Migrations Using Powerpath

7/22/2019 Migrations Using Powerpath

http://slidepdf.com/reader/full/migrations-using-powerpath 1/108

EMC Corporation

Corporate Headquarters :

Hopkinton, MA 01748-9103

1-508-435-1000

www.EMC.com

EMC ® PowerPath ® Migration EnablerVersion 5.5

USER GUIDEP/N 300-010-686

REV A02



EMC PowerPath Migration Enabler Version 5.5 User Guide 2

Copyright © 2006 - 2010 EMC Corporation. All rights reserved.

Published October 2010

THE INFORMATION IN THIS PUBLICATION IS PROVIDED “AS IS.” EMC CORPORATION MAKES NOREPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THISPUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY ORFITNESS FOR A PARTICULAR PURPOSE.

Use, copying, and distribution of any EMC software described in this publication requires an applicablesoftware license.

For the most up-to-date listing of EMC product names, see EMC Corporation Trademarks on EMC.com.

All other trademarks used herein are the property of their respective owners.




Preface ............................................................................................................................. 7

Chapter 1 PowerPath Migration Enabler Overview

Migration Enabler overview ........................................................... 12Supported migrations ................................................................12PowerPath Encryption with RSA.............................................12Pseudo and native devices ........................................................13

Technology types .............................................................................. 13Host Copy ....................................................................................13Encapsulation.............................................................................13

Open Replicator ..........................................................................14

TimeFinder/Clone......................................................................14Migration states................................................................................. 15Migration state descriptions......................................................16

Migration Enabler installation and licensing................................ 18Migration Enabler licensing......................................................18UNIX and Linux installations ...................................................18Windows installations................................................................19

Chapter 2 Migration Procedures

Migration process overview............................................................ 22Prerequisites ................................................................................22Setup .............................................................................................24Synchronization ..........................................................................24SourceSelected.............................................................................25TargetSelected .............................................................................26

Commit.........................................................................................26Cleanup ........................................................................................27

Contents




Contents

Post-migration steps...................................................................28

Reclaim unused space on the target for plaintext..................28Reclaim unused space on the target for encrypted text........30Migration procedure steps .............................................................. 31

Chapter 3 Migration Enabler Features

Migration Enabler commands ........................................................ 34Understanding throttle impact....................................................... 36

Understanding Host Copy ceiling copy rate................................ 39Powermig options command....................................................39Pause and resume commands ........................................................ 40

powermig pause ........................................................................ 40powermig resume...................................................................... 41

Virtually provisioned devices......................................................... 42Solutions Enabler thin client for Migration Enabler.................... 43

Changing the remote Solutions Enabler server .....................44

Viewing and updating disk labels with powerformat................ 45

Chapter 4 Volume Managers and Cluster Environments

Volume managers............................................................................. 48Native-named device migrations with VxVM .......................48LVM-controlled native devices on HP-UX and Linux hosts 49

Migrations in a cluster environment.............................................. 51

Migrations in an MSCS cluster .................................................51Migrations with HACMP..........................................................53Migrations in a VCS environment ...........................................55Migrations in a Sun Cluster ...................................................... 59

Appendix A Migration Procedure Examples

Host Copy migrations...................................................................... 64

Encrypted devices ......................................................................67Encapsulation migrations............................................................... 68Open Replicator migrations ............................................................ 72TimeFinder/Clone migrations ....................................................... 77

Appendix B Troubleshooting a Migration

Audit and error log messages......................................................... 82Reporting a problem ........................................................................ 83Graceful host reboot ......................................................................... 84Host crash .......................................................................................... 85



5EMC PowerPath Migration Enabler Version 5.5 User Guide

Contents

Other reported faults ........................................................................ 87

Resolving a process hang................................................................. 88

Appendix C Support Summary

Migration Enabler support summary ............................................ 90

Appendix D Disk Labeling powerformat Command for Solaris

Solaris and PowerPath Migration Enabler disk labeling ............ 94

Viewing and updating disk labels with powerformat ..........96

Index .............................................................................................................................. 105




Contents




Preface

As part of an effort to improve and enhance the performance and capabilitiesof its product line, EMC from time to time releases revisions of its hardwareand software. Therefore, some functions described in this guide may not besupported by all revisions of the software or hardware currently in use. Forthe most up-to-date information on product features, refer to your productrelease notes.

If a product does not function properly or does not function as described inthis guide, please contact your EMC representative.

Audience and

prerequisites

This manual is part of the PowerPath documentation set. It is

intended for use by storage administrators and other informationsystem professionals responsible for using PowerPath MigrationEnabler.

Readers of this guide are expected to be familiar with the operatingsystem on which you are running PowerPath Migration Enabler andwith EMC storage systems.

Related

documentation

Other related EMC PowerPath publications include:

◆ EMC PowerPath Installation and Administration Guide

◆ EMC PowerPath Encryption with RSA User Guide

◆ EMC PowerPath and PowerPath /VE for Windows Installation and Administration Guide

◆ EMC PowerPath Family CLI and System Messages Reference

◆ EMC PowerPath Family Product Guide

◆ EMC PowerPath Management Pack for Windows User Guide



8 EMC PowerPath Migration Enabler Version 5.5 User Guide

Preface

These documents are updated periodically. Electronic versions of theupdated manuals are available on the EMC Powerlink website.

Revision history Beginning with PowerPath 5.5, this document includes a revisionhistory.

The following table presents the revision history of this document.

Conventions used inthis document

EMC uses the following conventions for special notices.

Note: A note presents information that is important, but not hazard-related.

CAUTION A caution contains information essential to avoid data loss ordamage to the system or equipment. The caution may apply tohardware or software.

Typographical conventionsEMC uses the following type style conventions in this document:

Revision Date Description

A01 August 27, 2010 Firstreleaseof the EMC PowerPathMigration EnablerUser Guide Version 5.5.

A02 October, 2010 Second release of the EMC PowerPath MigrationEnabler User Guide Version 5.5.

Normal Used in running (nonprocedural) text for:

• Names of interface elements (such as names of windows,dialog boxes, buttons, fields, and menus)

• Names of resources, attributes, pools, Boolean expressions,buttons, DQL statements, keywords, clauses, environmentvariables, filenames, functions, utilities

• URLs, pathnames, filenames, directory names, computernames, links, groups, service keys, file systems, notifications

Bold: Used in running (nonprocedural) text for:

• Names of commands, daemons, options, programs,processes, services, applications, utilities, kernels,notifications, system call, man pages

Used in procedures for:

• Names of interface elements (such as names of windows,dialog boxes, buttons, fields, and menus)

• What user specifically selects, clicks, presses, or types




Preface

Where to get help EMC support, product, and licensing information can be obtained asfollows.

Product information — For documentation, release notes, softwareupdates, or for information about EMC products, licensing, andservice, go to the EMC Powerlink website (registration required) at:

http://Powerlink.EMC.com

Technical support — For technical support, go to EMC CustomerService on Powerlink. To open a service request through Powerlink,

you must have a valid support agreement. Please contact your EMCsales representative for details about obtaining a valid supportagreement or to answer any questions about your account.

Italic: Used in all text (including procedures) for:

• Full titles of publications referenced in text

• Emphasis (for example a new term)

• Variables

Courier: Used for:

• System output, such as an error message or script

• URLs, complete paths, filenames, prompts, and syntax whenshown outside of running text

Courier bold: Used for:

• Specific user input (such as commands)Courier italic: Used in procedures for:

• Variables on command line

• User input variables

< > Angle brackets enclose parameter or variable values supplied bythe user

[ ] Square brackets enclose optional values

| Vertical bar indicates alternate selections - the bar means “or”

{ } Braces indicate content that you must specify (that is, x or y or z)

... Ellipses indicate nonessential information omitted from theexample

http://powerlink.emc.com/





Preface

Your comments Your suggestions will help us continue to improve the accuracy,organization, and overall quality of the user publications. Please sendyour opinion of this guide to:

[email protected]

If you have issues, comments, or questions about specific informationor procedures, please include the title and, if available, the partnumber, the revision (for example, A01), the page numbers, and anyother details that will help us locate the subject you are addressing.



PowerPath Migration Enabler Overview 11

1Invisible Body Tag

This chapter provides an overview of the PowerPath MigrationEnabler product, discusses the various technology types andmigration states, and provides licensing information for migration

enabler.

◆ Migration Enabler overview............................................................. 12◆ Technology types ............................................................................... 13◆ Migration states.................................................................................. 15◆ Migration Enabler installation and licensing................................. 18

PowerPath MigrationEnabler Overview




PowerPath Migration Enabler Overview

Migration Enabler overview

EMC® PowerPath® Migration Enabler is a migration tool that enablesnon-disruptive or minimally disruptive data migration betweenstorage systems or between logical units within a single storagesystem. Migration Enabler resides on the host and allowsapplications continued data access throughout the migration process.Migration Enabler works independently of PowerPath multipathing.However, PowerPath must be fully installed.

PowerPath Migration Enabler integrates with other technologies tominimize or eliminate application downtime while migrating data.“Technology types” on page 13 contains additional information. Forprocedural information about the various technology types, refer toChapter 2, ”Migration Procedures.”

Supported migrationsSupported migration scenarios are included in the EMC PowerPathFamily Release Notes for each operating system on the EMCPowerlink® website. The release notes include information aboutsupported pseudo and native device names, technology types, datatypes (encrypted and plaintext), and operating systems.

PowerPath Encryption with RSA

The Migration Enabler Host Copy technology can be used to migratedata onto an encrypted LUN. The PowerPath Encryption with RSAUser Guide provides additional information.



Technology types 13


Pseudo and native devices

PowerPath Migration Enabler supports migrations using PowerPathdevices (pseudo devices) or native operating system devices.Migrations are non-disruptive with pseudo devices, which means anapplication never has to be brought down through the entiremigration process. With native devices, you must manually configurethe application after the migration to allow the host to see the newtarget device. Refer to the EMC PowerPath Product Guide on Powerlink

for more information.

Note: Pseudo devices are supported on all platforms except HP-UX.

Technology types

This section describes the Migration Enabler technology types.

Host Copy

PowerPath Migration Enabler Host Copy is a host-based migrationtool that uses host resources to move data nondisruptively from asource to a target. Host Copy supports the following:

◆

Migrating to and from a device managed by PowerPath (bootdevices are not supported)

◆ Migrations to and from virtually provisioned devices (thindevices)

◆ Migrations to and from encrypted LUNs

EncapsulationPowerPath Migration Enabler Encapsulation supports migration of EMC Symmetrix® or EMC CLARiiON® devices to EMC ® Invistadevices. The target logical unit in the migration is a Virtual Volumethat encapsulates a Symmetrix or CLARiiON Storage Element. Since

both logical units represent the same physical storage, no data isactually moved during the migration.





Open Replicator

EMC Open Replicator for EMC Symmetrix works in conjunction withPowerPath Migration Enabler and with Solutions Enabler andsupports the online pull method where data is copied through theSAN from a remote logical unit to a Symmetrix storage system.

Note: Source from CLARiiON or Symmetrix cannot be in the same array.

TimeFinder/Clone

EMC TimeFinder®/Clone migrations work in conjunction withPowerPath Migration Enabler and with Solutions Enabler to migratedata between devices within a single Symmetrix system, includingdevices with different RAID types.

Note: The source and target must be from the same Symmetrix array.



Migration states 15


Migration states

A migration session transitions through a number of states as youcomplete each step in the migration process.

Figure 1 on page 15 shows the overall migration workflow, includingthe sequence of powermig commands and migration states. Numbersone through eight in the graphic show the commands involved in anormal workflow.

Figure 1 Migration commands and states

CommittedAndRedirected

Setup

Synching

SourceSelected

TargetSelected

Committed

powermig cleanup

powermig commit

Source is Pseudo

powermig commit

Source is Native

powermig setup

powermig sync

powermig abort

powermig

selectTargetpowermig

selectSource

powermig cleanuppowermig undoRedirect

source & targetsynchronize

1

2

3

4

5

6 6

7

8

PP-000011

http://-/?-

http://-/?-





Migration state descriptions

Table 1 on page 16 describes the migration states. The PowerPathFamily CLI and System Messages Reference Guide provides additionalinformation.

Table 1 Migration state descriptions

state Description

Setup A migration is in the Setup state after a powermig setupcommand completes successfully. Before the migration canenter this state, synchronization prerequisites must be met,including technology-specific criteria, where relevant.

Syncing A migration enters the Syncing state after running thepowermig sync command.

Bulk data copy is initiated in the Syncing state. Data is read fromthe source and is written to the target. Application reads aredirected to the source. Application writes are written to thesource and cloned to the target.

SourceSelected The migration enters the SourceSelected state once thebulk data copy of the source and target logical units completes.

While the migration is in the SourceSelected state, thesource and target logical units continue to be synchronized, andI/O read requests are processed from the source logical unit.

TargetSelected The migration transitions to the TargetSelected state

after the powermig selectTarget command completes.While the migration is in the TargetSelected state, the source

and target logical units continue to be synchronized, and

I/O read requests are processed from the target logical unit.

CommittedAndRedirected If you are migrating data from a native-named device, this stateoccurs after you issue the powermig commit command. Whilein this state, the application using the native source device mustbe brought down and the user must run the powermig

undoRedirect command. The application can then bereconfigured and brought back up to use the target device. Themigration will now be in the Committed state.



Migration states 17


Committed If using a pseudo source device, this state is entered uponissuing the powermig commit command. During this transition,Migration Enabler swaps all underlying paths on both sourceand target logical units causing all I/O to be sent to the targetlogical unit without an application having to be brought down orreconfigured.

If using a native source device, this state is entered after issuing

the powermig undoRedirect command (as described above inthe CommittedAndRedirected state).

In the Committed state, source and target devices are nolonger kept synchronized and I/O only goes to the target device.

Table 1 Migration state descriptions (continued)

state Description





Migration Enabler installation and licensing

For UNIX and Linux operating systems, PowerPath MigrationEnabler automatically installs when PowerPath multipathing isinstalled. For Windows Server 2003 and 2008 operating systems, youmust choose Custom or Complete during installation, or chooseModify for an existing installation.

Migration Enabler licensingPowerPath Migration Enabler license keys are available at no cost tonew and existing PowerPath Multipathing customers. A PowerPathMigration Enabler license is technology-specific, which means thatyou can perform migrations with a specific technology type,including Host Copy, Open Replicator, TimeFinder/Clone, andEncapsulation.

Note: The PowerPath Encryption with RSA license supports data migrationto or from encrypted devices without the need to install the PowerPathMigration Enabler Host Copy license. The Host Copy license is requiredwhen migrating with only plain text devices (no encrypted devices as sourceor target).

You do not need to obtain a separate license for every server. A single

license per technology type supports multiple PowerPath migrations.

Note: Multiple PowerPath migrations are supported using differenttechnologies on the same host.

UNIX and Linux installations

If you have not done so already, install PowerPath. Follow theinstructions in the platform-specific PowerPath installation guide,available on the Powerlink website:

◆ PowerPath for Solaris Installation and Administration Guide

◆ PowerPath for AIX Installation and Administration Guide

◆ PowerPath for Linux Installation and Administration Guide

◆

PowerPath for HP-UX Installation and Administration Guide



Migration Enabler installation and licensing 19


1. On AIX or Linux hosts, type:

emcpreg -installOn Solaris hosts, enter:

/etc/emcpreg -install

2. Type y and press Enter.

3. Enter the PowerPath Migration Enabler registration key and

press Enter.

4. Press Enter.

Windows installations

To install PowerPath Migration Enabler on Windows hosts:

If you have not done so already, install the appropriate version of PowerPath for your host operating system. Follow the instructions inthe latest version of the PowerPath for Windows Installation and

Administration Guide, available on Powerlink.

1. Reboot the host for modifications to take effect.

2. Register the PowerPath Migration Enabler software. Go to:

Start > Programs > EMC > PowerPath Licensing Tool.3. In the EMC PowerPath Licensing Tool dialog box, enter the

PowerPath Migration Enabler key in the License Key field, andclick Add.







Migration Procedures 21

2Invisible Body Tag

This chapter provides the procedural steps for performing thevarious PowerPath technology type migrations, and includes themigration prerequisites and post-migration steps:

◆ Migration process overview ............................................................. 22◆ Migration procedure steps ................................................................ 31

Migration Procedures





Migration process overview

The following sections provide migration prerequisites, an overviewof the PowerPath Migration Enabler process, and post-migrationsteps.

Prerequisites

The following prerequisites apply to all the Migration Enablertechnology types:

Note: Some prerequisites may only apply to specific technology types, asnoted.

◆ You must have the appropriate migration license installed inPowerPath.

◆ Make target devices visible to the host, such as by provisioningSymmetrix LUs or adding volumes to the storage group.

◆ The target device must be the same size or larger than the sourcedevice.

◆ For PowerPath Encryption with RSA, the target logical unit must be at least 65 KB (AIX, Solaris, or Windows) and 68 KB (Linux)larger than the source when migrating from an unencrypted to an

encrypted virtual logical unit to allow space for PowerPathEncryption metadata. Refer to the PowerPath Encryption with RSAUser Guide for more information.

Note: Applies only to Host Copy with encryption.

◆ The target device cannot be in use by any application.

◆

If using the Solutions Enabler remote-only client with OpenReplicator or Timefinder/Clone technologies, download andinstall the Solutions Enabler Thin Client from Powerlink.

◆ For migrations with Volume Managers and cluster environments,refer to “Volume Managers and Cluster Environments” onpage 47.

◆ With replication, such as MirrorView or Symmetrix Remote Data

Facility (SRDF®), if your source is replicated to a remote site, youmust also replicate your target to a remote site.



Migration process overview 23


◆ The following applies to Windows:

• Ideally, the target logical unit is a freshly initialized MBRpartition on a basic disk (or a GPT basic disk if the platformsupports it). Regardless of the source type, the target logicalunit should not contain a volume or file system.

• Make sure that the Startup type is set to enabled to runPowerPath Migration Enabler. Services must be turned backon if you turned them off.

◆ The following applies to Solaris:

• Prior to using the powermig setup command, both the sourceand target must be labeled, and the labels must be of the sametype (VTOC/SMI or EFI).

◆ The following applies to Open Replicator:

• The target logical unit must reside on a Symmetrix systemwith Enginuity Operating Environment 5x71 or later.

• The two storage systems where the source and target residemust be connected directly or through a switch and they must

be able to communicate.

• Every port on the target array that allows access to the targetlogical unit must be accessible to the source logical unitthrough at least one port on the source array.

The following also applies to TimeFinder/Clone:

• Logical units involved in a migration should not be used asSymmetrix gatekeeper devices. Add the names of devices youwant to use as gatekeepers to the gatekeeper selection file,gkselect. Even if the device names are already configured asgatekeepers on the Symmetrix system, add them to thegkselect file to avoid a potential conflict between SYMAPI andMigration Enabler. The EMC Solutions Enabler Symmetrix Base

Management CLI Product Guide contains more informationabout gatekeeper devices.





Setup

After the prerequisites are met, run the powermig setup command toenter the Setup state. You perform the following actions:

◆ Specify the source and target device names.

Note: EMC PowerPath recommends that you do not rename a logicalunit that is involved in a migration session.

◆ Specify the technology type.

◆ Optionally set the throttle values to configure the migrationspeed of the synchronization. Refer to “Understanding throttleimpact” on page 36 for additional information.

◆ Optionally specify the remote host and/or the port for SolutionsEnabler (refers to Open Replicator and TimeFinder/Clonemigrations only). See “Solutions Enabler thin client for Migration

Enabler” on page 43 for additional information.

To successfully enter the Setup state, PowerPath Migration Enablerperforms the following actions:

◆ Checks devices for accessibility

◆ Checks that the devices are not part of another migration

◆ Checks the size of the devices

◆ Checks if they are virtually provisioned devices (Host Copy only)

Note: During Setup, the target is made unavailable to all applications on thehost for the duration of the migration.

Synchronization

After the Setup state, run the powermig sync command to enter theSyncing state. You must specify the handle (returned from Setup) forthe migration pair. In the Syncing state, host applications continue toaccess the source logical unit, as data is copied to the target.Migration Enabler also ensures all new host writes are written to bothsource and target. The time for completing synchronization dependsin part on the size of the source logical unit.

While in the Syncing state, you can perform the following options:





◆ Query one or more migrations to determine percentagecompletion and throttle value.

◆ Speed up or slow down the migration synchronization rate bychanging the throttle value. Refer to “Understanding throttleimpact” on page 36 for additional information.

◆ Optionally control the speed of migration synchronization bysetting the Host Copy ceiling (in MB/s). Refer to “UnderstandingHost Copy ceiling copy rate” on page 39 for additionalinformation.

◆ Optionally pause the migration and then resume at a later time.

◆ Abort the migration, returning it to the Setup state.

Once in the Syncing state, PowerPath Migration Enabler performs thefollowing actions:

◆ Initiates write cloning of application data. Any new I/O writes tothe source device are automatically cloned and written to the

target device.◆ Initiates bulk data copy from source device to target device.

SourceSelected

After all data has been copied from source to target, the migrationenters the SourceSelected state where the following occurs:

◆ I/O reads are directed to the source.

◆ I/O writes are written to the source and target to keep themsynchronized.

From the SourceSelected state, you can:

◆ Transition to the TargetSelected state.

◆ Optionally abort the migration and return the migration to the

Setup state.

Mi ti P d





TargetSelected

When you execute the powermig selectTarget command, themigration transitions to the TargetSelected state from theSourceSelected state. The TargetSelected state must be entered

before committing the migration. In this state, all I/O reads are sentto the target while writes continue to be written to both source andtarget. This gives you an opportunity to determine that all data isintact on the target and service levels are in line with application

requirements.After entering the TargetSelected state, perform the followingactions:

◆ Move to the Committed state, (or CommittedAndRedirectedstate if migrating a native source device).

◆ Optionally abort the migration and return the migration to theSetup state.

◆ Optionally return to the SourceSelected state.

Once in the TargetSelected state, PowerPath Migration Enabler behaves as follows:

◆ I/O reads are directed to the target to keep them synchronized.

◆ I/O writes are written to the source and target to keep themsynchronized.

Commit

The Commit state permanently designates the target as the recipientof all I/O requests. After this command is run, PowerPath MigrationEnabler no longer keeps the source and target synchronized, and I/Ois no longer sent to the source.

After entering the Commit state, perform the following action:

◆ Issue the Cleanup command.

To successfully enter the Commit state, PowerPath Migration Enablerperforms the following actions:

◆ When committing pseudo devices, source and target LUNs areswapped, allowing an application to continue using the originalsource pseudo device without any disruption.






◆ When committing native devices, the migration enters theCommittedAndRedirected state. In this state you must take

down any application using the source device and execute theundoRedirect command which stops the redirection of all I/Ofrom source to target. At this point the application can bereconfigured to use the new target device and brought backonline.

CAUTION Once in the Commit state, there is no returning to the previousstate.

◆ All reads and writes are directed to the target.

◆ The source and target are no longer synchronized.

Cleanup

Cleanup is the final step in the migration process, where thefollowing occurs:

◆ When entered from the Commit state, the powermig cleanupcommand removes enough data from the former source device toprevent the existence of two identical logical units, ensuring thatthe operating system or applications do not become confused.

Note: For Encapsulation migrations, remove source access to the host before cleanup.

Note: On some operating systems, powermig cleanup with the -formatoption can take substantially longer since it performs a full format of thedisk. This is not valid for Linux.

◆ When entered from the Setup state (after setup or abort) thiscommand removes enough data from the target device to preventthe possibility of two identical logical units.

◆ The migration handle is removed and the migration no longerexists.






Post-migration steps

Complete the post migration steps that apply to your environment:

◆ If the target is larger than the source, follow the steps in “Reclaimunused space on the target for plaintext” on page 28 or “Reclaimunused space on the target for encrypted text” on page 30 tomake unused space accessible.

◆ If you have EMC Ionix ControlCenter, update it to recognize thenew device name mappings after a migration:

1. In the navigation tree on the main window, select the host thatis running the migration.

2. Right-click the host, and select Rediscover.

◆ Update any replication software to ensure that the new targetdevice is replicated. This could also be done prior to the syncingstate.

Reclaim unused space on the target for plaintext

After a migration completes, the usable space on the target isequivalent to the size of the source. When the target logical unit islarger than the source, operating system commands must be run tomake the unused space accessible.

Note: This section provides sample procedures, but it does not include stepsfor all supported operating systems and versions.

Solaris hosts

For Solaris hosts, when both source and target devices are SMIlabeled, to reclaim unused space on the original target device aftermigration and cleanup, perform the following steps:

1. Umount the file system if already mounted:

Type: unmount < mount-point>

2. Use PowerPath powerformat utility to reclaim unused space onthe target device:

Type: powerformat -w -x <target_raw_device >

3. Verify using the Solaris format utility that the new label is applied

(use verify option).








4. Quit format.

5. Grow the file system:Type: growfs [-M mount-point] [newfs-options]raw-special-device

For Solaris hosts, where both the source and target are EFI labeled, toreclaim unused space on the original target device after migrationand cleanup, perform the following steps:

1. Unmount the file system if already mounted:

Type: umount < mount-point>

2. Use format utility to relabel the target device:

Type: format -e <target device >

3. Select menu option type.

4. Choose 0: Auto Configuration.

5. Select menu option label and choose 1 for EFI label.6. Quit format.

7. Grow the file system:

Type: growfs [-M mount-point] [newfs-options]raw-special-device

Linux hosts

To reclaim unused space on the target for Linux hosts, perform thefollowing steps:

1. Complete the migration procedure described earlier in thischapter.

2. Unmount the file system.

3. Make the volume group unavailable. Type:

vgchange -an <vol_grp >

where:<vol_grp > = volume group name

4. Resize the physical volumes. Type:

pvresize /dev/emcpowerx

where:x = the name of the target logical unit after the migration





g

5. Make the volume group resizeable if it is not already. Type:

vgchange --resizeable y

6. Verify how many physical extents are free. Type:

vgdisplay <vol_grp>

7. Make the volume group available. Type:

8. vgchange -ay <vol_grp >

9. Extend the logical volume. Type:lvextend -l < phys_ext> /dev/<vol_grp >/<log_vol#>

where:< phys_ext> — the number of physical extents you want to add<vol_grp> — volume group name<log_vol#> — logical volume number

Example:

lvextend -l +6343 /dev/vg01/lv01

10. Resize file system. For example, for ext3 file systems, type:

e2fsck -f /dev/<vol_grp >/<log_vol#>resize2fs -p /dev/<vol_grp >/<log_vol#>

11. Mount the file system.

Windows hostsTo claim space on the target:

1. Run diskpart

2. Type list volume

3. Type select volume 1

4. Type extend

Reclaim unused space on the target for encrypted text

To reclaim unused space on the target for encrypted text, follow thesteps below:

Solaris and Windows

1. Run powervt update2. Run powervt resize




Migration procedure steps 31

Solaris

For Solaris, after following steps 1 and 2 above, follow the steps

below:

3. Grow the file system:

a. Type: growfs [-M mount-point] [newfs-options]

raw-special-device

b. Verify using the Solaris format utility that the new label isapplied (use verify option).

Windows

For Windows, after following steps 1 and 2 above, follow the step below:

• If there is volume build on top of this partition and file system,then follow the steps under “Windows hosts” on page 30 toutilize new capacity available on the virtual encrypted disk.

Linux

Follow the procedure in “Linux hosts” on page 29 to reclaim unusedspace on the target for Linux hosts.

Migration procedure steps

Appendix A, “Migration Procedure Examples,” provides specificprocedures for each of the following technology types:

◆ “Host Copy migrations” on page 64

◆ “Encapsulation migrations” on page 68

◆ “Open Replicator migrations” on page 72

◆ “TimeFinder/Clone migrations” on page 77





3



Migration Enabler Features 33

3Invisible Body Tag

This chapter provides information about the following topics:

◆ Migration Enabler commands.......................................................... 34◆ Understanding throttle impact ........................................................ 36◆ Understanding Host Copy ceiling copy rate ................................. 39◆ Pause and resume commands.......................................................... 40◆ Virtually provisioned devices .......................................................... 42◆ Solutions Enabler thin client for Migration Enabler ..................... 43◆ Viewing and updating disk labels with powerformat ................. 45

Migration Enabler

Features

Migration Enabler Features




Migration Enabler commands

PowerPath Migration Enabler uses a common set of commands tosupport Host Copy, Encapsulation, Open Replicator, andTimeFinder/Clone migration technology types. Table 2 on page 34lists the migration powermig commands. The PowerPath Family CLI and System Messages Reference Guide provides additional information.

Note: Some commands may vary in usage and meaning depending on

technology type. Refer to the Family CLI and System Messages Reference Guidefor an explanation of how these commands are used with the varioustechnology types.

Table 2 PowerPath Migration Enabler commands

powermig command Description

setup Specifies the source and target devices and

technology type used in the migration.

sync Starts the bulk data copy and write cloningprocess while application continues uninterrupted.

info Displays information about one or all migrationsessions.

query Displays information about one migration session.

pause Suspends a migration in the Syncing state.

resume Resumes a paused migration.

throttle Changes the bulk data copy rate for a migration.

options Used to set values that are global to allmigrations, or all migrations that use a specifictechnology. For example, the option,

hostcopy_ceiling, can be used to specify an upperlimit on the aggregate rate of copying for all hostcopy migrations. When executed with no optionspecified, powermig options displays all availableoptions and their values. Currently,hostcopy_ceiling is the only available option.

abort Aborts a migration and returns the migrationsession to Setup state.


http://-/?-

http://-/?-



Migration Enabler commands 35

selectSource Designates the source logical unit as the recipientof all I/O Read requests. Continues to send writesto both source and target.

selectTarget Designates the target logical unit as the recipientof all I/O-Read requests. Continues to send writesto both source and target.

recover Recovers a migration that was interrupted by amigration error or process crash.

getHandle Retrieves the handle for a migration session.

commit Commits the migration by permanentlydesignating the target as the recipient of all I/Orequests. Source device is no longer synced withthe target.

undoRedirct Removes the migration. Stops the redirection ofI/O from the source to the target. Use thiscommand only when the source logical unit is anative device and when application I/O is stopped.

cleanup Removes selected data on the source or targetlogical unit so that the abandoned device is left ina safe state.

version Displays the version of PowerPath MigrationEnabler running on the host.

help Displays a brief description and usage summaryfor the specified powermig command.

Table 2 PowerPath Migration Enabler commands (continued)

powermig command Description





Understanding throttle impact

The powermig throttle command changes the synchronization speedfor a migration, and can be set to increase or decrease the rate atwhich bulk data is copying. As throttle value increases, the bulk datacopy rate decreases. The command works differently for thesupported technology types, as described in this section. The EMCPowerPath Migration Enabler Host Copy White Paper available onEMC Powerlink provides additional details about throttle impact.

This command does not apply to Encapsulation migrations.“Understanding Host Copy ceiling copy rate” on page 39 discussesan option for setting the ceiling rate, which disables powermigthrottle.

Note: Throttle can also be set during powermig setup, and can be changedwith powermig throttle while the migration is in the setup or syncing state,except in the case of OR, where throttle can be set when you run powermig

setup, but cannot be changed until after running powermig sync.

Syntax powermig throttle-handle < migrationHandle >-suspendTime <suspendValue >-throttleValue <throttleValue >[-noPrompt]

Note: For Solaris, HP-UX, Windows, and Linux, the -suspendTimecommand line argument is no longer valid for the Host Copy technologytype.

Arguments -handle <migrationHandle>Number that identifies the migration session. It is assigned whenyou issue the powermig setup command.

-throttleValue <throttleValue>Value of the throttle. Valid entries are 0–9, where 0 is the fastest.The default values are: 5 for Open Replicator; 2 for Host Copy;and 0 for TimeFinder/Clone. Use powermig pause and resume tostart and stop the migration when necessary.

-noPromptSpecifies you want to skip the confirmation prompt beforeexecuting the command.




Understanding throttle impact 37

Note: Depending on physical server characteristics, I/O workload, and thenumber of migrations, the throttle option can potentially impact hostapplications.

Migration state Required migration state to run the command: Syncing, Setup

Migration state after the command completes: This command doesnot change the migration state.

Example powermig throttle -handle 1 -throttleValue 0

No output appears when this command completes successfully.

Technology-specific notes

This section describes the throttle command and arguments in moredetail for each technology type.

Host Copy

The throttleValue defines the overall percentage of time the hostspends copying data.

Table 3 on page 37 shows the percentage of the host’s time spentcopying data as defined by the throttle value.

Note: For Solaris, HP-UX, Windows, and Linux, the -suspendTime commandline argument is no longer valid for the Host Copy technology type.

Open ReplicatorFor Open Replicator migrations, the throttle value adjusts the speedat which data is transferred between the source and target logicalunits. For Open Replicator, the throttle works the same as theSolutions Enabler symrcopy command with the set pace argument.The Solutions Enabler Symmetrix Open Replicator CLI Product Guide onEMC Powerlink provides additional information. A lower throttlevalue can significantly increase resource consumption on the target

logical unit’s array.

Table 3 Host Copy throttle values

Throttle Value 0 1 2 3 4 5 6 7 8 9

Percentage of time the hostspends copying data

100% 60% 36% 22% 13% 7.8% 4.7% 2.8% 1.7% 1.0%





TimeFinder/CloneFor TimeFinder/Clone migrations, the throttle value changes the

Symmetrix Quality of Service (QoS) value for the source and targetlogical units. The QoS value affects all I/O performance for thedevices, and not just migration performance. The powermig setupcommand sets the QoS value to 0, unless you specify a differentthrottle value during Setup. After a migration completes, the QoSvalue set during the migration remains on the source and target.Therefore, the last value set during the migration remains after themigration.

Symmetrix QoS values range from 0 to 16 and Migration Enablerthrottle values range from 0 to 9. The EMC Solutions EnablerSymmetrix Array Controls CLI Guide available on EMC Powerlinkprovides additional information. Migration Enabler maps the throttlevalues to QoS values as shown in Table 4 on page 38.

Table 4 TimeFinder/Clone throttle and QoS values

Throttle Value 0 1 2 3 4 5 6 7 8 9

Corresponding QoS value 0 2 4 5 7 9 10 12 14 16




Understanding Host Copy ceiling copy rate 39

Understanding Host Copy ceiling copy rate

The hostcopy_ceiling option lets you specify an upper limit on theaggregate rate of copying for all host copy migrations in the syncingstate on a host. In turn, this limits the I/O for all host copymigrations. With hostcopy_ceiling enabled, per-migration throttlevalues are disabled.

Powermig options commandThe powermig options command lets you set PowerPath MigrationEnabler options which apply to all migrations, or all migrationswithin a specific technology type. Running powermig options withno option merely displays all available options and their values.

Available options are: hostcopy_ceiling.

Syntax powermig options [-<option> <

value >]

Arguments -hostcopy_ceiling <ceilingValue|none>Used to control the aggregate copy rate for all host copymigrations in the syncing state. Once set, all per migrationthrottle values are ignored. Ceiling value is an integer between 1and 1000000 (1,000,000) per second, and disables per migrationthrottling. A value of none disables host copy ceiling andreenables per migration throttling.

Note: Host Copy ceiling value is not set on a per migration basis. Onceset, the value is applicable for all the host copy migrations on the server.The behavior is observed during syncing state. Therefore it is preferableto set the ceiling value before starting the host copy migrations.





Pause and resume commands

The powermig pause command lets you temporarily pause amigration, and the powermig resume command lets you resume themigration where it left off.

powermig pause The powermig pause command temporarily suspends the bulk datacopying between source and target logical units. The query commanddisplays the new status. When you resume the migration, the

synchronization process picks up where it left off.

Syntax powermig pause -handle <migrationHandle>[-noPrompt]

Migration state Required migration state to run the command: Syncing

Migration state after the command completes: Paused

Example powermig pause -handle 1


Note: This command is not supported with encapsulation.When the target is larger than the source, TimeFinder/Clone does notsupport this command.




Pause and resume commands 41

powermig resume The powermig resume command resumes a paused migration. When

restarted, the migration session resumes the synchronization of thesource and target. The synchronization resumes where it left off whenpowermig pause was run.

Syntax powermig resume -handle <migrationHandle>[-noPrompt]

Migration state Required migration state to run the command: Paused

Migration state after the command completes: Syncing

Example powermig resume -handle 1


Note: This command is not supported with encapsulation.When the target is larger than the source, TimeFinder/Clone does notsupport this command.





Virtually provisioned devices

Migrations to and from virtually provisioned devices (thin devices)offer improved storage capacity utilization and simplified storagemanagement. Virtual provisioning presents the application withsufficient capacity for an extended period of time, reducing the needto provision new storage frequently, and avoiding costly allocated

but unused storage.

In Host Copy migrations, the Setup state identifies the target devices

as either thick or thin. When copy operations begin with thin targetdevices, PowerPath Migration Enabler copies only those blocks withdata.

Host Copy operates in thin copying mode only if it recognizes thetarget device as a virtually provisioned device. It checks anarray-specific SCSI-mode page value to determine Symmetrix orCLARiiON virtual LUNs. Host Copy cannot determine whether or

not a LUN is thin on a third-party array.




Solutions Enabler thin client for Migration Enabler 43

Solutions Enabler thin client for Migration Enabler

Migration Enabler supports a Solutions Enabler (SE) thin client that isinstalled on the same host as Migration Enabler. The SE thin clientcommunicates with a remote server running a full version of SE. Theremote server can be a dedicated SYMAPI server, as shown inFigure 2 on page 44. The SE thin client allows you to limit the numberof servers running a full version of SE.

The SE thin client uses a reduced API and communicates with a

SYMAPI daemon on the remote server. The SE thin client host isunable to directly access a base daemon, Symmetrix devices, orsymmwin. Additionally, the thin client is restricted to the MigrationEnabler API, as SYMCLI is unavailable:

Note: The Solutions Enabler (SE) thin client for PowerPath Migration Enablerapplies only to technology types that require Solutions Enabler (that is, OpenReplicator and TimeFinder/Clone).

◆ For remote-only installations using the SE thin client, downloadthe latest SE Thin Client version from Powerlink.


http://-/?-

http://-/?-

http://-/?-




Note: Both the host where the SE thin client is installed and the serverrunning the full SE version must be able to see the source and target devices.

Figure 2 Solutions Enabler thin client for Migration Enabler

Changing the remote Solutions Enabler server

In rare cases if the remote server (or port) changes after a migrationhas started, two environment variables are provided to allowupdating values that were set during powermig setup. The variablesare:

PPME_REMOTE_SE_SERVERPPME_REMOTE_SE_PORT

◆ These variables can only be used with post-setup powermigcommands (that is, they cannot be used with setup).

◆ An error is returned if these variables are present duringpowermig setup.

◆ Once used, values are saved and used for the duration of thatmigration (variables can and should be removed).

Symmetrix array

SYMAPI server withfull version of SE

PowerPath MigrationEnabler and SEThin Client host

SE management

I/O conection

IP connection




Viewing and updating disk labels with powerformat 45

Viewing and updating disk labels with powerformat

For Solaris, PowerPath Migration Enabler includes a utility calledpowerformat that allows you to safely update disk-label information,preserve partition definitions and data, and make newly availabledisk capacity available for use.

When you migrate data to a target logical unit that is larger than thesource logical unit, powerformat can adjust the disk label andpartition table to make available the entire disk capacity on the target

logical unit. The powerformat command takes partitions thatextended to the end of the currently labeled logical unit (the labelmigrated from the source logical unit), and extends them to the endof the target logical unit. This allows you to use the additional spaceon the target. Appendix D, “Disk Labeling powerformat Commandfor Solaris” provides detailed information.





4Invisible Body Tag



Volume Managers and Cluster Environments 47

This chapter discusses support with volume managers andmigrations in cluster environments:

◆ Volume managers .............................................................................. 48

◆ Migrations in a cluster environment............................................... 51

Volume Managers

and ClusterEnvironments

Volume Managers and Cluster Environments




Volume managers

Currently, Veritas Volume Manager (VxVM) and Logical VolumeManager (LVM) are supported with PowerPath Migration Enabler:

Note: The target logical unit in a migration cannot be controlled by a volumemanager. For example, when migrating Logical Disk Manager (LDM) orVxVM logical units on a Windows host, the target logical unit should not be amember of an LDM dynamic disk group or a VxVM disk group.

Veritas VolumeManager

◆ When setting up a migration, with the target device online, if VxVM holds open the device, then PowerPath Migration Enablerreports a target device in use error message. You can resolve thisissue in one of the following ways:

• Remove the device from VxVM control

• Run the -force option

• Run fuser to discover which process is holding open thedevice (UNIX only)

Note: For Linux, the target logical unit should be removed from VxVMconfiguration. Run vxdisk rm <target device> to remove from VxVMconfiguration.

On Linux hosts:

◆ For native-named device migrations, refer to “Native-nameddevice migrations with VxVM” on page 48.

Logical VolumeManager

The Logical Volume Manager included with Linux is supported withPowerPath Migration Enabler. Refer to “LVM-controlled nativedevices on HP-UX and Linux hosts” on page 49.

Native-named device migrations with VxVM

Note: PowerPath for AIX does not support native devices.

Migrating a native-named device under Veritas Volume Manager(VxVM) control requires additional steps, as described in theprocedure below. Follow the standard migration procedure when

migrating a pseudo-named device:


1 Follow the migration steps in Chapter 2 ”Migration Procedures ”



Volume managers 49

1. Follow the migration steps in Chapter 2, Migration Procedures,until committing the migration.

2. After committing the migration, close the application andunmount the volumes in the disk group.

3. Type:

vxdg deport <disk_group name >

4. Turn off I/O redirection with the powermig undoRedirectcommand.

5. Import the disk group into VxVM, and restart the volumes.a. Type: vxdg import <disk_group_name>

b. Type: vxvol -g <disk_group_name> startall

6. Mount the volumes command.

7. Clean up the migration with the powermig cleanup command.

Note: During CLARiiON to CLARiiON Host Copy migration with I/Orunning on iSCSI devices, powermig cleanup -format takes substantiallylonger since it performs a full format of the disk. Not valid for Linux.

Optionally, use vxdisk resize to resize the logical unit that isunder VxVM control. If dealing with an encrypted LUN, usepowervt resize on applicable platforms and then vxdisk resizefollowed by growfs to reclaim unused space.

LVM-controlled native devices on HP-UX and Linux hosts

When migrating native-named devices controlled by LVM on HP-UXor Linux hosts:

1. Run powermig commit.

Note: On Linux, native to pseudo is supported on RHEL 5 and SLES 10SP3 only. Native to native is not supported on Linux.

Choose a convenient time to take applications or file systemsoffline and reconfigure them.

2. After committing the migration, close the application or unmountfile systems that use the volume group containing the

source-device name.


3 Deactivate the specified volume group Type:




3. Deactivate the specified volume group. Type:

vgchange -a n <VG >

where:<VG> is the name of the volume group

Note: You must close the logical volumes before executing thiscommand. For example, if the logical volume contains a file system,unmount the file system.

4. Capture all devices in the volume group so you can later import

them back into the system. Type:

vgexport /tmp/ VG _export_file <VG >

5. Edit the device mapping from the source to the target-devicenames. Use this command syntax:

sed -e s/source /target/ /tmp/ VG_export_file

/tmp/VG_import_file

where:<source> = source-device name in the migration<target> = target-device name in the migration<VG> = name of the volume group

Note: Do not use the fully qualified path names (/dev/rdsk/c#t#d#)when specifying the device names.

6. Turn off the redirection of I/O from the source to the target. Type:

powermig undoRedirect -hd <handle >

7. Make a directory, create a special file, and recreate thevolume-group device node entry to import. Type:

Note: The minor number must be unique among existing volumegroups.

mkdir /dev/<VG> mknod /dev/<VG>/group c 64 <minor_number>

8. Import the volume group back into the system. Type:

vgimport -f /tmp/ VG_import_file

9. Set the LVM volume group availability. Type:

vgchange -a y <VG >


10. Create or update the LVM volume group configuration backup



Migrations in a cluster environment 51

10. Create or update the LVM volume group configuration backupfile. Type:

vgcfgbackup <VG >

11. Restart applications or mount file systems that use the volumegroup.

12. Optionally, clean up temporary files:

rm /tmp/VG_export_file

/tmp/VG_import_file

Note: If the LVM-controlled device does not use LVM volumes forstorage, you must reconfigure the application to use the newtarget-device name before restarting it.

13. Clean up the migration with powermig cleanup.

Note: During CLARiiON to CLARiiON Host Copy migration with I/Orunning on iSCSI devices, powermig cleanup -format takes substantially

longer since it performs a full format of the disk. Not valid for Linux.

Migrations in a cluster environment

When migrating devices in a cluster configuration, you must selectone node to remain active during the migration. All failover groups

must be moved to the active node.

Migrations in an MSCS cluster

Note: This procedure is supported only with the Host Copy technology type.

To migrate a source device in a Microsoft Cluster Server (MSCS),

follow these steps:

1. Ensure the target disk is initialized and accessible from all nodes.

2. Select the node that will remain active during the migration.(Only one node with access to the source and target can remainactive.) Move all failover groups to the active node.

3. Shut down the remaining nodes in the cluster.


4. Ensure the underlying technology is configured, and then run




y g gy g ,powermig setup.

5. Begin syncing with the powermig sync command.

6. When the synchronization completes, type:

powermig info -query

7. Select the target logical unit with the powermig selectTargetcommand.

8. On Windows 2003, when ready to commit the migration, skip this

step and proceed to step 9.

On Windows 2008, bring the failover group(s) offline. This stopsthe applications.

9. Commit the migration with the powermig commit command.

10. On Windows 2003, skip this step and proceed to step 11.

On Windows 2008, bring the failover group (s) online again. This

starts the applications.

11. Clean up the migration with the powermig cleanup command.

Note: During CLARiiON to CLARiiON Host Copy migration with I/Orunning on iSCSI devices, powermig cleanup -format takes substantiallylonger since it performs a full format of the disk (this does not refer toWindows).

12. Start the remaining nodes and balance the I/O load as desired,ensuring that the source logical unit is inaccessible to the host. If LUN masking is not available, you can remove the physical path.Take care when removing paths to the source if the path is shared

by other logical units.


Mi ti ith HACMP




Migrations with HACMP

Note: This section applies only to AIX hosts.

PowerPath Migration Enabler supports migrations in a standbyconfiguration, where one cluster node is active and the other ispassive.

In the following procedure:

◆

hdiskpower7 = the source device◆ hdiskpower9 = the target device

Note: Steps 3 through 11 in the following procedure must be done for eachsource and target migration pair.

1. Map the intended target device(s) to the active and passive nodes,and then run the cfgmgr command.

2. Stop the HACMP service on the passive node.

3. On the active node, set up the migration and synchronize thedevices, as described in Chapter 2, ”Migration Procedures.”

4. On the active node, run powermt display dev=all and note theunique device ID for the target device (hdiskpower9) and thesource device (hdiskpower7) in the migration. The unique device

ID consists of the following information:• Symmetrix device ID = Symmetrix ID + logical device ID

• CLARiiON device ID = CLARiiON ID + logical device ID(LUN)

5. On the passive node:

a. Run powermt display dev=all, and locate the device whose IDmatches the device ID of the target device, noted in step 4.

(That is, the device ID associated with hdiskpower9 on theactive node.)

b. On the passive node, run the following command on thehdiskpower device identified in step 5 a:

chdev -a pv=yes -l hdiskpowerX

where:

X = the hdiskpower number identified in step 5.


For example:

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-




chdev -a pv=yes -l hdiskpower9

6. Verify the device you modified in step 5 b has the same PVID asthe source device on the active node. (Identified in step 4, in thisexample, the source-device name is hdiskpower7). Type:

lspv | grep hdiskpowerX

where:

X = the source device on the active node

7. Complete the remaining migration steps, as outlined inChapter 2, ”Migration Procedures.”

Note: After completing the migration steps, the device names on theactive node only are swapped.

8. On the active node, determine the device ID of the original sourcedevice (in this example, hdiskpower7, which is now the target

device on the active node).

9. On the passive node, run powermt display dev=all, and locatethe device whose device ID is the same as the ID of the originalsource device (hdiskpower7, which is now the target device onthe active node).

10. On the passive node, run this command on the hdiskpowerdevice identified in step 9:

chdev -a pv=clear -l hdiskpowerX

where:

x = the device identified in step 9.

For example:chdev -a pv=clear -l hdiskpower7

11. Verify each node has only one device name associated with thePVID of the target device on the active node. Use the 1spvcommand.

12. Synchronize the cluster on both nodes and make suresynchronization completes.

13. Start HACMP on the passive node, and ensure the cluster servicestarts.


14. Fail the resource group from the active to the passive node andif th V l G i ti t d

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-




verify the Volume Group is activated.

Migrations in a VCS environment

Follow the steps in the example below to migrate data betweenpseudo-named array devices in a Veritas Cluster Server (VCS)environment for Host Copy and Open Replicator technologies.

Note: PowerPath must be installed and properly configured on all nodes in

the cluster. The VCS cluster must be configured with at least one servicegroup running on the active node. The service group must be associated witha VxVM disk group with EMC devices. "Configuring applications andresources in VCS" in the Veritas Cluster Server User Guide provides additionalinformation.

1. Zoning recommendations:

Currently, VCS is only working properly in a single FA zoning forSymmetrix devices environment (each hosts HBA is zoned peronly one of the Symmetrix FAs). CLARiiON array ports should bevisible on Symmetrix arrays, and must be added to zoningconfiguration to perform PowerPath Migration Enabler OpenReplicator migrations from CLARiiON to Symmetrix.

HBA1 - FA1, SPA1, SPB0 (hba1, Target Symm FA1, SourceCLARiiON SP ports).

HBA2 - FA2, SPA0, SPB1 (hba2, Target Symm FA2, SourceCLARiiONs SP ports):


# powermt display paths | more




Symmetrix logical device count=14

==============================================================================----- Host Bus Adapters --------- ------ Storage System ----- - I/O Paths -### HW Path ID Interface Total Dead==============================================================================

4 lpfc 000187461234 FA 7dA 14 05 lpfc 000187461234 FA 7cA 14 0

CLARiiON logical device count=16==============================================================================----- Host Bus Adapters --------- ------ Storage System ----- - I/O Paths -### HW Path ID Interface Total Dead==============================================================================

4 lpfc APM00090500212 SP B0 16 04 lpfc APM00090500212 SP A1 16 05 lpfc APM00090500212 SP A0 16 05 lpfc APM00090500212 SP B1 16 0

2. In the active node, configure PowerPath Migration Enabler Open

Replicator or Host Copy as the underlying technology:

◆ Register Symmetrix ports on a CLARiiON side and addSymmetrix to CLARiiON storage group with VCS devicesconnected to nodes.

◆ Mask CLARiiON devices to the Symmetrix. Refer to the EMCPowerPath Migration Enabler User Guide.

3. Display information for devices associated with the servicegroup:

# vxdisk list -o alldgs | grep mydgemcpowerd auto:cdsdisk mydg01 mydg onlineemcpowere auto:cdsdisk mydg02 mydg onlineemcpowerf auto:cdsdisk mydg03 mydg online

4. Display source device information using the powermt command:

# powermt display dev=emcpowerd | grep "device ID\|LUN "Logical device ID=60060160574023003472AFDDA10ADE11 [LUN 7]

# powermt display dev=emcpowere | grep "device ID\|LUN "Logical device ID=60060160574023003672AFDDA10ADE11 [LUN 9]

# powermt display dev=emcpowerf | grep "device ID\|LUN "Logical device ID=60060160574023004172AFDDA10ADE11 [LUN 20]


5. Select target devices and display information using the powermtcommand:




command:

# powermt display dev=emcpowerw | grep "device ID\|LUN "Logical device ID=03A3

# powermt display dev=emcpowerx | grep "device ID\|LUN "Logical device ID=03A5

# powermt display dev=emcpowerz | grep "device ID\|LUN "Logical device ID=04C3

6. Before starting PowerPath Migration Enabler migrations, EMChighly recommends freezing any service group containing diskgroups associated with the migrations.

# hagrp -freeze service_group -persisitent

7. In the active node (service group is currently running) startPowerPath Migration Enabler migrations between source andtarget devices, and continue to completion.

# powermig setup -techType OR -src <source_dev> -tgt<target_dev> -no

# powermig sync -hd <handle_number> -no# powermig selecttarget -hd <handle_number> -no# powermig commit -hd <handle_number> -no# powermig cleanup -hd <handle_number> -no

Use the following to monitor the current state of the migration,

especially during the syncing state:# powermig info -query -all

8. Verify that emcpower names have swapped between source andtarget devices. Repeat Step 6 on page 57 and Step 7 on page 57 toverify the changes. If I/O was running to the disks during themigration, verify that there was not interruption or data

corruption.9. From the active node display disk information for the VxVM disk

group being used:

# vxdisk list -o alldgs | grep mydgemcpowerd auto:cdsdisk mydg01 mydg onlineemcpowere auto:cdsdisk mydg02 mydg onlineemcpowerf auto:cdsdisk mydg03 mydg onlineemcpowerw auto:cdsdisk - (mydg) online udid_mismatch

emcpowerx auto:cdsdisk - (mydg) online udid_mismatch


emcpowerz auto:cdsdisk - (mydg) online udid_mismatch




10. Import disks labeled as udid_mismatch as cloned disks by

assigning a new disk group name:

# vxdg -n newdg -o useclonedev=on -o updateid import mydg# vxdisk list -o alldgsemcpowerd auto:cdsdisk mydg01 mydg onlineemcpowere auto:cdsdisk mydg02 mydg onlineemcpowerf auto:cdsdisk mydg03 mydg onlineemcpowerw auto:cdsdisk - (newdg) online clone_diskemcpowerx auto:cdsdisk - (newdg) online clone_diskemcpowerz auto:cdsdisk - (newdg) online clone_disk

11. Unfreeze the service groups that were previously set to freezestate in Step 6 on page 57.

# hagrp -unfreeze service_group -persisitent

12. Fail resource group to a passive node by using the ClusterManagement Console or the following command:

# hagrp -switch service_group -to linx116

13. Verify that the service group has successfully switched to theother node:

# hastatus -sum

14. Verify that I/O is still running after switching the service group tothe new active node.

15. In the new active node, verify that the emcpower devicescurrently used in the VxVM disk group correspond to thephysical devices used as targets during the PowerPath MigrationEnabler migration.

Note: Names of pseudo devices may be different from the original node.

# powermt display dev=emcpowerd | grep "device ID\|LUN "Logical device ID=03A3

# powermt display dev=emcpowere | grep "device ID\|LUN "Logical device ID=03A5

# powermt display dev=emcpowerf | grep "device ID\|LUN "Logical device ID=04C3


Migrations in a Sun Cluster




g

For pseudo to pseudo migrations with VxVM:1. Configure target array devices only on the primary node.

2. Setup a migration on the primary node.

3. Sync the migration.

4. Prior to running powermig selectTarget, perform the followingcluster specific steps that prevent failover during migration:

Note: Full cluster functionality will not be available from Step 4 throughStep 12.

a. Set the number of the secondary node for VxVM disk group to1 if it is more than one:

Type: scconf -c -D name=<DGNAME >,numsecondaries=1

b. Remove all secondary nodes from the VxVM device group:

Type: scconf -r -Dname=<DG_NAME >,nodelist=<nodename >

c. Verify that no secondary node is listed:

Type: scstat -D

5. Run powermig selectTarget.6. Run powermig commit.

7. Configure target array devices on all secondary nodes.

8. Update the VxVM database on the primary node:

a. Type: rm /etc/vx/disk.info

b. Run: vxconfigd -kc

9. Run powermig cleanup.

Note: During CLARiiON to CLARiiON Host Copy migration with I/Orunning on iSCSI devices, powermig cleanup -format takes substantiallylonger since it performs a full format of the disk.

10. Optionally rename the pseudo devices on all secondary nodes.

11. Update the VxVM database on all secondary nodes:


a. Type: rm /etc/vx/disk.info

b. Run: vxconfigd -kc




b. Run: vxconfigd kc

12. Add all secondary nodes to the VxVM device group as follows:

Type: scconf -a -Dtype=vxvm,name=<DG-NAME >,nodelist=nodename

13. Set the number of secondary node for VxVM DG as required(optional: in case if more than one secondary node is required):

Type: scconf -c -Dname=<DGNAME >,numsecondaries=<no_of_secondarynode >

Note: Quorum device migration is not supported. You can add a newdevice (quorum) from the target array and then remove the older one.


Renaming a pseudo device

Follow the steps below to rename a pseudo device:




p p

Note: Renaming a pseudo device is not supported on HP-UX.

1. Run: nvxdisk offline <devices need to rename >.

2. Check free available pseudo instances:

Type: emcpadm getfreepseudos –n <# of pseudo instances>

3. Rename source pseudo to available free pseudo.4. Rename target pseudo to source pseudo name.

5. Type: vxdisk online <source pseudo name >.





AInvisible Body Tagt

Migration Procedure



Migration Procedure Examples 63

This appendix provides information about the PowerPath MigrationEnabler technology types, and includes migration procedureexamples for each of the technology types as follows:

◆

Host Copy migrations ....................................................................... 64◆ Encapsulation migrations ................................................................. 68◆ Open Replicator migrations ............................................................. 72◆ TimeFinder/Clone migrations......................................................... 77

g

Examples

Migration Procedure Examples

Host Copy migrations




Host Copy migrations work in conjunction with the host operatingsystem to migrate data from the source logical unit to the target. AHost Copy migration does not use or require a direct connection

between the arrays containing the source and target logical units.

Host Copy procedure

The following steps apply to Host Copy type migrations forPowerPath:

1. Type the following command to setup the migration from thesource to the target, and to specify the technology type:

powermig setup -src <sourceName > -tgt <targetName >-techType <techType >

Example:

powermig setup -src emcpower72a -tgt emcpower52a-techType hostcopy

a. Optionally, you can set throttle during Setup, but cannotchange it until the migration enters the Syncing state. Refer tothe specific technology type sections in this chapter for throttleinformation and refer to “Understanding throttle impact” onpage 36.

2. From the Setup state, start the bulk copying and cloning of application write I/Os:

Type: powermig sync -handle <handle >

a. Optionally, you can set a throttle value of 0 through 9 in theSyncing state. Refer to the specific technology type sections inthis chapter for throttle information and refer to

“Understanding throttle impact” on page 36. b. Optionally, you can pause the migration during peak I/O

times. When you resume the migration, it continues fromwhere it left off. See “Pause and resume commands” onpage 40 and the PowerPath Family CLI and System MessagesReference Guide for more information.

c. Optionally, you can abort the migration.

3. While the migration is syncing, check the status. Type:


powermig query -handle <handle >

4. Once synchronized, the migration automatically goes into the



Host Copy migrations 65

SourceSelected state.5. Designate the target logical unit as the recipient of all I/O read

requests:

Type: powermig selectTarget -handle <handle >

Note: Before you commit the migration, you have the option of returningto the SourceSelected state or aborting the migration.

6. Type the following to commit the migration to designate thetarget as the recipient of all I/O requests:

powermig commit -handle <handle >

Note: There is no returning to the source after you perform a commit.

a. With native operating system devices as the source, the

commit command places the migration into theCommittedAndRedirected state.

b. Also with native devices, you must execute the undoRedirectcommand to point to the new device and restart theapplication. This places the migration into the Commit state.

7. Clean up the migration. Type:

powermig cleanup -handle <handle >

The migration ends after the cleanup step.

8. Complete the “Post-migration steps” on page 28.


Host Copy throttle settings




The Host Copy technology type copies data from source to target incycles that last five seconds each. Table 5 on page 66 shows thepercentage of the host’s time spent copying data as defined by thethrottle value. The default value for Host Copy is 2.

Host Copy throttle settings are disabled when you enable the HostCopy ceiling function. “Host Copy ceiling settings” on page 67provides additional information.

The EMC PowerPath Migration Enabler Host Copy White Paperavailable on EMC Powerlink provides additional details aboutthrottle settings.

Note: You can put as many Host Copy migrations into the Syncing state asdesired; however only eight will be actively syncing at any given time. InHost Copy, a daemon in UNIX and a Service in Windows manages the bulkdata copy. The daemon or Service initiates periodic calls to the kernel toperform copy operations in 5 second cycles, and maintains a separate logshowing the progress of each migration it is managing.

During synchronization, Host Copy reads from the source device andwrites to the target device in chunks of 256 KB.

Note: Synchronizing multiple migrations impacts host performance. One ormore migrations may be paused, or throttle values may be modified toreduce application impact.

Table 5 Host Copy throttle values

Throttle Value 0 1 2 3 4 5 6 7 8 9

Percentage of time the hostspends copying data

100% 60% 36% 22% 13% 7.8% 4.7% 2.8% 1.7% 1.0%


Host Copy ceiling settings

http://-/?-

http://-/?-



Host Copy migrations 67

When you enable Host Copy ceiling, this disables per migrationthrottling. Host Copy ceiling is specified on the powermig optionscommand. “Understanding Host Copy ceiling copy rate” on page 39provides additional information.

Encrypted devices

PowerPath Encryption with RSA is a software-based encryption tool

that leverages RSA technology to encrypt user-specifiedPowerPath-controlled devices. Writes are encrypted and reads aredecrypted to and from the encrypted LUN. Data must be migrated toan encrypted target device to encrypt plain text data. MigrationEnabler Host Copy is a non-disruptive method for performing thismigration. The PowerPath Encryption with RSA User Guide onPowerlink provides more information.

Note: When migrating to an encrypted device, the target must be at least65 KB (AIX, Solaris, or Windows) and 68 KB (Linux) larger than the sourcedevice to accommodate metadata requirements.


Encapsulation migrations




To set up a storage element for PowerPath Migration Enabler,identify the LU that will be the source for the migration and expose itto the Invista (for Symmetrix LUs, set up masking; for CLARiiONLUs, add it to the appropriate storage group). Then, using the InvistaElement Manager (IEM):

1. Right-click on Invista Instance and select Rediscover BackendSEs and Paths.

2. Right-click an unimported Storage Element that corresponds tothe source device (compare the "Unique ID" to the CLARiiONWWN or to the Symmetrix ID and LUN), and select Set NotReady.

Putting the element in the Not Ready state (also called theFaulted state in some IEM views) disables I/O to and from theStorage Element and prevents from using it in any Clones, Data

Mobility, or volume resizing operations.3. Import the Storage Element.

4. Create a Virtual Volume from the Storage Element. Make thevolume the full size of the Storage Element.

Note: You cannot expand this Virtual Volume until after the migrationhas been completed.

5. Add the volume to a Virtual Frame and expose it to the host asyou would any other Virtual Volume.

The EMC Invista Element Manager Administration Guide containsmore information on Virtual Frames and Virtual Volumes.

6. On the host running PowerPath Migration Enabler, make thevolume visible to the host. Follow the steps for your operating

system.

Platform Command or procedure

Solaris Type:

devfsadm -C

AIX Type:

emc_cfgmgr or

cfgmgr -v


Linux To make LUNs visible on a Linux host, refer to Dynamic LUN




Encapsulation migrations 69

On UNIX or Linux hosts7. Update the configuration. Type:

powermt config

8. Type:

powermt save

Identify target-device name

Once the Virtual Volume is visible on the host, identify thetarget-device name to use when you set up the migration session. Theexamples in this section use pseudo-device names in a Solaris hostenvironment. For example: emcpower44a:

To identify the device names for a migration type:

powermt display nonvirtual

The nonvirtual information is displayed for a device only if thatdevice is qualified to be the target of an INVE (or) migration:

◆ The Virtual Volume is associated with exactly one StorageElement.

◆ The Storage Element is associated with exactly one VirtualVolume.

◆ The Virtual Volume and the Storage Element have the samecapacity.

◆

The Storage Element has been set to Not Ready.

addition and removal in the EMC PowerPath Installation and Administration Guide for Linux .

Windows Use the Disk drives option in the Device Manager utility to list thestorage devices visible to the host:

a. Open the Device Manager, click Start > Settings > ControlPanel.

b. Double-click System, click the Hardware tab, and then clickDevice Manager.

c. Click Disk drives to see the storage devices visible to the host.d. If necessary, rescan the storage devices using the Rescan

Disk option.

e. Refresh the PowerPath Administrator view if AutoRefresh is notenabled. Right-click the root node and select Refresh.


In the nonvirtual output:

Pseudo name = the target-device name used setting up a

i i




migration.Nonvirtual pseudo name = the source-device name used whensetting up a migration.

The following output is truncated:

Pseudo name=emcpower44aID=APM00051002822

Logical device ID=60001440B0601C6D00589C10390003B2

NonVirtual Symmetrix ID=000187461238NonVirtual device ID=067CNonVirtual pseudo name=emcpower15astate=alive; policy=ADaptive; priority=0; queued-IOs=0==============================================================================---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---### HW Path I/O Paths Interf. Mode state Q-IOs Errors==============================================================================2306 pci@1e,600000/lpfc@2 c5t6d2s0 10 active alive 0 0

2306 pci@1e,600000/lpfc@2 c5t7d2s0 11 active alive 0 02307 pci@1f,700000/lpfc@2 c6t6d2s0 0C active alive 0 02307 pci@1f,700000/lpfc@2 c6t7d2s0 0D active alive 0 0

Pseudo name=emcpower9aID=APM00051002822

Logical device ID=60001440B0601C6D00589C80390003B6NonVirtual device ID=60060160EEE215000A5F05D3A915DB11NonVirtual pseudo name=emcpower35a

state=alive; policy=ADaptive; priority=0; queued-IOs=0==============================================================================---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---### HW Path I/O Paths Interf. Mode state Q-IOs Errors==============================================================================2306 pci@1e,600000/lpfc@2 c5t6d4s0 10 active alive 0 02306 pci@1e,600000/lpfc@2 c5t7d4s0 11 active alive 0 02307 pci@1f,700000/lpfc@2 c6t6d4s0 0C active alive 0 02307 pci@1f,700000/lpfc@2 c6t7d4s0 0D active alive 0 0

Note: HP-UX displays differently. You would not seePseudo name=emcpower as shown in the output above. Also, inNonVirtual pseudo name=emcpower the word pseudo would bereplaced with the word device.

For more information on the powermt display nonvirtual output,refer to the PowerPath CLI and System Messages Reference Guide,available on Powerlink.


Encapsulation procedure

Th f ll i t l t t i ti f P P th



Encapsulation migrations 71

The following steps apply to type migrations for PowerPath:1. Type the following command to setup the migration from the

source to the target, and to specify the technology type.

powermig setup -src <sourceName> -tgt <targetName >-techType <techType>

Example:

powermig setup -src emcpower72a -tgt emcpower52a

-techType INVE

2. Move to the Syncing state:

Type: powermig sync -handle <handle >

a. Optionally, you can abort the migration.

Encapsulation migrations are instantaneously synchronized;

there is no bulk copying or cloning of I/Os. Once synchronized,the migration automatically goes into the SourceSelected state.

Note: Throttle and Pause and Resume commands are not supported forthe Encapsulation migrations.

3. Check the status. Type:


4. Designate the target logical unit as the recipient of all I/O readrequests: Type:

powermig selectTarget -handle <handle >


5. Type the following to commit the migration to designate thetarget as the recipient of all I/O requests:



a. With native operating system devices as the source, thepowermig commit places the migration into the

CommittedAndRedirected state.


b. In this case, the application must now be reconfigured to pointto the new (target) device and restarted. Then executepowermig undoRedirect to place the migration into the

Commit state




Commit state.

6. Remove host access to the source device. Unmask the LU on aSymmetrix array or remove the LU from the storage group on aCLARiiON array.




8. Complete the“Post-migration steps” on page 28.

Open Replicator migrations

PowerPath Migration Enabler works in conjunction with the OpenReplicator hot pull method, which copies data through the SANinfrastructure from a remote logical unit to a Symmetrix storagesystem. Migration Enabler mirrors I/O to keep the source and targetlogical units synchronized throughout the migration process. Formore information on Open Replicator, refer to the EMC SolutionsEnabler Symmetrix Open Replicator CLI Product Guide, available onPowerlink.

Symmetrix-to-Symmetrix migrations

Set up your environment for a Symmetrix-to-Symmetrix migrationwith Open Replicator:

1. Run powermt display dev=all to verify the host can see thesource and target devices.

2. Verify and update the zoning configuration as necessary. If

zoning is enabled on the switch, every port on the target arraythat allows access to the target logical unit must have access to thesource logical unit through at least one port on the source array.

3. If the storage port with the source logical unit is running VolumeLogix (VCM), grant permission to access the source logical unitthrough the target storage port.


4. On the host running PowerPath Migration Enabler, make thevolume visible to the host. Follow the steps for your operatingsystem in the table in Step 4 on page 74.

5 I ll i h S l i E bl S l i E bl hi li



Open Replicator migrations 73

5. Install either Solutions Enabler or Solutions Enabler thin client onthe host.

On UNIX or Linux hosts1. Run powermt config to update the configuration.

2. Type: powermt save.

CLARiiON-to-Symmetrix migrations

To migrate data from a CLARiiON storage system to a Symmetrixstorage system:

1. Run powermt display dev=all to verify the host can see thesource and target devices.

2. Verify and update the zoning configuration as necessary. If zoning is enabled on the switch, every port on the target arraythat allows access to the target logical unit must have access to thesource logical unit through at least one port on the source array.

3. Add the Symmetrix target WWN to the Storage Group thatcontains the source logical unit(s).

Note: A Symmetrix storage port must be registered as an initiator to be

added to a CLARiiON Storage Group. You must try to initiate aPowerPath Migration Enabler session for the Symmetrix to appear inconnectivity status.


4. On the host running PowerPath Migration Enabler, make thevolume visible to the host. Follow the steps for your operatingsystem.




On UNIX or Linux hosts

5. Run: powermt config6. Run: powermt save

Migration Enabler supports Solutions Enabler (SE) thin client forOpen Replicator migrations. Refer to “Solutions Enabler thin clientfor Migration Enabler” on page 43 for additional information.


Solaris Type:

devfsadm -C

AIX Enter:

emc_cfgmgr | cfgmgr -v

Linux To make LUNs visible on a Linux host, refer to Dynamic LUNaddition and removal in the EMC PowerPath Installation and Administration Guide for Linux .

Windows Use the Disk drives option in the Device Manager utility to list thestorage devices visible to the host:

a. Open the Device Manager, click Start > Settings > ControlPanel.

b. Double-click System, click the Hardware tab, and then click

Device Manager.c. Click Disk drives to see the storage devices visible to the host.

d. If necessary, rescan the storage devices using the RescanDisk option.

e. Refresh the PowerPath Administrator view if AutoRefresh is notenabled. Right-click the root node and select Refresh.


Open Replicator procedure

The following steps apply to Open Replicator type migrations forPowerPath:



Open Replicator migrations 75

PowerPath:

1. Setup the migration from the source to the target, and specify thetechnology type:

powermig setup -src <sourceName > -tgt <targetName >-techType <techType>

Example:

powermig setup -src emcpower72a -tgt emcpower52a-techType OR

If using the SE thin client, specify the server that has the fullSolutions Enabler.

powermig setup -src <sourceName > -tgt <targetName >-techType OR -techHost <hostname >

Example:

powermig setup -src harddisk7 -tgt harddisk5 -techTypeOR -techHost 192.168.1.1

a. Optionally, you can set throttle during Setup, but cannotchange it until the migration enters the Syncing state. Refer tothe specific technology type sections in this chapter for throttleinformation and refer to “Understanding throttle impact” on

page 36.

Note: In the case of Open Replicator, throttle can be set during Setup, but can only be changed once the migration has started syncing.

2. In the Syncing state, start the bulk copying and cloning of application write I/Os. Type:

powermig sync -handle <handle >

a. Optionally, you can set a throttle value of 0 through 9 in theSyncing state. Refer to the specific technology type sections inthis chapter for throttle information and refer to“Understanding throttle impact” on page 36.


b. Optionally, you can pause the migration during peak I/Otimes. When you resume the migration, it continues fromwhere it left off. Refer to “Pause and resume commands” on

page 40 and the PowerPath Family CLI and System Messages




p g y y gReference Guide for more information.


Once synchronized, the migration automatically goes into theSourceSelected state.



4. Designate the target logical unit as the recipient of all I/O readrequests. Type:



5. Enter the following to commit the migration to designate thetarget as the recipient of all I/O requests. Type:



a. With native operating system devices as the source, the

commit command places the migration into theCommittedAndRedirected state.

b. You must execute the undoRedirect command to point to thenew device and restart the application. This places themigration into the Commit state.


powermig cleanup -handle <handle >The migration ends after the cleanup step.

7. Complete the“Post-migration steps” on page 28.


Open Replicator throttle settings

For Open Replicator migrations, the throttle value adjusts the speedat which data is transferred between the source and target logical



TimeFinder/Clone migrations 77

g gunits. For Open Replicator, the throttle works similarly to theSolutions Enabler symrcopy command with the set pace argument. Alower throttle value can significantly increase resource consumptionon the target logical unit’s array.

TimeFinder/Clone migrationsThe TimeFinder/Clone technology is used to copy from LUNs withinthe same Symmetrix array.

Migration Enabler supports Solutions Enabler (SE) thin client forTimeFinder/Clone migrations. Refer to “Solutions Enabler thin clientfor Migration Enabler” on page 43 for additional information.

TimeFinder/Clone procedure

The following steps apply to TimeFinder/Clone type migrations forPowerPath:

1. Setup the migration from the source to the target, and specify thetechnology type:

powermig setup -src <sourceName > -tgt <targetName >-techType <techType>

Example:

powermig setup -src emcpower72a -tgt emcpower52a-techType symclone

If using the SE thin client, specify the server that has the fullSolutions Enabler.

powermig setup -src <sourceName > -tgt <targetName >-techType symclone -techHost <hostname >

Example:

powermig setup -src harddisk7 -tgt harddisk5 -techTypesymclone -techHost 192.168.1.1


Note: Symclone is the name that represents the TimeFinder/Clonetechnology type.




a. Optionally, you can set throttle during Setup, but cannotchange it until the migration enters the Syncing state. Refer tothe specific technology type sections in this chapter for throttleinformation and refer to “Understanding throttle impact” onpage 36.

b. Optionally, you can set the thin client. See “Solutions Enabler

thin client for Migration Enabler” on page 43 for additionalinformation.

2. In the Syncing state, start the bulk copying and cloning of application write I/Os. Type:

powermig sync -handle <handle >

a. Optionally, you can set a throttle value of 0 through 9 in the

Syncing state. Refer to the specific technology type sections inthis chapter for throttle information and refer to“Understanding throttle impact” on page 36.

b. Optionally, you can pause the migration during peak I/Otimes. When you resume the migration, it continues fromwhere it left off. Refer to “Pause and resume commands” onpage 40 and the PowerPath Family CLI and System MessagesReference Guide for more information.


Once synchronized, the migration automatically goes into theSourceSelected state.



4. Designate the target logical unit as the recipient of all I/O readrequests. Type:



5. Commit the migration to designate the target as the recipient of

all I/O requests. Type:




a With native operating system devices as the source the



TimeFinder/Clone migrations 79

a. With native operating system devices as the source, thecommit command places the migration into theCommittedAndRedirected state.

b. You must execute the undoRedirect command to point to thenew device and restart the application. This places themigration into the Commit state.




7. Complete the“Post-migration steps” on page 28

TimeFinder/Clone throttle settingsFor TimeFinder/Clone migrations, the throttle value changes theSymmetrix Quality of Service (QoS) value for the source and targetlogical units. The QoS value affects all I/O performance for thedevices, and not just migration performance. The powermig setupcommand sets the QoS value to 0, if not specified with the throttleoption during setup. After a migration completes, the QoS value setduring the migration remains on the source and target.

Symmetrix QoS values range from 0 to 16 and Migration Enablerthrottle values range from 0 to 9. Migration Enabler maps the throttlevalues to QoS values as shown in Table 6 on page 79.

Table 6 TimeFinder/Clone throttle and QoS values

Throttle Value 0 1 2 3 4 5 6 7 8 9

Corresponding QoS value 0 2 4 5 7 9 10 12 14 16


http://-/?-

http://-/?-

http://-/?-




BInvisible Body Tag

Troubleshooting aMigration



Troubleshooting a Migration 81

This appendix provides guidance for troubleshooting problems youmay encounter during a migration:

◆ Audit and error log messages .......................................................... 82◆ Reporting a problem.......................................................................... 83◆ Graceful host reboot .......................................................................... 84◆ Host crash............................................................................................ 85◆ Other reported faults ......................................................................... 87◆ Resolving a process hang.................................................................. 88

Troubleshooting a Migration

Audit and error log messages

PowerPath Migration Enabler uses audit logs for recording eventsthat change the migration state, and error logs that capturet d t th t d i th f i ti Th




unexpected events that occur during the course of a migration. ThePowerPath Family CLI and System Messages Reference Guide onPowerlink describes audit logging, including instructions for settingup logging in your environment. It also lists Migration Enabler errormessages, descriptions, and corrective action suggestions.

Audit log message example

The following audit log entry indicates that a powermig synccommand was executed on Migration Handle number 3 onDecember 21st at approximately 3 p.m. The state of the migration isnow syncing:

Dec 21 15:00:23 Ippa020 emcpAudit: PPME: Info:

handle=3,event=stateChanged, state=syncing,cmd=sync

Error log message example

The following error log entry indicates that the powermig setupcommand failed because either the source or the target device wasalready a part of another Migration Enabler migration session:

Dec 21 14:49:16 Ippa020 PPME: API: Error: Setup failed:

PPME error(14): Devices already involved in a

migration


Reporting a problem

If you encounter a problem with PowerPath Migration Enabler 5.5,run the emcgrab utility, which collects system and softwareconfiguration information and then follow the instructions in this



Reporting a problem 83

configuration information, and then follow the instructions in thissection to collect Migration Enabler information. You can downloademcgrab from the Powerlink website. Choose the following links:Support > Product and Diagnostic Tools > Grab Utilities. Refer tothe emcgrab ReadMe.txt file on Powerlink for prerequisiteinformation and instructions for using the utility.

Note: EMC Grab 3.9 or later is required to collect information aboutPowerPath Encryption with RSA or Host Copy migrations. EMC Grabversion 3.7 or later collects information about PowerPath Migration Enablerfor Open Replicator or Encapsulation.


Graceful host reboot

When a host gracefully shuts down, there is no impact to themigration, and PowerPath Migration Enabler Host Copy records acheckpoint in the Migration Enabler database The migration




checkpoint in the Migration Enabler database. The migrationprogress checkpoint is recorded approximately once every 30 secondson a local or SAN disk, depending on where the boot disk is located.

When the host starts up after a graceful shutdown, the migrationresumes from its last checkpoint. This behavior is for migrations thatare syncing or paused.

Note: Graceful host reboot applies only to Host Copy.


Host crash

When a host crashes, PowerPath Migration Enabler cannot guaranteethe success of application writes, and cannot guarantee that the dataon the source and target devices are the same after such a crash. The



Host crash 85

on the source and target devices are the same after such a crash. Thefault state (either TargetLUfault or SourceLUfault) appears in thepowermig info output. PowerPath Migration Enabler reactsdepending on the state of the migration at the time of the crash.Faults are recorded in the Syncing, SourceSelected, andTargetSelected states. Table 7 on page 86 shows the three states when

Logical Unit faults can be reported.

Note: This does not apply to migrations.

States before the host crash:

◆ Syncing

◆ SourceSelected

◆ TargetSelected

Each state causes a different result when the host is rebooted after thecrash:

1. In the Syncing state, the synchronization has been stopped.PowerPath Migration Enabler reports a TargetLUFault. Themigration is returned to the Setup state. Application reads andwrites are still directed to the source device. The migration can berestarted with the powermig sync command.

2. In the SourceSelected state, the migration has completed. Bothsource and target devices are fully synchronized. However, whenthe host crashes, PowerPath Migration Enabler reports aTargetLUFault. I/O writes may have been in flight during thetime of the crash. Therefore, Host Copy can no longer guaranteethat the data is the same on both the source and the target. The

migration returns to the Setup state where it can be restarted withthe powermig sync command.

3. In the TargetSelected state, the migration has completed. Bothsource and target devices are fully synchronized. However, whenthe host crashes, PowerPath Migration Enabler reports aSourceLUFault. I/O writes may have been in flight during the


time of the crash. Therefore, Host Copy can no longer guaranteethat the data is the same on both the source and the target. Themigration proceeds to the Commit state.

R t d f lt ft h t h




When a host crashes, PowerPath Migration Enabler returns errors to both reads and writes for any device that is in a migration where thesource and target is kept synchronized; until PowerPath MigrationEnabler startup recover is run, which is done automatically. Thisprotects data integrity until device default states are processed. Once

startup recovery is run, it is safe for reads and writes to flow to thesedevices.

Table 7 Reported faults after a host crash

state at time ofHost Crash

Hostcrash

Fault statereported

state aftercrash

Powermigcommand

Dataloss

Syncing Yes TargetLUfault Returns t o Setupstate

sync orcleanup

No

SourceSelected Yes TargetLUfault Returns toSetupstate

sync orcleanup

No

TargetSelected Yes SourceLUfault pseudo: movesto Committedstate

cleanup No

TargetSelected Yes SourceLUfault native: moves toCommitted andRedirected state

undoRedirect No


Other reported faults

In addition to a host crash, errors or faults may occur due to changesin the storage, the host, or the SAN. Table 8 on page 87 shows thethree states when Logical Unit faults are reported. If an I/O failure

http://-/?-

http://-/?-



Other reported faults 87

g poccurs in the course of a write I/O operation, and if MigrationEnabler is currently cloning write I/Os (migrations do not clonewrites or reads) and then Migration Enabler records a Logical Unitfault, the fault state (either TargetLUfault or SourceLUfault) appearsin the powermig info output. Migration Enabler reports different

types of faults depending on the state of the migration at the time theI/O failure occurs. Faults are recorded in the Syncing,SourceSelected, and TargetSelected states.

States at the time of the fault:

◆ Syncing

◆ SourceSelected

◆

TargetSelectedTable 8 Other reported faults

Current state

Sourcedevicefailure

Targetdevicefailure

Fault statereported Cause

Powermigcommand

Dataloss

Syncing Yes N/A SourceLUfault I/O write tosource fails

aborta

a. When this fault occurs under these conditions, errors will be returned on allI/Os until you perform this transaction.

No

Yes TargetLUfault I/O write totarget fails

abort No

SourceSelected Yes N/A SourceLUfault I/O write tosource fails

selectTargetb No


abort No

TargetSelected Yes N/A SourceLUfault I/O write tosource fails

commit No


abortc No


b. When this fault occurs under these conditions, errors will be returned on allI/Os until you perform this transaction.

c. When this fault occurs under these conditions, errors will be returned on all

I/Os until you perform this transaction.




Resolving a process hang

Some powermig commands do not respond when you enter CTRL-C.If a powermig command hangs, you will be unable to enter anysubsequent powermig commands. To clear the hung process, use the

kill -9 command (UNIX only).

CInvisible Body Tag

Support Summary



Support Summary 89

This appendix provides a table of the commands supported for thevarious platforms and technology types.

◆ Migration Enabler support summary ............................................. 90

Support Summary

Migration Enabler support summary

Table 9 on page 90 lists for all versions of PowerPath MigrationEnabler the commands supported with each platform and technologytype.

http://-/?-

http://-/?-




Table 9 PowerPath Migration Enabler command support (1 of 2)

Commands Supported technology types

Versions 5.0a & 5.1 Version 5.2 Version 5.3 Version 5.5

UNIXb &Windows Linux

Solaris &Windows

AIX, Linux,Solaris,Windows

Windows,Linux

powermig abort INVE, OR OR HOSTCOPY,INVE, OR

HOSTCOPY,INVE, OR,SYMCLONE

HOSTCOPY,INVEc, OR,SYMCLONE powermig

cleanup

powermigcommit

powermiggetHandle

powermig help

powermig info

powermigoptions N/A N/A N/A N/A Host Copy(Linux only)

powermig pause Not supported Not supported HOSTCOPY, OR HOSTCOPY, OR,SYMCLONE

HOSTCOPY,OR,SYMCLONE

powermig query INVE, OR OR HOSTCOPY,INVE, OR



powermigrecover

powermigresume

Not supported Not supported HOSTCOPY, OR HOSTCOPY, OR,SYMCLONE


Support Summary

powermigselectSource

INVE, OR OR HOSTCOPY,INVE, OR

HOSTCOPY,INVE, OR,

SYMCLONE

HOSTCOPY,INVE, OR,

SYMCLONE powermigselectTarget

Table 9 PowerPath Migration Enabler command support (2 of 2)



Migration Enabler support summary 91

powermig setup

powermig sync

powermigthrottle

Not supported Not supported HOSTCOPY, OR HOSTCOPY, OR,SYMCLONE


powermig version

INVE, OR OR HOSTCOPY,INVE, OR



a. Version 5.0 is the first Migration Enabler release.

b. UNIX includes AIX, HP-UX and Solaris.

c. Encapsulation (INVE) is only supported with Invista storage systems. It is not supported for VPLEX storage systems.

Support Summary




DInvisible Body Tag

Disk Labelingpowerformat

Command for Solaris



Disk Labeling powerformat Command for Solaris 93

This appendix provides detailed information about the powerformatcommand for Solaris.

◆ Solaris and PowerPath Migration Enabler disk labeling ............. 94

Disk Labeling powerformat Command for Solaris

Solaris and PowerPath Migration Enabler disk labeling

When you perform a heterogeneous migration with PowerPathMigration Enabler on a Solaris host, the source logical unit’s disklabel is copied to the target along with the source’s data. PowerPathMigration Enabler changes the ASCII name of the target logical unit’s




g g g gdisk label to one that matches the logical unit type of the target array.

For example, when migrating data from a CLARiiON logical unit(DGC-RAID-0206) to a Symmetrix logical unit(EMC-SYMMETRIX-5671) using Open Replicator as the underlying

technology, the CLARiiON logical unit’s disk label (DGC-RAID-0206)is copied to the target Symmetrix logical unit. InCLARiiON-to-Symmetrix migrations only, PowerPath MigrationEnabler changes portions of the ASCII disk label name to match thetarget logical-unit type. In this example, the ASCII name of the disklabel is changed from DGC-RAID-0206 cyl… toEMC-SYMMETRIX-5671 cyl…. This change to the ASCII disk-labelname ensures the correct logical-unit type (for example,

EMC-SYMMETRIX-5671) appears when you run the Solaris formatutility.

Multiple array type entries in the format utility outputPowerPath Migration Enabler changes the ASCII disk-label name of the target logical unit as part of the migration process; however, thedisk geometry on the target logical unit is unchanged. (It has the diskgeometry of the source.) You can change the disk geometry and

partition information after a migration using the powerformatcommand. Refer to “Viewing and updating disk labels withpowerformat” on page 96 for more information.

Usually the format type command displays one entry for each drivetype. After a heterogeneous migration, the format type commandsees a different geometry for the target (Symmetrix) logical unit. Itsees the source (CLARiiON) logical unit’s geometry. The logical-unitlabel EMC-SYMMETRIX-5671 appears in the format output twice

because the two Symmetrix 5671 logical units are recognized asdifferent array types.

format exampleThe following example shows the format utility before and after aCLARiiON logical unit (DGC-RAID-0206) is migrated to a Symmetrixlogical unit (EMC-SYMMETRIX-5671) using Open Replicator as theunderlying technology.


Before the migration

The Solaris format utility shows the following output.

Note: A logical unit with the EMC-SYMMETRIX-5671 identification exists onthe host before the migration is performed.

f t>t



Solaris and PowerPath Migration Enabler disk labeling 95

format>type

AVAILABLE DRIVE TYPES:0. Auto configure1. Quantum ProDrive 80S2. Quantum ProDrive 105S

3. CDC Wren IV 94171-3444. SUN01045. SUN04246. SUN1.0G7. SUN1.058. SUN1.3G9. SUN2.1G10. SUN2.9G11. Zip 10012. EMC-SYMMETRIX-556813. EMC-SYMMETRIX-5671 <— original drive type14. DGC-RAID5-0206

After the migration

After you migrate a CLARiiON source logical unit with disk labelDGC-RAID-0206 to a Symmetrix target logical unit with disk labelEMC-SYMMETRIX-5671, format displays the following output.

Note: Two entries now appear for EMC-SYMMETRIX-5671.

format>typeAVAILABLE DRIVE TYPES:

0. Auto configure1. Quantum ProDrive 80S2. Quantum ProDrive 105S3. CDC Wren IV 94171-3444. SUN0104

5. SUN04246. SUN1.0G7. SUN1.058. SUN1.3G9. SUN2.1G10. SUN2.9G11. Zip 10012. EMC-SYMMETRIX-556813. EMC-SYMMETRIX-5671 <— original drive type

14. DGC-RAID5-0206


15. EMC-SYMMETRIX-5671 <— new drive type with a differentgeometry

The format utility recognizes that the original logical unit and the

migrated logical unit have different disk geometries. Therefore, twoentries for EMC-SYMMETRIX-5671 appear.

Note: More duplicate drive entries are created with PowerPath Migration




Note: More duplicate drive entries are created with PowerPath MigrationEnabler and Open Replicator if you migrate from logical units with differentgeometries to the same type of EMC-SYMMETRIX-5671 logical units.

Exception to disk-label renaming

If PowerPath Migration Enabler cannot open the target logical unit(for example, because all slices are exclusively opened), thepowermig cleanup command displays the following message:

Unable to edit array name in disk label

EMC does not recommend manually editing the disk label. However,you can use the powerformat utility to update the label, as describedin the next section.

Viewing and updating disk labels with powerformat

On Solaris hosts, a logical unit’s disk label contains information aboutthe vendor, product, geometry, and slices. PowerPath MigrationEnabler includes a utility called powerformat that allows you tosafely update disk-label information, preserve partition definitions

and data, and make newly available disk capacity available for use.When you migrate data to a target logical unit that is larger than thesource logical unit, powerformat can adjust the disk label andpartition table to make available the entire disk capacity on the targetlogical unit. The powerformat command takes partitions thatextended to the end of the currently labeled logical unit (the labelmigrated from the source logical unit), and extends them to the endof the target logical unit. This allows you to use the additional space

on the target.

The powerformat three main operation modes are as follows:

◆ Query — Allows you to view the current disk informationassociated with the specified logical unit. Refer to “Query fordisk-label information” on page 97.


◆ Preview — Allows you to view the current disk informationassociated with the specified logical unit, as well as new diskinformation calculated by powerformat. You can preview any

changes before implementing them. Refer to “Preview diskinformation” on page 98.

◆ Execute — Executes disk-label changes on the specified disk.Refer to “Install a new disk label” on page 102

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://-/?-




Refer to Install a new disk label on page 102

Query for disk-label information

The powerformat utility allows you to query the specified logicalunit for information about the logical unit and its label. The output

includes the logical unit’s vendor, product, and revision; and modesense, disk geometry, and VTOC information.

Table D-1 Disk-label

Action

To view disk information for the specified logical unit, run:

powerformat <device >

Where:

<device > = device name whose disk information you want to display.

Example:

powerformat /dev/rdsk/c1t2d0s0

Output

Inquiry: "EMC ", "SYMMETRIX ", "5568"

Mode Sense: pcyl 17726, nhead 15, nsect 64, capacity17016960

GEOM: pcyl 10000, ncyl 9998, acyl 2, nhead 10, nsect 32

VTOC: "DGC-RAID5-0217 cyl 9998 alt 2 hd 10 sec 32"

v_part p_start p_size------ ------- ------

0 0 263040

1 263040 2630402 0 31993606 526080 2673280

Notes

Refer to Table 10 on page 100 for an explanation of the output. The output in this example isa subset of the output in Table 10 on page 100. Refer to the relevant output lines and theircorresponding descriptions.


Preview disk information

Preview mode allows you to view the current disk information aswell as the new disk information calculated by powerformat without

implementing any changes. The procedure in this section does notwrite a new disk label.

Table D-2 Preview disk

http://-/?-

http://-/?-




Action

To preview the new disk information for the specified logical unit, type:

powerformat -<n|g|a|w> [-p <partition>] <device >

Where:<device > = device name whose disk information you want to change.

-a installs a new ASCII text label. The ASCII text label includes the vendor, product, and revision; and thenumber of cylinders, alternate cylinders, heads, and sectors associated with the logical unit. This optiondoes not change the disk geometry.

-g installs a new geometry and partitions based on the label cylinder geometry of the specified logicalunit.

The new partition table shows the information calculated by powerformat. When a logical unit hasmore space than is designated by the partition table (for example, when you migrate data to a targetlogical unit that is larger than the source) powerformat displays how the partitions would beextended to take advantage of the additional disk space. By default, powerformat grows all partitionsthat extend to the end of the currently labeled logical unit.

The -g option uses the label cylinder geometry of the specified logical unit, and increases thenumber of cylinders to accommodate a larger target logical unit than the label implies. If themaximum number of cylinders allowed on a logical unit is reached before the entire disk is covered,then the maximum number of cylinders is used, and additional space becomes unavailable. Whenthe logical unit is not in use, use the -n option to avoid this scenario.

-n tries the native disk geometry before installing a new geometry, as follows:

Note: Do not use the -n option while the logical unit is in use. Use the -g option instead.

• If the partitions can be preserved using the native geometry, and the native geometry can berepresented by the VTOC label, then the native geometry is installed.

• Otherwise, if the label cylinder geometry does not result in wasted disk space, then the label cylindergeometry is installed. In this case, the number of cylinders is increased to cover the whole disk.

• Otherwise, a new geometry that can preserve partitions and cover as much of the disk as possible isinvented and installed.

By default, powerformat grows all partitions that extend to the end of the currently labeled logical unit.

-p allows you to specify partitions to preserve. Without this option, all partitions are preserved. When youspecify partitions, powerformat selects a new geometry based on the partitions you want topreserve. It then preserves other partitions as possible given the new geometry, and deletes theremaining partitions. To specify multiple partitions, include multiple -p options on a single command line.

-w tells powerformat to extend only the partitions that were whole disk partitions before the migration.Partitions that start in the middle of the disk and continue to the end are not extended.


Output

Inquiry: "EMC " "SYMMETRIX " "5568"Mode Sense: pcyl 17726 hd 15 sec 64 cap 17016960GEOM (before): cyl 9998 alt 2 hd 10 sec 32 cap 3200000 VTOC (before): "DGC-RAID5-0217 cyl 9998 alt 2 hd 10 sec 32"

t t t i

Table D-2 Preview disk (continued)




Table 10 on page 100 explains the output shown in the previousexamples.


0 0 2630401 263040 2630402 0 3199360

6 526080 2673280GEOM (after): cyl 53176 alt 2 hd 10 sec 32 cap 17016960 VTOC (after): "EMC-SYMMETRIX-5568 cyl 53176 alt 2 hd 10 sec 32"v_part p_start p_size------ ------- ------

0 0 2630401 263040 2630402 0 170163206 526080 16490240

Notes

Refer to Table 10 on page 100 for an explanation of the output.


Table 10 Output explanation: preview mode

Output Line Description

Inquiry: "EMC" "SYMMETRIX" "5568" Inquiry displays the vendor, product, and revision information for thespecified logical unit.

This information does not derive from the disk label. When you runpowerformat on the target logical unit after a migration, the inquirydata for the target displays




data for the target displays.

Mode Sense: cyl 17726 hd 15 sec 64cap 17016960

Mode Sense displays the physical logical unit’s mode sense data,including the number of physical cylinders, heads, and sectors; andthe disk capacity.

This information does not derive from the disk label. When you runpowerformat on the target logical unit after a migration, the modesense data for the target displays.

GEOM (before): cyl 9998 alt 2 hd 10sec 32 cap 3200000

GEOM displays the number of cylinders, alternate cylinders, heads,and sectors; and the disk capacity.

This information derives from the disk label. When you runpowerformat on the target logical unit after a migration, thesource logical unit’s disk geometry displays.

VTOC (before): "DGC-RAID5-0217 cyl9998 alt 2 hd 10 sec 32"

VTOC displays the VTOC information for the logical unit, includingthe type of array and the number of cylinders, alternate cylinders,heads, and sectors.

This information derives from the disk label. When you runpowerformat on the target logical unit after a migration, the sourcelogical unit’s VTOC information displays (except in cases whereMigration Enabler automatically updates parts of the ASCIIname—as in CLARiiON-to-Symmetrix migrations).


0 0 2630401 263040 2630402 0 31993606 526080 2673280

The partition table for the specified logical unit.v_part — volume partition number

p_start — sector number where the partition starts

p_size — number of sectors in the partition

This information derives from the disk label. When you runpowerformat on the target logical unit after a migration, the sourcelogical unit’s partition table displays.

Note: If the target to which you migrate data is larger than the

source, you cannot take advantage of additional space on the targetlogical unit until you update the target’s partition table.

GEOM (after): cyl 53176 alt 2 hd 10sec 32 cap 17016960

The new geometry for the specified logical unit. This new geometryis installed if you execute the changes.

The display includes the number of cylinders, alternate cylinders,heads, and sectors; and the disk capacity.


VTOC (after): "EMC-SYMMETRIX-5568cyl 53176 alt 2 hd 10 sec 32"

VTOC (after) displays the new VTOC information. The displayincludes the type of array and the number of cylinders, alternatecylinders, heads, and sectors.

In this example, the CLARiiON logical unit information copied to thetarget during the migration would be replaced by the target logical

Table 10 Output explanation: preview mode

Output Line Description




g g g p y g gunit’s Symmetrix information shown here.

v_part p_start p_size

------ ------- ------

0 0 2630401 263040 263040

2 0 17016320

6 526080 16490240

This output shows the new partition information.

In this example, the values in the p_size column for partitions 2 and

6 increased compared with the before partition information earlier inthis table.

This output shows how partitions 2 and 6 would be extended if youinstalled the new label.


Install a new disk label

Before you install a new disk label, preview the changes as describedin “Preview disk information” on page 98. Once you preview the

information and decide you want to install a new label, you canexecute the changes by including the -x option.

Table D-3 Install new disk label




Action

To install the new disk label, type:

powerformat -x <n|g|a|w> -p <partitions > <device >

Where:<device >—device name whose disk information you want to change

-xn executes the changes previewed with the powerformat -n <device > command. -xn tries the nativedisk geometry before installing a new geometry.

Note: Do not execute this command while the logical unit is in use. If you want to run powerformat while

the logical unit is in use, then use the -g option instead of -n.

-xg — Executes the changes previewed with the powerformat -g <device > command. -xg installs the

new disk-label geometry and the current and new partition tables.-xa — Executes the changes previewed with the powerformat -a <device > command. -xa installs thenew ASCII disk label.

Note: Refer to “Preview disk information” on page 98 for more information on these options.

-w — Tells powerformat to extend only the partitions that were whole-disk partitions before the migration.Partitions that start in the middle of the disk and continue to the end are not extended.

-p — Allows you to specify particular partitions to preserve. Without this option, all partitions are

preserved.Examples:

powerformat -xg /dev/rdsk/c1t2d0s0

powerformat -xg -p 0 -p 6 /dev/rdsk/c1t5006048AD52CB8EDd342s0


Output

powerformat -xg /dev/rdsk/c1t2d0s0Inquiry: "EMC " "SYMMETRIX " "5568"Mode Sense: pcyl 17726 hd 15 sec 64 cap 17016960GEOM (before): cyl 9998 alt 2 hd 10 sec 32 cap 3200000 VTOC (before): "DGC-RAID5-0217 cyl 9998 alt 2 hd 10 sec 32"

Table D-3 Install new disk label (continued)




( ) yv_part p_start p_size------ ------- ------

0 0 2630401 263040 263040

2 0 31993606 526080 b2673280

GEOM (after): cyl 53176 alt 2 hd 10 sec 32 cap 17016960 VTOC (after): "EMC-SYMMETRIX-5568 cyl 53176 alt 2 hd 10 sec 32"


0 0 2630401 263040 263040

2 0 170163206 526080 16490240

powerformat -xg -p 1 -p 2 /dev/rdsk/emcpower17cInquiry: "EMC " "SYMMETRIX " "5671"Mode Sense: pcyl 25314 hd 15 sec 64 cap 24301440GEOM (before): cyl 32766 alt 2 hd 8 sec 16 cap 4194304 VTOC (before): "EMC-SYMMETRIX-5671 cyl 32766 alt 2 hd 8 sec 16"


------ --------- ---------0 0 1310721 131072 2621442 0 41940486 393216 3800832

Using max dkg_ncyl. Wasting capacity.GEOM (after): cyl 65535 alt 2 hd 8 sec 16 cap 8388736 VTOC (after): "EMC-SYMMETRIX-5671 cyl 65535 alt 2 hd 8 sec 16"


------ --------- ---------0 0 1310721 131072 2621442 0 83884806 393216 7995264


Inquiry: "EMC " "SYMMETRIX " "5771"Mode Sense: pcyl 1916 hd 15 sec 128 cap 3678720

GEOM (before): cyl 2078 alt 2 hd 15 sec 64 cap 1996800 VTOC (before): "EMC-SYMMETRIX-5267 cyl 2078 alt 2 hd 15 sec 64"

v_part p_start p_size------ --------- ---------

0 0 2054402 0 1994880

Table D-3 Install new disk label (continued)




2 0 19948804 205440 2054405 410880 2054406 616320 1024320

GEOM (after): cyl 3830 alt 2 hd 15 sec 64 cap 3678720 VTOC (after): "EMC-SYMMETRIX-5771 cyl 3830 alt 2 hd 15 sec 64"

v_part p_start p_size------ --------- ---------

0 0 2054402 0 36768004 205440 2054405 410880 2054406 616320 1024320

Notes

Refer to Table 10 on page 100 for an explanation of the output.

Index




Aabort

command 34audit logs 82

Cceiling copy rate 34CLARiiON to Symmetrix migration 73Cleanup 26cleanup

command 35cluster environments 51commands

migration enabler 34comments by email 10

commitcommand 35

Commit state 26Committed state 26Committed state, 17CommittedAndRedirected 17CommittedAndRedirected state 26

Ddisk labelinstall 102previewing 98querying 97updating 102

disk labeling 94documentation, related 7

EEFI label 29Encapsulation 13

procedure 71encapsulation

post-migration steps 28encryption

migrations 67Encryption with RSA

prerequisites 22error logs 82Execute

powerformat 97

F

feedback by email 10

GgetHandle

command 35Graceful host reboot 84

Hhelp

command 35Host Copy 13

throttle settings 66host copy ceiling 39

powermig options 39Host Copy type migrations 64host crash 85, 87

Hostcopy

Index

post-migration steps 28

Iinfo

command 34install

Linux 18UNIX 18

with volume managers 48MirrorView 22

Nnative devices 13

O

http://-/?-

http://-/?-




UNIX 18Windows 19

Install new disk label 102installation 18

installing Migration Enabler 18interoperability

cluster environments 51Ionix ControlCenter 28

Llabeling

disk 94

license keys 18Linux 18Logical Volume Manager 48Logical Volume Manager support 48

MMicrosoft Cluster Server 51migration

states 15Migration Enableroverview 12

migration enablercommands 34

migration session 15Migration states

Committed 17CommittedAndRedirected 16

Setup 16SourceSelected 16Synching 16TargetSelected 16

migration workflow 15migrations

in cluster environments 51prerequisites 22procedure examples 63

Open Replicator 14migrations 72prerequisites 23procedure 75throttle settings 77

overviewMigration Enabler 12

Ppatch information 9pause

command 34pausing a migration 40powerformat 93

utility 45powerformat utility 96powermig

commit 16setup 16sync 16

undoRedirect 17powermig cleanup 27, 52powermig commands 15powermig commit 52powermig options

hst copy ceiling 39powermig pause 40powermig resume 40, 41powermig selectTarget 52

powermig setup 24, 44throttle 38

powermig sync 24powermig throttle 36prerequisites

migrations 22Preview

powerformat 97Preview disk information 98

Index

problemreporting 83

procedure 71

procedure examplesmigrations 63

proceduresmigration 31

process hang 88d d i

powerformat utility 96Solaris hosts 28Solutions Enabler 14

Solutions Enabler (SE) thin client 43source

size 22SourceSelected state 25states

i ti 15



107EMC PowerPath Migration Enabler Version 5.5 User Guide

pseudo devicerenaming 61

pseudo devices 13

QQuery

powerformat 96query

command 34

R

reclaim unused space 30Linux 31Windows 31

recovercommand 35

related documentation 7renaming

pseudo device 61Reported faults 86

resumecommand 34

resuming a paused migration 41

SselectSource

command 35selectTarget

command 35session

migration 15Setup 24setup 34

command 34Setup state 16SMI label 28Solaris

migration 15steps

migration 31Sun Cluster

migrations 59support summary table 90support, technical 9Symclone. See also TimeFinder/Clone 36, 38Symmetrix Remote Data Facility 22Symmetrix-to-Symmetrix migration 72sync 34

command 34

Synchronization 24synchronization speed 36Syncing state 24, 28

Ttarget device 22TargetSelected state 25, 26technical support 9

technology types 13Hostcopy 64thin client 74thin devices 42throttle

command 34throttle settings

Host Copy 66TimeFinder/Clone 79

TimeFinder®/Clone 14TimeFinder/Clone

prerequisites 23procedure 77QoS (Quality of Service) values 38, 79throttle 38throttle settings 79throttle value 38, 79

Timefinder/Clone

Index

prerequisites 22troubleshooting 81tuning migration speed 36

UundoRedirct

command 35unused space

http://-/?-

http://-/?-




preclaim 30

update disk-label information 96updating disk labels (Solaris) 96

VVeritas Cluster Server 55Veritas Volume Manager 48version

command 35virtually provisioned devices 42volume managers 48

WWindows

prerequisites 23workflow

migration 15

Documents

Migrations Using Powerpath