler unit from...


vfiler run vfilertemplate quota

The host storage system

Enter the following command through an RSH connection tothe vFiler unit:

quota

The vFiler unit

The command displays quota status information about all volumes in which the vFiler unit ownsstorage space. The following messages are sample messages in the command output:

vol0: quotas are on.vol1: quotas are off.vol2: quotas are disabled.

For more information on quotas, see the Data ONTAP Storage Management Guide


Displaying a quota reportYou display a quota report using the quota report command. By specifying a path name, you candisplay a quota report for all quotas or for a specific file, directory, qtree, or volume.

Step

1. To display a quota report, enter the following command:

quota report

path

You can control the format and fields displayed using the quota report command options. Formore information on the available options, see the na_quota(1) man page.


Disaster recovery and data migration

You can prepare for recovery from a a potential disaster by creating a backup vFiler unit that can beused for disaster recovery. You can copy data from one storage system to another or use vFiler unitsand migrate them by using SnapMover between active/active nodes.

Next topics

How to prepare for a disaster on page 87

Responding to a disaster on page 98

Deleting the disaster-recovery vFiler unit on page 109

How to move (migrate) a vFiler unit on page 109

How to prepare for a disasterBefore a disaster occurs, you can safeguard information by creating vFiler units on a destination storagesystem that remain inactive unless a disaster occurs.

Next topics

How to prepare the destination storage system on page 87

Creating a disaster-recovery vFiler unit on page 95

What the vFiler dr configure command does on page 97

Related tasks


How to prepare the destination storage systemBefore moving (migrating) a vFiler unit or setting up a disaster-recovery vFiler unit, you should performchecks to ensure that the storage system and network are ready.

Next topics

Checking and preparing the storage system on page 88

Storage checklist on page 90

Checking the network on page 90

Network checklist on page 94

Related tasks

Activating the disaster-recovery vFiler unit on page 98

Disaster recovery and data migration | 87

Reactivating the original vFiler unit using vFiler dr commands on page 105

Related references


Checking and preparing the storage system

You must ensure that your destination storage system can support your disaster-recovery vFiler unit.You can use the storage checklist as you are checking and preparing the storage system. However, youdo not need to do the storage checks if you are using SnapMover vFiler unit migration to migrate thevFiler unit.

Check the storage system and record information in the storage checklist as a reference.

Steps

1. Check that the destination storage system has enough storage space to hold the source vFiler unit’svolumes.

On the source storage system, use the vfiler status -r command to see what volumes thevFiler unit is using. Then use the df command on each of those volumes to check how much spaceis being used. The destination volumes must have at least the amount of space in use on the sourcevolumes; run the df command on the destination storage system to check.

Fill in the df results in the storage checklist.

Note: If the source and destination storage systems use different-sized disks and have differentblock sizes, adjust the df numbers accordingly.

2. Make sure there is enough space on the destination volumes.

Then...If...

Go to the next step.There is enough space on the destinationvolumes

1. If necessary, install new disk shelves.2. Use the aggr add command to add new disks to the

destination volumes.

There is not enough space

3. Make sure that the destination storage system has the same volume structure as the source and thatthe volumes to be used by the destination vFiler unit are not used by any other vFiler unit.

The volumes to be used by the destination vFiler unit must have the same path names as those usedby the source vFiler unit.


Then...If the destination storage system...

Go to step 4.Has a volume whose path namesmatches that used by the source vFilerunit and the volumes to be used by thedestination vFiler unit are not used byany other vFiler unit

Do one of the following:

• If the volume is the root volume of the vFiler unit, use thevfiler destroy command to destroy the vFiler unit.

• If the volume can be removed, use the vfiler removecommand to disassociate the volume from that vFiler unit.

• If the volume cannot be destroyed or removed, use the volrename command to rename the volume. Then, create a newvolume with the old name of the volume you just renamed.

Has a volume whose path name matchesthat of the source vFiler unit, howeverthe volume is used by another vFilerunit

Do one of the following:

• For traditional volumes, use the aggr create commandon the destination storage system to create volumes whosenames match those being used by the source vFiler unit, oruse aggr rename to rename a volume.

• For FlexVol volumes, use the aggr create command onthe destination storage system to create volumes whose namesmatch those being used by the source vFiler unit, or use aggrrename to rename a volume. For FlexVol volumes, use thevol create command. For more information, see the DataONTAP Storage Management Guide.

Does not have any path name thatmatches those used by the two choicesabove

4. Make sure there are no qtrees in the destination volumes whose names match those of qtrees in thesource volumes.

The migration process, which is used in the disaster recovery uses SnapMirror. SnapMirror whichwill replicate qtree names from the source to the destination volume, therefore these names shouldnot already exist on the destination.

Then...If there are...

Rename the matching qtrees in the destination volumes.

To rename a qtree, move it from a client the same way youmove a directory or folder. For more information, see theData ONTAP Storage Management Guide.

Qtrees in the destination volumes whose namesmatch those of qtrees in the source volumes

Go to Step 5.No matching qtree names in the destinationvolumes


5. Check whether quotas are being enforced from the host storage system.

Quotas enforced from the vFiler unit will be copied to the new vFiler unit, but quotas enforced fromthe host storage system will not.

To see where quotas are being enforced from, use the host storage system to enter the followingcommand:

quota report

Then...If quotas for qtrees used by the vFiler unit...

Go to the Step 6.Are being enforced

You have completed checking the storage system.Are not being enforced

6. Keep a record of the storage system's /vol/vol0/etc/quotas file for future reference.

7. Copy the relevant entries into the destination storage system's /vol/vol0/etc/quotas file.

Storage checklist

You can use the storage checklist to record storage system information and make sure your systemsare ready to use for disaster recovery.

• How much disk space is used on the source storage system’s volumes?df of source storage system’s volumes:________________________________________

• How much disk space is free on the destination storage system’s volumes?df of destination storage system’s volumes:_____________________________________

• Have you added enough disks to the destination volumes, if required?________

• Do the path names of the source and destination volumes match?_________

• If you are managing qtree-based vFiler units, do any destination volume qtree names match thoseon the source volume? ________

• Have you copied storage system-based quota information from the source to the destination storagesystem’s /etc/quotas file, if possible?__________

Checking the network

Before migrating a vFiler unit or setting up a disaster recovery vFiler unit, you must perform somechecks and actions. Use the network checklist to check if the network supports the disaster-recoveryvFiler unit.


Steps

1. Determine if the destination vFiler unit can take over the source vFiler unit’s IP addresses.

To display information about all their network interfaces, use the ifconfig -a command on thesource vFiler unit and on the destination storage system .

2. You can reuse the source IP addresses and aliases on the destination vFiler unit if the destinationvFiler unit is on the same subnet as the source vFiler unit.

Then...If the source and destinationstorage systems...

Go to Step 3.Are on the same subnet

Note: This is the default caseassumed by the vfilermigrate command.

1. Obtain as many new IP addresses for the destination vFiler unit asare in use on the source vFiler unit.

Note: You might need to replicate subnet-separationarrangements that exist on the source vFiler unit. For example,the source vFiler unit might use one IP address for a servicenetwork and another for an administration network.

2. Make a note of the new IP addresses on the worksheet. The vfilerdr configure command will prompt you for these addresseswhen you create the disaster-recovery vFiler unit.

3. The vfiler migrate command will also prompt you for theseaddresses, but you might then need to run setup on the destinationvFiler unit.

Are on different subnets

3. Check if the source vFiler unit is using the default IPspace.

To display information about IPspaces and the interfaces assigned to them, use the ipspace listcommand on the source vFiler unit.

Then...If...

Go to Step 4.The ipspace list command reportsdefault-ipspace

1. Use the ipspace create command to create acorresponding IPspace with the same name on thedestination storage system.

2. Use the ipspace assign command to assign physicalinterfaces to the IPspace. These interfaces must all beattached to the same physical network.

The ipspace list command reportssomething other thandefault-ipspace


4. Check if the destination vFiler unit has access to the same NIS servers as the source.

Note: You can skip this check if the source and destination vFiler units are on the same subnet.

To see what NIS servers are available to the source vFiler unit, use the nis info command.

Note: The ypwhich command shows to which server the storage system is currently bound.

Then...If...

Go to Step 5.The destination vFiler unit can use thesame NIS servers as the source vFiler unit

Note: This is the default case assumedby the vfiler migrate command.

1. Find out which NIS servers are available to the destinationstorage system.

2. Make a note of the IP addresses of those servers on thenetwork checklist.

The vfiler dr configure command will prompt you forthese addresses when you create the disaster-recovery vFilerunit.

The vfiler migrate command will not prompt you forthese addresses. If you move a vFiler unit to a different subnet,you might need to run setup on the destination vFiler unit.

The destination vFiler unit cannot use thesame NIS servers

5. Check if the destination vFiler unit has access to the same DNS servers as the source.

Note: You can skip this check if the source and destination vFiler units are on the same subnet.

To see what DNS servers are available to the source vFiler unit , use the command dns info on

Then...If...

Go to Step 6.The destination vFiler unit can use the sameDNS servers as the source vFiler unit

Note: This is the default case assumedby the vfiler migrate command.


Then...If...

1. Find out which DNS servers are available to the destinationstorage system.

2. Make a note of the IP addresses of those servers on theworksheet.

The vfiler dr configure command will prompt youfor these addresses.

The vfiler migrate command will not prompt you forthese addresses. If you move a vFiler unit to a different subnet,you might need to run setup on the destination vFiler unit.

The destination vFiler unit cannot use thesame DNS servers

6. Check if the destination vFiler unit will have access to the same WINS servers and the same Windowssecurity network as the source.

Then...If the destination vFiler unit...

Go to Step 7.Can use the same WINS servers andWindows security network as the sourcevFiler unit

1. Find out the name and type (Windows NT 4 or Windows2000) of the domain the destination vFiler unit will be in.

2. Note this information on the network checklist.

When you activate the disaster-recovery vFiler unit, you willneed to configure it into the new domain.

If you move a vFiler unit into a different domain, you will needto configure it into the new domain.

Cannot use the same WINS servers andWindows security network

7. Check if the destination vFiler unit can use the same trusted host for vFiler unit administration asthe sourcevFiler unit.

Then...If the destination vFiler unit...

You are done.Can use the same trusted host as thesource vFiler unit

1. Find out the name of the new trusted host.2. Note this information on the network checklist

You will need to configure the new trusted-host informationafter configuring the disaster-recovery vFiler unit, or aftermoving thevFiler unit.

Cannot use the same trusted host


Related concepts


Related tasks


Adjusting client and network configurations if migrating to a different subnet on page 113


Related references


Network checklist

You can use the network checklist to record network information and make sure your systems are readyto use for disaster recovery.

• Are there enough IP addresses available for the vFiler unit on the destinationnetwork?______________old interface: _______________ new interface:_______________old interface: _______________ new interface:_______________old interface: _______________ new interface:_______________old interface: _______________ new interface:_______________

Note: Check syntax carefully; interface names are case-sensitive.

• Have you created the number of non-default IPspaces, if any are required?

• Have you gathered all the authority servers?old NIS domain: ____________ new NIS domain: ______________old NIS IP address:____________ new NIS IP address:_ _____________old NIS IP address : ____________ new NIS IP address:______________old DNS domain:____________new DNS domain: _____________old DNS IP address : ____________ new DNS IP address:_____________old DNS IP address :____________new DNS IP address: _____________old DNS IP address :____________ new DNS IP address:_____________old WINS IP address : ___________ new WINS IP address:_ ___________old WINS IP address : ___________ new WINS IP address:____________old NT domain type: NT4 W2Kold domain name (FQDN and NetBIOS):_______________________new NT domain type: NT4 W2Knew domain name (FQDN and NetBIOS):_______________________

• Can you use the same trusted host for vFiler unit administration?old trusted host name:_________new trusted host name:___________


Related tasks


Creating a disaster-recovery vFiler unitYou can create a disaster-recovery vFiler unit on a destination storage system that has the storagecapacity, characteristics, and the network connectivity to host an identical copy of the vFiler unit if theoriginal vFiler unit is not able to serve data.

Before You Begin

Before you begin creating the disaster-recovery vFiler unit, ensure the following :

• You have prepared the destination storage system.

• SnapMirror is licensed and enabled on both the source and the destination storage system.

• The source and destination storage systems can communicate with each other over the network (forexample, by means of DNS lookup or entries in the /etc/hosts file).

• The destination volumes are online.

• You know the source storage system’s administrative ID and password.

Considerations

Attention: In the disaster-recovery storage system, protect any volumes that have the same namesas volumes on the original vFiler unit. Otherwise, data in those volumes will be lost.

Steps

1. On the destination storage system, enter the following command:

vfiler dr configure source_vfiler@source_filer

Note:

• If you want to set up synchronous SnapMirror between the source and destination storagesystems, use the -s option of the vfiler dr configure command. For more informationabout the -s option, see the na_vfiler man page.

• The vfiler dr configure command uses the Data ONTAP SnapMirror feature as itsunderlying technology. You can set multiple paths from the source to the destination storagesystems in SnapMirror. The -a option of the vfiler dr configure command enables youto set multiple paths for the configure operation. For more information, see the section onusing SnapMirror over multiple paths in the Data ONTAP Data Protection Online Backupand Recovery Guide.

2. Respond to the login prompt with a valid administrative ID and password for the source storagesystem.


3. Respond to the IP address and binding prompts, using the addresses you entered on the “newinterface” lines on the network checklist worksheet.

4. Respond to the NIS and DNS server prompts, using the addresses you entered in New NIS addr andNew DNS addr in the network checklist sheet.

5. (Optional) Monitor progress using the following command:

vfiler dr status source_vfiler@source_filer

When the vfiler dr status command reports all the storage units of the source vFiler unit asmirrored. The disaster-recovery vFiler unit now exists, but has not been started. Go to Step 6.

Note: The vfiler dr configure command might take some time to complete, especially ifa source qtree has many millions of inodes.

6. If you copied quota information into the destination storage system’s /etc/quotas file, activatethe quotas on that storage system. For each volume you are activating quotas on, use the followingcommand:

quota on volume_name

7. Edit the disaster-recovery vFiler unit’s /etc/hosts.equiv file, adding the name of the trustedhost for administering the disaster-recovery vFiler unit. (This is the host name you entered on thenetwork checklist as the new trusted host.)

Note: If the trusted host is a Windows system, or if it is a UNIX system and the trusted user isnot the root user, you need to add the user name as well; for example,

adminhost joe_smith.

8. Add the path to the root volume and the name of the trusted host to the disaster-recovery vFilerunit’s /etc/exports file.

Example

/vol/vf1_root access=adminhost, root=adminhost

9. If the vFiler unit’s storage units contain iSCSI LUNs, reconfigure iSCSI authentication. Forinstructions, see the Data ONTAP Block Access Management Guide for iSCSI and FCP.

Related concepts

What the vFiler dr configure command does on page 97

Related tasks






Related references


Network checklist on page 94

What the vFiler dr configure command doesThe vFiler dr configure command uses the Data ONTAP SnapMirror feature as its underlyingtechnoloy.

Multiple paths can be set from the source to the destination storage systems in SnapMirror. The -a

option of the vfiler dr configure command enables you to set multiple paths for the configurationoperation.

The vFiler dr configure command does the following:

• Checks that the destination storage system is capable of receiving the source data

• Configures and runs SnapMirror to copy the data from the source to the destination vFiler unit

• iSCSI LUNs (including the LUN maps) are copied from the source vFiler unit to the destinationvFiler unit

• igroups and the iSCSI configuration, including node names and the iSCSI service state, are alsocopied to the destination vFiler unit

• iSCSI authentication is not copied to the destination vFiler unit

• Saves the IP configuration and binding information you supplied when you created thedisaster-recovery vFiler unit

• Saves the NIS and DNS server information you supply

• Saves the quota information from the source vFiler unit's /etc/quotas file

• Causes a baseline transfer to occur from the source to the destination.

• Sets the incremental update interval from the source to the destination to be once every 3 minutes

• Three minutes is the default setting. If you want to change the default setting, edit theetc/snapmirror.conf file as described in the the Data ONTAP Data Protection OnlineBackup and Recovery Guide.

• The vfiler dr configure command automatically configures everything that SnapMirrorrequires for regular updates. No other SnapMirror configuration is necessary. If any SnapMirrorconfiguration requirements are missing from your system (for example, a missing volume orlicense), the vfiler dr configure command returns errors

• Overwrites all data on the volumes of the destination vFiler unit. You must protect any volumes onthe destination storage system that have the same names as the volumes on the source vFiler unit.Otherwise, data in those volumes will be lost.

• Creates a special type of vFiler unit, known as a DR backup vFiler unit, on the destination storagesystem. This vFiler unit is stopped and cannot be started except as described in “Activating thedisaster-recovery vFiler unit.” Before activation, the vFiler unit will respond only to the vfiler


dr delete , vfiler dr status , and vfiler dr resync commands. You should not useifconfig to configure its addresses.

Related tasks





Responding to a disasterAfter a disaster occurs, you will want to keep serving data on the disaster-recovery vFiler unit until youcan reactivate the disaster-recovery configuration.

Next topics


What activating the disaster-recovery vFiler unit does on page 99

How to reactivate the original vFiler unit after damage repair on page 99

Activating the disaster-recovery vFiler unitYou can keep serving data by switching to the disaster-recovery vFiler unit while you determine if youcan recover the original vFiler unit after the disaster.

Steps


vfiler dr activate source_vfiler@source_filer

2. Activating the vFiler unit in the event of a disaster can cause the /etc/hosts.equiv file to beoverwritten. If you specified a different set of DNS or NIS servers for the DR location when youcreated the dissaster-recovery vFiler unit, the existing /etc/hosts.equiv file is overwritten, andthe old file is copied to an /etc/hosts.equiv.bak file.

Then...If you specified...

Copy the /etc/hosts.equiv.bak file to the/etc/hosts.equiv file.

A different set of DNS or NIS servers

Go to Step 3.The same set of DNS or NIS servers

3. To change the name of the Windows domain controller, use the cifs prefdc command.


4. To change the Windows WINS server, use the cifs setup command.

Note: If the Windows domain has changed, you might need to change the permissions on theWindows data files to allow your users the same access they had in the old domain.

5. Make adjustments on the clients, such as remounting volumes and qtrees.

What activating the disaster-recovery vFiler unit doesWhen you activate a disaster-recovery vFiler unit, a number of activities occur that replaces the originalvFiler unit with the disaster-recovery vFiler unit.

The vfiler dr activate command does the following:

• Breaks the SnapMirror relationships between the source and destination storage systems

• Converts the vFiler unit into an ordinary vFiler unit that can be started and that responds to all thecommands vFiler units support

• Brings LUNs online

• Configures IP bindings on the destination vFiler unit according to the information you provided tothe vfiler dr configure command, adding the destination IP information to the destination/etc/rc file. (Any IP information that pertains only to the source vFiler unit is removed from thedestination /etc/rc file.)

• Configures the NIS and DNS servers according to the information you provided to the vfiler drconfigure command.

• Configures any quota information saved by the vfiler dr configure command

How to reactivate the original vFiler unit after damage repairYou can re-create the disaster-recovery configuration using the original storage system if it is notdamaged but suffers only a temporary failure; or you can use a different storage system if the originalstorage system was severely damaged.

Next topics

Resynchronizing the vFiler unit on page 99

Handling resynchronization failures on page 102

Reactivating the original vFiler unit using SnapMirror commands on page 103


Re-creating the vFiler unit on a replacement storage system on page 107

Resynchronizing the vFiler unit

You can resynchronize your original vFiler unit with the currently activated disaster-recovery vFilerunit before bringing up the original vFiler unit. This method not only saves you from deleting theoriginal vFiler unit and creating a new vFiler unit, but also does not require a baseline transfer (involving


a large amount of data), which would otherwise occur between the new vFiler unit and thedisaster-recovery vFiler unit.

You can resynchronize the vFiler unit if the following prerequisites are met:

• The storage elements for the vFiler unit are only volumes (traditional or FlexVol) and not qtrees.

• The source and destination vFiler units contain identical volumes.

• The size of volumes on the source and the destination vFiler units is the same.

• The vFiler unit from which you are updating is activated.

• The original vFiler unit is not in the process of migration.

• If new storage elements have been added to a disaster-recovery activated vFiler unit, make sure thatthe newly added storage elements exist on the original storage system also.

Attention:

On the storage system on which you are resynchronizing the original vFiler unit, protect any volumesthat have the same names as volumes on the disaster-recovery vFiler unit.

If a volume with the same name exists, the volume is automatically added and initialized forSnapMirror transfers from the disaster-recovery vFiler unit. Any existing data on the newly addedvolume is lost.

Steps

1. Ensure that the original vFiler unit you are resynchronizing is in a stopped state.

If not, enter the following command to stop the vFiler unit:

vfiler stop vfilername

2. On the original storage system, enter the following command:

vfiler dr resync [-l authinfo] [-a alt-remote, alt-local] [-s]

dr_vfilername@disaster_recovery_filer

authinfo is the authentication information specified in the username:password format. username

is the login name of the administration host on the disaster recovery storage system and password

is the password for that user name. If you do not specify the authentication information in the vfilerdr resync command, you are prompted for it when you run the command.

alt-remote is the alternate host name or IP address of the source (the disaster recovery storagesystem, in this case).

alt-local is the alternate host name or IP address of the destination (the original storage system,in this case).

-s enables you to set up a synchronous SnapMirror relationship between the source and destinationstorage systems.


dr-vfilername is the name of the disaster-recovery vFiler unit that is currently in the activatedstate.

disaster_recovery_filer is the name of the storage system on which the currently activateddisaster-recovery vFiler unit exists.

Note:

• The vfiler dr resync command resynchronizes all storage elements belonging to thedisaster-recovery vFiler unit, including the volumes that were added to the disaster-recoveryvFiler unit after it was activated.

• When you run the vfiler dr resync command on the disaster-recovery vFiler unit toresync it with the original vFiler unit, you do not need to specify the -l, -a, and -s options.The values stored when the vfiler dr configure command was run (for a baselinetransfer) to back up the original vFiler unit are used.

3. After the resync operation is complete, enter the following command on the storage system on whichthe original vFiler unit exists to check the status of the vFiler unit that was resynchronized:

vfiler status -r original_vfilername

The original vFiler unit will be in a stopped, DR backup state. This is because the vfiler drresync command does not activate the vFiler unit upon resynchronizing. The vFiler unit continuesto behave like a backup disaster-recovery vFiler unit until you use the vfiler dr activatecommand to reactivate it.

4. Reinstate the disaster-recovery vFiler unit on the disaster-recovery storage system.

Then...If the storage elements forthe vFiler unit are...

On the disaster-recovery storage system, enter the following command:

vfiler dr resync [-l authinfo] [-a alt-remote,alt-local] [-s] original_vfilername@original_filer

Only volumes

1. On the disaster-recovery storage system, destroy the disaster-recovery vFilerunit by entering the following command on the disaster-recovery host storagesystem:

vfiler destroy vfilername

vfilername is the name of the original vFiler unit.

2. On the disaster-recovery storage system, re-create the disaster-recoveryvFiler unit.

qtrees

Related tasks








Handling resynchronization failures

You can take some corrective action if the vfiler dr resync operation is interrupted or does notcomplete.

Steps

1. You take different corrective actions depending on the error you get.

Then...If...

1. Make the volume online or create it.2. Go to Step 4.

There was a “volume offline or does not exist”error

Perform the steps to reactivate the original vFiler unitusing SnapMirror commands.

There were volume resync errors

Go to Step 2.There were no volume resync errors

2. Enter the following command to check if the vFiler unit you were resynchronizing exists on thestorage system on which you were running the vfiler dr resync command:

vfiler status

3. If the vFiler unit exists, enter the following command to destroy it:


4. Enter the following command to re-create the vFiler unit you destroyed earlier:

vfiler create -r vfilername pathname

5. Enter the following command to resynchronize the vFiler unit:

vfiler dr resync [-l authinfo] [-a alt-remote, alt-local] [-s]

dr_vfilername@disaster_recovery_filer

Related tasks

Reactivating the original vFiler unit using SnapMirror commands on page 103



Reactivating the original vFiler unit using SnapMirror commands

If the storage system was not damaged but failed temporarily and if the storage element associated withthe vFiler unit is a qtree, you can reactivate the vFiler unit using SnapMirror commands.

Steps

1. Boot the original storage system and interrupt the boot process by pressing the Del or Esc key whilethe memory self-test is in progress.

Note: If you do not press the Del or Esc key in time, you can press Ctrl-C when prompted laterduring the boot; choose option 5 (maintenance mode); and enter the halt command.

2. At the OK prompt, set the no-vfiler-ips? variable as follows:

setenv no-vfiler-ips? true

This ensures that the storage system does not try to bind IP addresses already being used by thedisaster-recovery vFiler unit. When the storage system boots, the original vFiler unit starts running,but it does not accept any read or write requests because its interfaces are not configured.

3. Resynchronize the mirrored volumes and qtrees.

For each volume and qtree owned by the original vFiler unit, enter the following command on theoriginal storage system that you are trying to activate:

snapmirror resync -S disaster_recovery_filer:/pathnameoriginal_filer:/pathname

Examplesnapmirror resync -S drfiler:/vol/vfiler1/qtree1 prfiler:/vol/vfiler1/qtree

If the snapmirror resync command fails, and informs you that there are no matching Snapshotcopies, you might have accidentally deleted the Snapshot copies that SnapMirror depends on, andto initialize SnapMirror using the snapmirror initialize command (see the na_snapmirrorman page for details), then go to Step 4.

Note: If the snapmirror resync command fails, and informs you that there are no matchingSnapshot copies , then skip Step 7.

4. Find out if the snapmirror.access option on the disaster-recovery storage system is set to legacy.Enter the following command on the disaster-recovery storage system:

options snapmirror

Then...If the options snapmirror commandreturns...

Edit the /etc/snapmirror.allow file and add the hostname of the original storage system if it is not already there.

snapmirror.access legacy


Then...If the options snapmirror commandreturns...

Use the options snapmirror command to add the hostname of the original storage system. For example,

options snapmirror.access host=fridge,toaster,prfiler

A list of host names, and the name of theoriginal storage system is not in the list

5. Run the setup command on the disaster-recovery vFiler unit and unconfigure its IP addresses.

6. Update the data on the original vFiler unit.

For each volume and qtree owned by the original vFiler unit, enter the following command on theoriginal storage system:

snapmirror update -S disaster_recovery_filer:/pathnameoriginal_filer:/pathname

Example

snapmirror update -S drfiler:/vol/vfiler1/qtree1 prfiler:/vol/vfiler1/qtree1

7. Stop SnapMirror transfers to the disaster-recovery vFiler unit.


snapmirror quiesce pathname

Example

snapmirror quiesce /vol/vfiler1/qtree1

Note: This operation can take a long time. Use Ctrl-C to interrupt it if you need to.

8. Check that all the paths are quiesced by entering the following command:

snapmirror status

The status column in the output should show each path as Quiesced.

9. Break the SnapMirror relationship.


snapmirror break pathname


Example

snapmirror break /vol/vfiler1/qtree1

10. On the original vFiler, run the setup command to configure the vFiler unit’s IP addresses andconfigure the NIS and DNS servers.

11. If the storage units that have been copied contain iSCSI LUNs, check that the iSCSI configurationon the original vFiler unit is intact and still makes sense.

You might need to remap the LUNs and re-create initiator groups (igroups).

12. If the storage units that have been copied contain iSCSI LUNs, bring the LUNs back online on theoriginal vFiler unit.

13. Stop the disaster-recovery vFiler unit. On the disaster-recovery storage system, enter the followingcommand:

vfiler stop vfilername

vfilername is the name of the disaster-recovery vFiler unit.

You have completed reactivating the original storage system. To ensure that you have a disasterrecovery vFiler unit available for the vFiler unit on the reactivated original storage system, go toStep 14.

14. Resynchronize the disaster-recovery vFiler unit.

From the disaster-recovery storage system, perform the procedure for resynchronizing the originalvFiler unit.

Related tasks


Reactivating the original vFiler unit using vFiler dr commands

If the storage system was severely damaged, or you are not comfortable using SnapMirror commands,you can reactivate the vFiler unit using vFiler dr commands.

Steps

1. Boot the original storage system and interrupt the boot process by pressing the Del or Esc key whilethe memory self-test is in progress.

Note: If you do not press the Del or Esc key in time, you can press Ctrl-C when prompted laterduring the boot; then choose option 5 (maintenance mode); then enter halt.

2. At the OK prompt, set the no-vfiler-ips? variable as follows:


setenv no-vfiler-ips? true

This ensures that the storage system does not try to bind IP addresses already being used by thedisaster-recovery vFiler unit.

3. At the OK prompt, enter the boot command.

4. Destroy the original vFiler unit. Enter the following command on the original storage system:



5. Stop the disaster-recovery vFiler unit by using the vfiler stop command.

6. Prepare for the new vFiler unit on the original (or its replacement) storage system using thepreparation steps for the destination storage system

7. Update the data on the original storage system:

For each volume and qtree owned by the new vFiler unit, enter the following commands on theoriginal storage system (or its replacement)

snapmirror break name

snapmirror resync -S disaster_recovery_filer:name production_filer:name

disaster_recovery_filer is the name of the disaster recovery. storage system

production_filer is the name of the original storage system.

name is the volume name or path name of the qtree.

ExampleFor volume vol1, enter the following commands:

snapmirror break drfiler:vol1

snapmirror resync -S drfiler:vol1 prfiler:vol1

ExampleFor qtree qtree1, enter the following commands:

snapmirror break drfiler:/vol/vol2/qtree1

snapmirror resync -S drfiler:/vol/vol2/qtree1 prfiler:/vol/vol2/qtree1

8. Create a new vFiler unit on the original storage system(or its replacement) , following the procedurefor creating a vFiler unit.

The original vFiler unit is now reactivated on the original storage system or its replacement.



Then...If the storage elementsfor the vFiler unit are...



Only volumes




2. On the disaster-recovery storage system, re-create the disaster-recovery vFilerunit.From the original storage system or its replacement, perform the procedurefor creating the disaster-recovery vFiler unit.

Qtrees

Related tasks




Re-creating the vFiler unit on a replacement storage system on page 107

Re-creating the vFiler unit on a replacement storage system

You can re-create the original vFiler unit on a replacement storage system if the original storage systemis damaged beyond repair.

Steps

1. Boot the replacement storage system.

2. Stop the disaster-recovery vFiler unit by using the vfiler stop command.

3. Prepare for the new vFiler unit on the original (or its replacement) storage system using thepreparation steps for the destination storage system

4. Update the data on the original storage system:

For each volume and qtree owned by the new vFiler unit, enter the following commands on theoriginal (or its replacement) storage system:

snapmirror break name

snapmirror resync -S disaster_recovery_filer:name production_filer:name

disaster_recovery_filer is the name of the disaster recovery storage system


production_filer is the name of the original storage system

name is the volume name or path name of the qtree

ExampleFor volume vol1, enter the following commands:

snapmirror break drfiler:vol1

snapmirror resync -S drfiler:vol1 prfiler:vol1

ExampleFor qtree qtree1, enter the following commands:

snapmirror break drfiler:/vol/vol2/qtree1

snapmirror resync -S drfiler:/vol/vol2/qtree1 prfiler:/vol/vol2/qtree1

5. Create the new vFiler unit on the original (or its replacement) storage system, following the procedurefor creating the vFiler unit.

The original vFiler unit is now reactivated on the original storage system or its replacement.


Then...If the storage elementsfor the vFiler unit are...



Only volumes




2. On the disaster-recovery storage system, re-create the disaster-recovery vFilerunit.From the original storage system or its replacement, perform the procedureto creating the disaster-recovery vFiler unit.

Qtrees

Related tasks








Deleting the disaster-recovery vFiler unitYou can delete the disaster-recovery vFiler unit at any time after setting it up.

Step


vfiler dr delete source_vfiler@source_filer

Before removing the disaster-recovery vFiler unit, the vfiler dr delete command also removesall SnapMirror relationships, and any other configuration information related to the disaster-recoveryvFiler unit, from the source vFiler unit.

If any errors are detected in SnapMirror relationships, the deletion of the vFiler unit is aborted. Toignore SnapMirror errors and remove the disaster-recovery vFiler unit, you can use the -f optionavailable in the vfiler dr delete command.

To remove a disaster-recovery vFiler unit on the destination storage system "StorageSystem 2"created for a vFiler unit "vfiler1" on source storage system "StorageSystem1" even if SnapMirrorerrors exist, enter the following command on "StorageSystem 2":

vfiler dr delete -f vfiler1@StorageSystem1

How to move (migrate) a vFiler unitIf you want to move (migrate) a vFiler unit from a storage system that is overloaded to one that hasspare capacity, or from an old storage system to a new one, you can do so by either copying data froma source vFiler unit to a destination vFiler unit or by using SnapMover data migration technology.

Next topics

Prerequisites for migrating a vFiler unit by copying data on page 110

How migrating a vFiler unit affects clients on page 110

What the vFiler unit migrate commands do on page 111

Migrating thevFiler unit by copying data on page 112


What SnapMover vFiler unit migration is on page 114

Requirements for vFiler unit migration between active/active nodes on page 115


Guidelines for setting up volumes to support vFiler unit migration on page 115

What SnapMover vFiler unit migration does on page 116

How to migrate the vFiler unit between active/active nodes with SnapMover on page 116

Disabling SnapMover vFiler unit migration on page 118

Related tasks





Migrating thevFiler unit by copying data on page 112


Prerequisites for migrating a vFiler unit by copying dataEnsure that the following before you begin the migration.

• You performed all the storage checks and actions and the network checks and actions.

• You licensed and enabled SnapMirror on both the source and the destination storage systems.

• Preferably, the source and destination storage systems are on the same subnet. If they are not on thesame subnet, you need to configure authority servers on the new vFiler unit after the migration

• The source and destination storage systems can communicte with each other over the network ( forexample, using a DNS lookup or entries in the /etc/hosts file).

• If any of the source qtrees or volumes contain LUNs, the destination storage system is running aversion of Data ONTAP that supports LUNs.

• If the source vFiler unit owns iSCSI LUNs, the destination vFiler unit must be running a version ofData ONTAP that supports iSCSI LUNs on vFiler units.

• The destination volumes are online and writable.

• You know the source storage system's administrative ID and password.

Related tasks




How migrating a vFiler unit affects clientsMigrating a vFiler unit to another storage system on the same subnet has the following affects.

• NFS volume-level mounts move transparently.

• NFS qtree-level mounts must be remounted.

• CIFS clients experience the equivalent of a reboot.


• LUN access continues uninterrupted.Although an iSCSI host is briefly disconnected from the source vFiler unit, an initiator hides thisbrief disruption from applications accessing the LUNs. To the applications, access continues withoutinterruption.

Related tasks




What the vFiler unit migrate commands doMigration uses two vFiler migrate commands, vfiler migrate start and vfiler migratecomplete.

The vfiler migrate start command does the following:

• Checks that the destination storage system is capable of receiving the source data

• Configures and runs SnapMirror to copy the data from the source to the destination vFiler unit

• Saves the IP configuration and binding information you supplied when you created thedisaster-recovery vFiler unit


The vfiler migrate complete command does the following:

• Stops the source vFiler unit

• Updates the data on the destination vFiler unit

• Breaks the SnapMirror relationships

• Configures IP bindings on the destination vFiler unit according to the information you provided tothe vfiler migrate start command, adding the destination IP information to the destination/etc/rc file.Any IP information that pertains only to the source vFiler unit is removed from the destination/etc/rc file

• Configures any quota information saved by the vfiler migrate start command

• Destroys the source vFiler unit

• Brings LUNs online using migrated LUN maps and igroups.

Related tasks



Migrating thevFiler unit by copying dataYou might want to copy data from one storage system to another to maintain a backup copy of the dataso that if the primary disks fail or the data on them becomes inaccessible, you can use the backup copyto restore services to clients.

Before You Begin

Ensure that you perform all of the steps listed in the storage checklist and network checklist.

Considerations

Attention: This procedure destroys the original vFiler unit after replicating it on the destinationstorage system. To leave the original vFiler unit intact and create its backup copy for disaster-recoverypurposes, use the procedure for creating a disaster-recovery vFiler unit.

Steps

1. Qualify and prepare the destination storage system using the storage checklist and network checklist.

2. On the destination storage system, enter one of the following commands:

Then...If the time the console is locked ...


source_vfiler @source_filer

Concerns you, (for example, if you want to lockthe console for a minimum amount of time)

vfiler migratesource_vfiler@source_filer

Note: Skip Step 5 and Step 6 if you use thiscommand. For more information, see the na_vfilerman page.

Is of no concern to you (for example, if you wantto migrate the vFiler unit overnight )

3. Respond to the login prompt with a valid administrative ID and password for the source storagesystem.

4. Respond to the IP address and binding prompts, using the addresses you entered on the “newinterface” lines on the worksheet.

5. Monitor the progress of the migration with the fillowing command:

vfiler migrate status source_vfiler@source_filer

Note: The vfiler migrate command might take some time to complete, especially if a sourceqtree has many millions of inodes.


6. When the status command reports that SnapMirror has replicated all the storage units of the sourcevFiler unit, you can either complete the migration or cancel the migration.

Then...If you want to...


vfiler migrate complete source_vfiler@source_filer

Note: If the vfiler migrate complete command reports that theprocess cannot complete, wait a few minutes and then enter the same vfilermigrate complete command again.

Complete the migration


vfiler migrate cancel source_vfiler@source_filer

This destroys the destination vFiler unit and removes the SnapMirror and othermigration-related configuration information from the source vFiler unit.

Cancel the migration,

7. If you copied quota information into the destination storage system’s /etc/quotas file when youprepared the destination storage system, activate the quotas on that storage system. For each volumeyou are activating quotas on, use the following command:

quota on volume_name

8. Make adjustments on the clients.

You will need to remount qtrees, but volume-level mounts should be intact. If you have moved thevFiler unit to a different subnet or Windows domain, you need to configure the network servers andmake adjustments on the clients and you need to modify data-file security attributes.

9. Reconfigure iSCSI authentication if the vFiler unit’s storage units contain iSCSI LUNS,

See the Data ONTAP Block Access Management Guide for iSCSI and FCP for instructions.

Related tasks






Adjusting client and network configurations if migrating to a different subnetIf you have moved the vFiler unit to a different subnet, you might need to configure the network serversand make adjustments on the clients. If you have moved the vFiler unit to a different Windows domain,you might also need to modify data-file security attributes.


Considerations

Perform the following steps on the destination vFiler unit.

Steps

1. To configure NIS and DNS servers, run setup.

2. To change the name of the Windows domain controller, use the cifs prefdc command.

3. To change the Windows WINS server, run the cifs setup command.

Note: If the Windows domain has changed, you might need to change the permissions on theWindows data files to allow your users the same access they had in the old domain.

4. To change the trusted host, do the following:

a) Edit the vFiler unit’s /etc/hosts.equiv file, adding the name of the trusted host foradministering the vFiler unit. (This is the hostname you entered on the worksheet as “new trustedhost.”)

b) Add the path to the root volume and the name of the trusted host to the vFiler unit’s /etc/exportsfile.

Example

/vol/vol0 access=adminhost, root=adminhost

5. Make adjustments on the clients.

These include remounting volumes and qtrees.

What SnapMover vFiler unit migration isSnapMover vFiler unit migration on an active/active system is the no-copy transfer of a volume-levelvFiler unit from one host node to the other.

For example, if heavy traffic to one vFiler unit is affecting the performance of its host node, and theother active/active configuration node is lightly loaded, you can transfer ownership of that vFiler unitto the host node’s active/active configuration partner to balance the load processing on the two nodes.

The SnapMover feature uses software-based disk ownership to transfer ownership of the aggregate thatcontains the vFiler unit from the original host node to its partner. Because the SnapMover feature carriesout the vFiler unit migration through transfer of disk ownership rather than by copying data from oneset of disks to another, the migration operation is quickly completed.


Requirements for vFiler unit migration between active/active nodesEnsure that the following points are true before you use SnapMover to migrate a vFiler unit betweenactive/active nodes.

• The MultiStore license and the SnapMover license must be installed on each active/active node.

• Other licenses used by the vFiler unit need to match. For example, if CIFS is licensed on the sourceactive/active node, it must also be licensed on the destination active/active node. Otherwise, movingthe vFiler unit causes CIFS to be unavailable for that vFiler unit after it is moved.

• Both the source active/active node and the destination active/active node must be connected to thesame storage subsystem. The disks must be visible to both the source and destination active/activenodes.

• To ensure that software-based disk ownership changes are transparent to NFS users, the destinationactive/active node must have an Ethernet connection to the same subnet that the source active/activenode uses.

• The volumes assigned to the vFiler unit can be either traditional volumes or flexible volumes. If thevolumes are flexible volumes, the containing aggregate must only contain volumes belonging tothe migrating vFiler unit. For information about traditional and FlexVol volumes, see the DataONTAP Storage Management Guide.

• The vFiler unit’s storage units must all be composed of complete volumes — that is, the vFilerunit’s paths must use the form /vol/volname. SnapMover migration of storage units namingspecific volume subdirectories—for example, /vol/volname/directory or /vol/volname/qtree,is not supported.

• The volume containing the configuration information for the vFiler unit (/vol/volname) must bewritable.

Guidelines for setting up volumes to support vFiler unit migrationWhen a vFiler unit is migrated, all volumes associated with that vFiler unit are moved. Therefore, whenyou create the vFiler units, consider the following points.

• SnapMover cannot migrate just a subset of the volumes that are managed by the vFiler unit it ismigrating.

• The destination active/active node must be able to accommodate the size of the vFiler unit with allits associated volumes.

• You can create up to 64 vFiler units on a storage system.

• The names of the vFiler units and volumes being moved from the source to the destination must beunique on the destination.Although you can rename a volume, it is best not to do so if you have NFS clients, because therenaming of the volume is not transparent to NFS clients. When the storage system uses NFS toexport a file system, the volume name is part of the exported path name. NFS clients try to mountusing the old path name; to access the data after the vFiler unit has been moved, clients will needto remount using the new path name.


What SnapMover vFiler unit migration doesYou use the vfiler migrate -m nocopy command to carry out SnapMover vFiler unit migration.

This command does the following:

• Ensures that no vFiler unit with the same name exists on the destination node

• Ensures that the SnapMover license is installed on both the source and destination active/activenodes

• Ensures that the licenses and Data ONTAP versions are the same on both the source and destinationactive/active nodes

• Saves the IP configuration and binding information that you supplied when you created the vFilerunit


• Stops the source vFiler unit

• Destroys the source vFiler unit

• Rewrites the disk ownership information so that the ownership of the vFiler unit volumes passesfrom the source active/active node to the destination active/active node

• Re-creates the vFiler unit on the destination active/active node

How to migrate the vFiler unit between active/active nodes with SnapMoverIf heavy traffic to one vFiler unit is affecting the performance of its host node, and the other active/activeconfiguration node is lightly loaded, you can transfer ownership of that vFiler unit to the host node'sactive/active configuration partner to balance the load processing on the two nodes.

Enabling SnapMover vFiler unit migration on page 1161.Performing a vFiler unit migration on volumes on page 1172.

Enabling SnapMover vFiler unit migration

You must reconfigure the active/active configured storage system and license SnapMover before youcan migrate a vFiler unit.

Steps

1. On each active/active configuration node, confirm that MultiStore is licensed. At the commandprompt, enter the following command and make sure that MultiStore is listed as a licensed feature:

license

2. Temporarily disable the active/active configuration by entering the following command:

cf disable


3. On each active/active configuration node, confirm that the software-based disk ownership is enabledby entering the following commands in the command-line interface of one of the active/activeconfiguration nodes:

disk show

• If the system displays disks in the active/active configuration to which disk ownership is assigned,then software-based disk ownership is enabled. Go to Step 10.

• If the system does not displays disks in the active/active configuration to which disk ownershipis assigned, then go to Step 4.

4. Reboot each active/active configuration node. During the boot process, press Ctrl-C to display theboot menu options.

5. Enter the choice for booting in maintenance mode.

6. On each active/active configuration node, in maintenance mode, enter the following command:

disk upgrade-ownership

This command writes software-based disk ownership information to enable SnapMover.

7. Enter the following command on each active/active configuration node to confirm the newsoftware-based disk ownership scheme:

disk show

The system displays all the disks on the active/active configuration to which ownership has beenassigned. Disks assigned to either active/active configuration node will be visible.

8. Halt each active/active configuration node to exit maintenance mode, by entering the followingcommand:

halt

9. Reboot each active/active configuration node in normal mode.

10. On each active/active configuration node, install the license for SnapMover:

license add snapmover_license

11. Reenable the active/active configuration operation, by entering the following command:

cf enable

Performing a vFiler unit migration on volumes

You can balance the load processing on an active/active configured system by using SnapMover tomigrate a vFiler unit from one active node to the other.


Steps

1. On the destination active/active configuration, enter the following command:

vfiler migrate -m nocopy vfilername@source_cl_partner

vfilername is the name of the vFiler unit that you are migrating.

source_cl_partner is the active/active configuration from which you are moving the vFiler unit.

For detailed information, see the na_vfiler(1) man page.

2. Answer the prompts, including the following information:

• A valid administrative login ID and password

• The IP address and binding information for the destination vFiler unit

The vFiler unit is migrated from the source active/active configuration to the destination active/activeconfiguration .

3. Verify that the vFiler unit was moved by entering the following command on the destinationactive/active configuration :

vfiler status -r vfilername

Disabling SnapMover vFiler unit migrationIf you decide that the load on the active/active system is too high and you want to physically migratevFiler units to another storage system, you need to disable the SnapMover vFiler unit migration.

Steps

1. On one of the active/active nodes, enter the following command to ascertain whether or not thecurrent software-based disk ownership assignments align with the active/active configuration ’s Aloop B loop topology:

aggr status -r

For backward compatibility, you can also enter the following command:

vol status -r

In the output of this command, all disks assigned to the current active/active node are listed asvolume disks or as spare disks; all disks assigned to the active/active configuration ’s partner arelisted as partner disks.

All disks on the same shelf must be assigned to the same active/active configuration node and alldisk shelf assignments must be consistent with the A loop B loop topology of the active/activeconfiguration .

2. You have to change the disk ownership if it deviates for the A loop B loop topology.


Then...If the disk ownership...

Change it back with the vfiler migrate -mnocopy command using the procedure you usedwhen migrating the vFiler unit, then go to Step 3.

Deviates from the A loop B loop topology becauseyou migrated a vFiler unit with the vfilermigrate -m nocopy command.

Modify the disk assignment using the diskassign command to assign all disks on the shelfto the active/active configuration node attached tothe shelf’s A loop connection then, go to Step 3.

Deviates from the A loop B loop topology becauseyou split ownership of a single expansion disk shelfbetween the two active/active configuration nodes.

Modify the disk assignment using the diskassign command to change it back, then go to Step3.

Deviates from the A loop B loop topology becauseyou reassigned a spare disk with the disk assigncommand.

Go to Step 3.Reflects the active/active configuration ’s A loop Bloop topology.

3. Disable active/active configuration by entering the following command:

cf disable

4. On each active/active configuration node, uninstall the license for SnapMover:

license delete snapmover

5. Reboot each active/active configuration node, and when prompted, press Ctrl-C to display the bootmenu options.

6. On each active/active configuration node, enter the choice for booting in maintenance mode.

7. In the maintenance mode on each active/active configuration node, enter the following command:

disk remove_ownership

This command disables software-based disk ownership.

8. Halt each active/active configuration node to exit maintenance mode:

halt

9. Reboot each active/active configuration node in normal mode.

The system reverts to ownership based on A loop and B loop connections.

10. Re-enable the active/active configuration operation by entering the following command:

cf enable

Related tasks



Virus protection for CIFS

You can perform virus scanning for vFiler units that run the CIFS protocol.

Next topics

Who can configure virus scanning on page 121

Storage systems with which virus scanners can be registered on page 121

Virus scanning for vfiler0 on page 121

Requirements for virus scanning on a vFiler unit other than vfiler0 on page 122

Effect of virus scanner availability on CIFS access on page 122

Configuring virus scanning for a vFiler unit on page 122

Who can configure virus scanningThe host storage system administrator) can specify how virus scanning works on files owned by thehost storage system and files owned by vFiler units. vFiler unit administrators, however, cannot configurevirus scanning for vFiler units.

Storage systems with which virus scanners can be registeredVirus scanners can be registered with the host storage system or any vFiler unit. You can determinewhether files on a vFiler unit are scanned by virus scanners registered with the host storage system orthe vFiler unit.

A virus scanner local to a vFiler unit always takes precedence over the virus scanner registered withthe host storage system. If a virus scanner is registered with a vFiler unit and that scanner is functional,the files on the vFiler unit will be scanned by this scanner. If this scanner becomes unavailable, thevscan options use_host_scanners command is set to On, a scanner is registered with the hoststorage system, the files on the vFiler unit will be scanned by the host storage system’s virus scanneruntil the scanner local to the vFiler unit becomes available.

Virus scanning for vfiler0The vfiler0 vFiler unit is the host storage system; therefore, files on vfiler0 can be scanned only byvirus scanners registered with the host storage system.

Virus protection for CIFS | 121

Requirements for virus scanning on a vFiler unit other thanvfiler0

There are requirements that must be met before you can scan files on a vFiler unit other than vfiler0.

• Virus scanning must be enabled for each vFiler unit.

• A virus scanner must be available for the vFiler unit because one of the following requirements ismet:

• A virus scanner must be registered with the vFiler unit.

• A virus scanner must be registered with the host storage system and the vFiler unit must beallowed to use it.

Effect of virus scanner availability on CIFS accessAlthough virus scanning is enabled and the mandatory_scan option for the vscan command is setto On, CIFS clients of the vFiler unit are not allowed to open any files on the vFiler unit if no virusscanner is available.

Configuring virus scanning for a vFiler unitYou can configure virus scanning either to use or not use the virus scanner registered with the hostsystem, if the virus scanner registered with the vFiler unit is unavailable.

Steps

1. Enter the following command to specify the virus scanner for scanning files on a vFiler unit:

vfiler run vfilertemplate vscan options use_host_scanners on | off

Set the use_host_scanners option to On to allow the vFiler unit to use the virus scanner registeredwith the host storage system. Doing so enables the host storage system and its vFiler units to sharethe virus scanner. However, the vFiler unit uses the virus scanner registered with the host storagesystem only when the virus scanner registered with the vFiler unit is unavailable.

Set the use_host_scanners option to Off if you do not want to allow the vFiler unit to use thevirus scanner registered with the host storage system.

Note: The use_host_scanners option is applicable only to a vFiler unit you created. Youcannot set it on vfiler0 or a storage system.


2. Enter the following command to enable or disable virus scanning:

vfiler run vfilertemplate vscan on | off

Virus protection for CIFS | 123

Index

/etc/dgateways file 58/etc/exports file

exporting all file systems in 79hanging path names in after renaming volume 51

A

active/active configuration 69adding and removing vFiler unit resources, 36addresses relative to IPspace 66aggr add command

adding new disks 88aggr create command 89aggr status command 118allowing or disallowing

CIFS or NFS, effects of 45audience

intended for this book 15

B

backing up a vFiler unit 52backup vFiler unit 87, 95

C

CIFSsupport for a vFiler unit 44local user accounts 80path names for shares 77prefdc command (domain controller) 98statistics 63virus scanning 122

cifs setup command 34cifs stat command 63clients, effect of vFiler unit move on 110commands

halt 119vfiler migrate 91, 92cf disable 119cf enable 119ipspace assign 91ipspace create 91

commands (continued)ipspace list 91license delete 119options snapmirror 103options snapmirror.access 104remove_ownership 119vfiler dr configure 91, 96vfiler migrate 91, 92ypwhich 92

commands for a vFiler unitvfiler command, purpose of 31entering through rsh 48, 49vfiler context command; 46vfiler dr activate command 98vfiler migrate start command 97, 111vfiler move command 37vfiler remove command 37vfiler setup command 34vfiler start command 44vfiler status command 46, 88

configuringvirus scanning 121, 122

consolidating multiple servers 21context of vFiler unit, switching to 46create a vFiler unit 28creating a vFiler unit

equired status of network interface 32n nondefault IPspaces 74

creating volumes 89

D

daemonenabling the routed 28

default IPspace, interfaces in 68destination storage system, preparing for migration 88destroying a vFiler unit 38destroying IPspaces 73disabling MultiStore license 29disabling SnapMover vFiler unit migration 118disallowing or allowing

CIFS or NFS, effects of 45disaster recovery

quotas 90

Index | 125

disaster recovery (continued)vFiler unit 95

disk upgrade-ownership command 117displaying

quota status 84, 90distinct IP address space 65DNS servers

migration disaster recovery 92domain controller name, changing 98domains

DNS servers:and vFiler unit 22multiple 22NIS servers

vFiler unit 22

E

ebooting storage system, effects on a vFiler unit 50

F

FCP LUNs 62FilerView

configuring resources with 33using to manage host storage system resources 25

FlexVol volumefor a vFiler unit 50

FTP support for a vFiler unit 44

H

host storage system 24, 25, 35, 47, 90access to data on a vFiler unit 24administering vFiler unit from 35administration tasks 25anaging a vFiler unit from 25quotas not copied to new vFiler unit 90

HTTP support for a vFiler unit 44

I

incoming packets for an IPspace 68IP addresses

and migration, disaster recovery 91assigned to a vFiler unit 32configuring for vFiler unit creation 32removing 36removing from an interface 72

IPsec, configuring vFiler unit for 58

IPspaces 38, 65, 66, 68, 69, 71, 72, 91and migration, disaster recovery 91creating 71destroying vFiler units in different 38guidelines for using 65incoming and outgoing packets in 68interfaces in 68listing 72names for in a active/active configuration 69routing tables for 68typical applications of 66

iSCSI and a vFiler unit 56iSCSI LUNs 55iSCSI support for a vFiler unit 44

L

language, guidelines for/etc/quotas file

etc/usermap.cfg ,language for encoding 30language for encoding 30

license for MultiStore, enabling or disabling 28limit 39local user accounts 80LUNs on a vFiler unit 55

M

managing storage systemvolumes, disks, and RAID groups

backups and data recovery 25mandatory_scan option 122maximum number of vFiler units 41migrate

by copying data 110storage system data 23ways to 109

moving a vFilerhow to 109

moving a vFiler unitquotas 90

moving resources, about 36multiple security domains 22multiple server consolidation 21

N

NDMP 53


network checks for migration to destination storage system90Network Data Management Protocol 53network interfaces

configuring down 32emoving an IP alias from 32for a vFiler unit 24

network resources 32, 36base IP address

IP alias 32requirements for moving and removing 36

NFSsupport for a vFiler unit 44path names for exporting 77starting the protocol 78statistics 63

nfsstat command 63NIS servers

and migration, disaster recovery 92no-vfiler-ips? variable, setting 105

O

outgoing packets for an IPspace 68

P

partitioning storage system resourcesstorage system resources

partitioning 22path names, for NFS exports and CIFS shares 77performance, monitoring 62primary unit 29protocols

allowing or disallowing 45supported for a vFiler unit 44

pseudo-root, definition of 77

Q

qtree command output (how it differs for a vFiler unit andhost storage system) 52qtrees

who can create on a vFiler unit 51quota report command 90quotas

effects of destroying a vFiler unit on 38guideline for 30prerequisite for turning on and off 81, 83

quotas (continued)resizing 84seeing where enforced from 90that are copied to new vFiler unit 90types supported for a vFiler unit 81vFiler unit names in report 85who specifies 82

R

removing IP addresses from an interface 72removing IP aliases from an interfaceIP alias, removing 32renaming a vFiler unit 42resizing quotas 84resources

assigning 33changing 36guidelines for assigning 29

responding to a disaster 98restricting storage system traffic 22routed daemon

effects of disabling 57enabling 28

routing tablefor the storage system 57

routing table:or a vFiler unit in default IPspace 58

rshaccess to vFiler unit from clients 48enable option 48support for a vFiler unit 44

RSH 48

S

SAN guidelines 31secure routing and IPspace 66Server Manager 79setting up a vFiler unit

ntroduction to 34setup command for a vFiler

unitpurpose of 34SnapMirror

options snapmirror.access command 104using to reactivate vFiler unit 103

snapmirror commandssnapmirror break command 105snapmirror quiesce 104snapmirror status 104snapmirror update command 104

Index | 127

snapmirror update command 107snapmirror.allow file 103SnapMover

disabling 118vFiler unit migration 87

SnapMover:enabling 116

software-based disk ownershipdisk upgrade-ownership command 117enabling 117

ssh 44SSH 49starting a vFiler unit

meaning of 44step for 44

storage checks for migration to destination storage system88, 94storage resources 29, 36

assigning volumes 29requirements for moving and removing 36

storage system partitioning 21storage system reboot, effects on a vFiler unit 50subnet, moving vFiler unit to a different 114

T

traditional volume for a vFiler unit 50trusted host

changing after migration 114

U

uptime command 63User Manager 79

V

vFiler unit 23, 24, 34, 36, 39, 42, 43, 44, 46, 87, 90, 94, 99, 102, 103, 105, 107, 109, 112, 114, 116, 118

increasing the limit of 39backup for disaster recovery 87changing resources for 36decreasing the limit of 42default 23disaster recovery

checking storage space 94preparing for 87, 90

disaster recovery:creating a 99

vFiler unit (continued)displaying status 46maximum number of 24, 39migration

copying data for 112disabling 118SnapMover, enabling 116using the vfiler migrate command for 116

migration formethod 109

movingto different subnet 114to different Windows domain 114

protocols supported 24reactivating using vfiler dr commands 105reactivating via SnapMirror commands 103rename 42replacing on original storage system 107resynchronization (resync)

handling failures 102setup 34SnapMover 87states, stopped or running 43

vFiler unit limit 41vFiler unit:

disaster recoveryactivating a 98

vFiler unit:limit per host storage systemmaximum number of vFiler units 24

vfiler0included in vFiler unit limit 39

vfilertemplate, defined 31virus scanning

registering scanners for;scanners, virus 121requirements for 122

VLAN tagging 68VLANs 69volumes in a vFiler unit

assigning 29effects of renaming 51taking offline 51

W

Windows domainchanging 99

WINS serverchanging 99changing after migration 114


WINS serversand migration, disaster recovery 93

Index | 129

vfiler

Documents

vfiler unit data

vfiler unit storage

vfiler unit resources

vfiler units

vfiler commands

default vfiler unit

multistore license

multistore management