300-008-703

EMC® PowerPath®/VEfor VMware vSphereVersion 5.4 and Service Packs

Installation and Administration GuideP/N 300-008-703

REV A11

EMC CorporationCorporate Headquarters:

Hopkinton, MA 01748-9103

1-508-435-1000www.EMC.com

2

Copyright © 2011 EMC Corporation. All rights reserved.

Published April 2011

EMC believes the information in this publication is accurate as of its publication date. The information issubject to change without notice.

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 regulatory document for your product line, go to the Technical Documentation andAdvisories section on EMC Powerlink.

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.

PowerPath/VE for VMware vSphere V5.4 and Service Packs Installation and Administration Guide

Contents

Preface

Chapter 1 PowerPath/VE OverviewPowerPath/VE for VMware ESX and ESXi platforms ................ 14

PowerPath Multipathing Plugin ..............................................14Components of a PowerPath/VE configuration.......................... 15

PowerPath/VE multipathing software ...................................15Remote PowerPath CLI (rpowermt) server ............................15ELM license server..................................................................... 16

Differences from other PowerPath platforms............................... 18PowerPath/VE features ................................................................... 21

Dynamic path failover ...............................................................21Proactive path testing and automatic path recovery.............21Dynamic multipath load balancing .........................................22Storage system types ..................................................................22PowerPath/VE and NMP coexistence.....................................25

Chapter 2 Installing PowerPath/VEInstallation components and workflow......................................... 28Task 1: Installing PowerPath/VE for VMware vSphere ............. 33

Supported PowerPath/VE installation scenarios ..................33PowerPath/VE pre-installation tasks ......................................33Installing PowerPath/VE using VMware vCLI .....................37Installing PowerPath/VE using VUM.....................................39Upgrading PowerPath/VE .......................................................45Upgrading PowerPath/VE in a PowerPath/VE-controlled boot-from-SAN setup .......................................45PowerPath/VE post-installation tasks ....................................48

PowerPath/VE for VMware vSphere V5.4 and Service Packs Installation and Administration Guide 3

Contents

Task 2: Installing the PowerPath/VE remote CLI (rpowermt).. 49PowerPath/VE rpowermt installation requirements............ 49PowerPath/VE rpowermt pre-installation tasks ................... 49PowerPath/VE rpowermt installation procedure................. 50PowerPath/VE rpowermt upgrade procedure...................... 51PowerPath/VE rpowermt post-installation tasks ................. 52

Task 3: Installing and configuring PowerPath/VE licenses....... 53PowerPath/VE licenses prerequisites ..................................... 53Installing and configuring PowerPath/VEelectronic licenses ....................................................................... 54PowerPath/VE licenses post-installation andconfiguration tasks ..................................................................... 59Removing licenses ..................................................................... 63Re-installing and configuring after converting licensetypes ............................................................................................. 63Rehosting .................................................................................... 64

Chapter 3 Managing PowerPath/VEManaging claim rules....................................................................... 66

Listing claim rules ...................................................................... 68Adding claim rules..................................................................... 69Preventing PowerPath/VE from claiming devices ............... 79Renumbering claim rules .......................................................... 79

Configuration changes in PowerPath/VE .................................... 81Adding new paths to a PowerPath/VE logical device ............... 82

Path limit..................................................................................... 82Procedure.................................................................................... 82

Adding new logical devices to a configuration............................ 83Procedure.................................................................................... 84

Removing paths or logical devices from a configuration ........... 85Procedure to remove an entire HBA ....................................... 85Procedure to remove logical devices ....................................... 86

Managing PowerPath/VE using VMware commands ............... 87Managing PowerPath/VE using rpowermt ................................. 88

Setting the rpowermt lockbox .................................................. 89Configuring path latency monitoring ..................................... 91

Boot-from-SAN configuration management ................................ 92Diagnostic partition.................................................................... 92

PowerPath/VE for VMware vSphere V5.4 and Service Packs Installation and Administration Guide4

Contents

Chapter 4 Removing PowerPath/VETask 1: Removing PowerPath/VE from a VMware ESX host .... 94

Pre-removal tasks....................................................................... 94Removing PowerPath/VE.........................................................94Removing PowerPath/VE in a PowerPath/VE-controlled boot-from-SAN setup ..............................................95Post-removal tasks ......................................................................96

Task 2: Removing the PowerPath/VE rpowermt ........................ 97PowerPath/VE rpowermt pre-removal tasks.........................97Removing PowerPath/VE rpowermt ......................................97Post-removal tasks ......................................................................98

Chapter 5 Troubleshooting PowerPath/VE InstallationCollecting logs from vSphere Center Server orvSphere client................................................................................... 100Log files............................................................................................. 101TCP port conflicts............................................................................ 102Problems when performing vihostupdate .................................. 103Incompatible rpowermt lockbox versions................................... 104Reporting a problem....................................................................... 105

Appendix A Files Installed and Modified by PowerPath/VEFiles installed and modified by the PowerPath/VEinstallation........................................................................................ 108

Appendix B rpowermt Command Line interfacerpowermt command summary ..................................................... 110rpowermt syntax summary ........................................................... 112rpowermt error and warning messages....................................... 113rpowermt server lockbox ............................................................... 117

rpowermt CST libraries environment variable.....................118rpowermt check_registration: Check PowerPath/VElicense information.......................................................................... 120rpowermt config: Retrieve CLARiiON nice names.................... 121rpowermt display: Monitor HBAs and devices.......................... 122rpowermt display latency: Monitor path latency....................... 128rpowermt display options: Displays option settings forstorage-system classes .................................................................... 131rpowermt display paths: Display path information .................. 132rpowermt display ports: Display port information ................... 135

5PowerPath/VE for VMware vSphere V5.4 and Service Packs Installation and Administration Guide

Contents

rpowermt help: Display rpowermt command syntax............... 139rpowermt register: Register a PowerPath/VE license onan ESX host ...................................................................................... 141rpowermt restore: Restore paths .................................................. 142rpowermt set mode: Set paths to active or standby mode ....... 144rpowermt set path_latency_monitor: Enables or disablespath latency monitoring ................................................................ 147rpowermt set path_latency_threshold: Sets a time limitfor the completion of I/O .............................................................. 148rpowermt set periodic_autorestore: Enable or disableperiodic autorestore ....................................................................... 150rpowermt set policy: Set the load-balancing and failoverpolicy ................................................................................................ 151rpowermt set streamio_threshold: Set I/O threshold count .... 153rpowermt setup add_host ............................................................. 155rpowermt setup list_hosts ............................................................. 157rpowermt setup remove_host....................................................... 158rpowermt setup update_host........................................................ 159rpowermt unregister: Remove a PowerPath/VE licensefrom an ESX host............................................................................. 160rpowermt update lun_names: Retrieve CLARiiONnice names........................................................................................ 161rpowermt version: Display PowerPath/VE version number .. 162

Appendix C PowerPath/VE MessagesError log messages.......................................................................... 164VM kernel errors............................................................................. 165

Operation .................................................................................. 165Message sources....................................................................... 165

PowerPath/VE error messages .................................................... 167

Glossary

Index

PowerPath/VE for VMware vSphere V5.4 and Service Packs Installation and Administration Guide6

Preface

As part of an effort to improve and enhance the performance and capabilitiesof its product lines, EMC periodically releases revisions of its hardware andsoftware. Therefore, some functions described in this document may not besupported by all versions 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 document, please contact your EMC representative.

Audience andprerequisites

This document is part of the PowerPath documentation set. It isintended for use by storage administrators and other informationsystem professionals responsible for installing, using, andmaintaining PowerPath.

Readers of this manual are expected to be familiar with the VMwareESX host, storage-system management, and the applications usedwith PowerPath.

Relateddocumentation

The following is a complete set of PowerPath documentation; allmanuals are available from EMC Corporation:

◆ PowerPath Product Family Guide◆ PowerPath Family CLI and System Messages Reference Guide◆ PowerPath for VMware vSphere Installation and Administration Guide◆ PowerPath/VE for VMware vSphere Licensing Guide◆ PowerPath for AIX Installation and Administration Guide◆ PowerPath for HP-UX Installation and Administration Guide


8

Preface

◆ PowerPath for Linux Installation and Administration Guide◆ PowerPath for Solaris Installation and Administration Guide◆ PowerPath and PowerPath/VE for Windows Installation and

Administration Guide◆ PowerPath for VMware Release Notes◆ PowerPath Family for Solaris Release Notes◆ PowerPath Family for AIX Release Notes◆ PowerPath Family for Linux Release Notes◆ PowerPath and PowerPath/VE Family for Windows Release Notes◆ PowerPath Migration Enabler User Guide◆ PowerPath Encryption with RSA User Guide◆ PowerPath Management Pack for Microsoft Operations Manager User

Guide

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

If your environment includes Symmetrix storage systems, refer alsoto the EMC host connectivity guides, which are available on thePowerlink website.

If your environment includes VNX™ Operating Environment (OE)and CLARiiON storage systems, refer also to the following manuals:

◆ EMC host connectivity guides

◆ VNX OE Storage-System Support website

(www.emc.com/vnxsupport)

◆ CLARiiON Storage-System Support website

(www.emc.com/clariionsupport)

If your environment includes other vendors’ storage systems, refer tothe appropriate documentation from your vendor.

Another related EMC interoperability resource is the E-LabInteroperability Navigator, available on the Powerlink website:http://Powerlink.EMC.com.

PowerPath/VEdocumentation set

changes

Note the following changes in the PowerPath/VE documentation set:

Beginning with PowerPath/VE 5.4 SP2, we are including a revisionhistory in all documents of the documentation set. In thePowerPath/VE documents that include information for versions


http://powerlink.emc.com


www.emc.com\clariionsupport

http://www.emc.com/clariionsupport

Preface

previous to version 5.4 SP2, the revision history reflects changes tothe document from version 5.4 SP2 and moving forward.

Revision history The following table presents the revision history of this document:

Revision Date Description

A08 September 9, 2010 Modification of the following sections in relation to therelease of PowerPath/VE 5.4 SP2 for VMwarevSphere:• “Differences from other PowerPath platforms” on

page 18• “PowerPath/VE pre-installation tasks” on page 33• “Installing PowerPath/VE using VUM” on page 39• “Problems with PowerPath/VE upgrade” on

page 46• “Upgrade order for PowerPath/VE and VMware

ESX” on page 47• “PowerPath/VE rpowermt pre-installation tasks” on

page 49• “PowerPath/VE licenses prerequisites” on page 53• “Verify PowerPath/VE configuration” on page 61• “Managing claim rules” on page 66• “Listing claim rules” on page 68• “Managing PowerPath/VE using rpowermt” on

page 88• “Setting the rpowermt lockbox” on page 89• “Log files” on page 101• “rpowermt command summary” on page 110• “rpowermt syntax summary” on page 112• “rpowermt error and warning messages” on

page 113• “rpowermt set mode: Set paths to active or

standby mode” on page 144• “rpowermt set policy: Set the load-balancing and

failover policy” on page 151• “rpowermt set streamio_threshold: Set I/O

threshold count” on page 153

A09 September 10, 2010 Addition of “PowerPath/VE rpowermt upgradeprocedure” on page 51.


10

Preface

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.

IMPORTANT!An important notice contains information essential to operation ofthe software.

WARNING

A warning contains information essential to avoid a hazard that cancause severe personal injury, death, or substantial property damageif you ignore the warning.

DANGER

A danger notice contains information essential to avoid a hazardthat will cause severe personal injury, death, or substantial propertydamage if you ignore the message.

A10 March 15, 2011 Addition of:• “VNX OE Storage-System Support website” on

page 8• “PowerPath/VE Messages” on page 163.Modification of:• “Storage system claim rules” on page 67• “rpowermt set policy: Set the load-balancing and

failover policy” on page 151

A11 April 15, 2011 Modification of “rpowermt display latency: Monitorpath latency” on page 128.

Revision Date Description


Preface

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

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

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, functions, utilities

• URLs, pathnames, filenames, directory names, computernames, filenames, 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 calls, 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

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 when

shown 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


12

Preface

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 WebSupporton Powerlink. To open a case on EMC WebSupport, you must be aWebSupport customer. Information about your site configuration andthe circumstances under which the problem occurred is required.

Your comments Your suggestions help us continue to improve the accuracy,organization, and overall quality of the user publications. Please senda message to [email protected] with your opinions ofthis guide.

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.


[email protected]


1

This chapter provides an overview of PowerPath/VE for VMwarevSphere. Topics include:

◆ PowerPath/VE for VMware ESX and ESXi platforms ................. 14◆ Components of a PowerPath/VE configuration ........................... 15◆ Differences from other PowerPath platforms................................ 18◆ PowerPath/VE features .................................................................... 21

PowerPath/VEOverview

PowerPath/VE Overview 13

14

PowerPath/VE Overview

PowerPath/VE for VMware ESX and ESXi platformsPowerPath/VE uses redundant physical path components—hostbased adapters (HBAs), switches, storage processors (SPs), andcables—between an ESX and ESXi host and an external storagedevice to provide fault tolerance. If one or more path componentsfails, the ESX and ESXi host can use a viable alternate path to accessan external storage device. The process of detecting a failed path andswitching to another path is called path failover. Path failover helpsensure uninterrupted I/O between an ESX/ESXi host and externalstorage devices, allowing applications to continue to access theirdata.

PowerPath/VE also redistributes the I/O load across multiple pathsbetween an ESX/ESXi host and an external storage device. Thisprocess is called load balancing. Load balancing improves a host’sability to manage heavy I/O loads by continually balancing the loadacross all paths, eliminating the need for repeated staticreconfigurations as workloads change.

Note that throughout this document, wherever VMware ESX ismentioned, ESX also refers to VMware ESXi, unless otherwise noted.

PowerPath Multipathing PluginThe default multipathing functionality for ESX hosts is provided bythe generic Native Multipathing Plugin (NMP). The VMwarevStorage APIs for Multipathing provides a framework for integratingthird-party multipathing plugins (MPPs) into the ESX platform.Third-party MPPs, like PowerPath, run in parallel with the NMP and,for specified arrays, replace the default NMP behavior by takingcontrol of the path failover and load balancing operations.PowerPath/VE works with VMware vSphere to provide enhancedpath management capabilities to ESX and ESXi hosts. Havingmultiple paths enables the ESX and ESXi host to access a storagedevice even if a specific path is unavailable. Multiple paths can alsoshare the I/O traffic to a storage device.



Components of a PowerPath/VE configurationA PowerPath/VE configuration has the following major components:

◆ The PowerPath/VE multipathing software, which resides onVMware ESX host

◆ The remote PowerPath CLI (rpowermt) server

◆ The Electronic Licensing Management (ELM) license server

Note: The ELM license server is part of the PowerPath/VE configurationfor served licenses only. If you use unserved licenses, the ELM server isnot a component of your PowerPath/VE configuration. Figure 1 onpage 16 and Figure 2 on page 17 provide high-level views of thecomponents of the PowerPath/VE served and unserved configurations.

PowerPath/VE multipathing softwareThe PowerPath/VE multipathing software resides on the ESX host.

The PowerPath/VE multipathing software resides within the ESXkernel so that the multipathing and load-balancing functionalities aretransparent to VMware and the Guest OSes that are running onVMware. The PowerPath/VE software delivers the multipathing andload-balancing capabilities and manages license keys. The rpowermtCommon Interface Model (CIM) client enables communicationbetween the VMware ESX host and the rpowermt host (see “ELMlicense server” on page 16 for a description of the rpowermt host) formanagement of PowerPath/VE using the PowerPath/VE remoteCLI, called rpowermt. In the case of the served licenses, the rpowermtCIM client also carries out licensing queries.

Remote PowerPath CLI (rpowermt) serverThe remote PowerPath/VE, or rpowermt, server is a machine onwhich the PowerPath remote multipathing (rpowermt) CLI andunserved license files are installed. It is also responsible for acquiringserved license files from the ELM server. It can be a Virtual Machine(VM), a 64-bit RHEL 5.2 machine, or a 32-bit Windows 2003 machine.You use rpowermt to manage PowerPath/VE on an ESX host.

Figure 1 on page 16 shows the components of a PowerPath/VEenvironment for unserved licenses.

Components of a PowerPath/VE configuration 15

16


Figure 1 PowerPath/VE for VMware vSphere unserved license configuration

ELM license server The ELM license server is a component of the PowerPath/VE forVMware vSphere configuration only if you use served licenses. It is acentralized licensing management application that keeps track of thenumber of licenses in use at a given time. When a host requires aparticular license functionality, the license for that functionality ischecked out from the license pool, which is stored in the licenseserver. License keys are released back to the license pool when theyare no longer being used and are available for use by another ESXhost. The check-in and check-out queries are carried out usingrpowermt commands between the ELM server and the rpowermtserver.

PowerPath/VE for VMware vSphere uses a third-party product as itsELM license server, which is available for download on Powerlink.See “Task 3: Installing and configuring PowerPath/VE licenses” on



page 53,” in Chapter 2, “Installing PowerPath/VE,”for moreinformation.

You can install the ELM license server on a Virtual Machine (VM) aslong as the VM is running the supported operating system. See thePowerPath 5.4 and Service Pack Releases for VMware vSphere ReleaseNotes for ELM license server supported operating systems.

Figure 2 on page 17 shows a high-level view of the components of thePowerPath/VE for VMware vSphere ESX served licenseconfiguration.

Figure 2 PowerPath/VE served license configuration

ELM license server availabilityFor information on served license configurations, see thePowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, for more information.

Overdraft licenses Overdraft licenses are available for served licenses. Overdraft givesyou a grace period for peak use for a limited period of time ofPowerPath/VE for VMware vSphere in the case that you somehowcheck out more licenses than you have purchased. Overdraft licenseshave the same functionality as licenses you have purchased and haveno impact on PowerPath/VE for VMware vSphere functionality or

Components of a PowerPath/VE configuration 17

18


performance. For more information on overdraft licenses, see thePowerPath/VE for VMware vSphere Licensing Guide.

Differences from other PowerPath platformsPowerPath/VE is different from PowerPath on other supportedplatforms in the following ways:

◆ Electronic Licensing

PowerPath/VE for VMware vSphere supports Electronic LicenseManagement (ELM). The rpowermt server stores unservedlicense information.

For served licenses, PowerPath/VE for VMware vSphere uses athird-party license server to manage license information. See“ELM license server” on page 16 for more information.

Note: PowerPath/VE for VMware vSphere licenses are available onlythrough Powerlink Licensing. The LAC (license authorization code)letter, which can be physical or electronic, contains additionalinformation on obtaining a PowerPath/VE for VMware license. ThePowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, provides more information on ELM for PowerPath/VE forVMware vSphere.

◆ Behavior in different licensed states

PowerPath/VE behaves in the following ways when operating indifferent licensed states:

• Unlicensed state

When PowerPath/VE is installed but not licensed,PowerPath/VE claims devices but does not export them. Thismeans that devices show as claimed by PowerPath/VE butalso as unlicensed, and I/O to these devices is not supported.Device paths display as unlicensed when you run rpowermtdisplay dev=# as well. PowerPath/VE shows the devices butVMware Virtual Center does not. The load-balancing andfailover policy is Basic failover for Symmetrix, CLARiiON,VNX OE, Invista, and VPLEX, and No Redirect for Celerraand supported third-party devices. This is the only time that adevice is set to Basic Failover or No Redirect. I/O is notsupported when PowerPath/VE is unlicensed.



Note: The EMC PowerPath/VE 5.4 and Service Pack Releases for VMwarevSphere Release Notes provides information on supported Celerradevices.

• Licensed state

When the PowerPath/VE license is installed, PowerPath/VEoperates with full functionality and normal PowerPath/VEbehavior. The policy transitions to Symmetrix Optimized,CLARiiON Optimized, or Adaptive, as appropriate.

• Expired license state

When an existing device functions on an expired license,PowerPath/VE functions the same as a device in a licensedstate; that is, with full functionality, until the next reboot.

If a new device is added to the PowerPath/VE for VMwarevSphere environment where the license is expired, the devicebehaves the same as a device in an unlicensed state (see“Unlicensed state” on page 18). If you install a new license,the device then operates with full PowerPath/VEfunctionality, as in the licensed state.

◆ License requirement for CLARiiON AX-series

Unlike PowerPath supported on other platforms, PowerPath/VEfor VMware vSphere does not provide full load-balancing andfailover functionality exclusively to CLARiiON AX-series storagesystems without a valid PowerPath license. The ESX host musthave a valid PowerPath/VE license, even if it is connectedexclusively to CLARiiON AX-series storage arrays.

◆ Installation

PowerPath/VE for VMware vSphere has a different installationprocess from PowerPath on other supported platforms. It is amulti-task installation process that includes some of the followingsteps, depending on whether you use served or unservedlicenses:

• Installing PowerPath/VE on the ESX host

• Installing rpowermt on the remote rpowermt server

• Installing PowerPath/VE for VMware vSphere unservedlicenses on the rpowermt server

• Installing a third-party ELM license server on a remote serverfor management of served license information

Differences from other PowerPath platforms 19

20


• Installing Powerpath/VE for VMware vSphere served licenseson the ELM license server

“Installation components and workflow” on page 28 inChapter 2, “Installing PowerPath/VE,” gives a high-leveloverview of the PowerPath/VE for VMware vSphere installationprocess and provides noteworthy information. Figure 4 onpage 31 illustrates the PowerPath/VE configuration andinstallation workflow for unserved licenses. Figure 5 on page 32illustrates the PowerPath configuration and installation workflowfor served licenses.

◆ Commands

Unlike other PowerPath platforms, which use a local CLI tomanage their environment, PowerPath/VE uses a remote CLI,rpowermt, to configure and manage PowerPath and paths underits control on ESX hosts in the PowerPath for VMware vSphereenvironment.

Not all of the powermt commands that are supported on otherPowerPath platforms have an rpowermt equivalent onPowerPath/VE. Other differences between powermt supportedin other PowerPath platforms and rpowermt supported inPowerPath/VE are:

• The host option syntax in rpowermt requires specifying the IPaddress or the FQDN (fully qualified domain name) of thehost server. It also uses an equal sign (=) between the optionand its parameter.

• There are different rpowermt commands for registering andunregistering PowerPath/VE licenses. The new commandsare:

– rpowermt register– rpowermt unregister

Also, the command used for registration in other platforms,emcpreg, is not valid for PowerPath/VE for VMware vSphere.

For information about rpowermt for PowerPath/VEconfiguration and management and the supported rpowermtcommands, see “rpowermt Command Line interface” onpage 109.



PowerPath/VE featuresPowerPath/VE provides intelligent high-performance pathmanagement with path failover and load balancing optimized forSymmetrix, VNX OE, CLARiiON, and selected third-party storagesystems. PowerPath/VE supports multiple paths between an ESXhost and an external storage device. Having multiple paths enablesthe ESX host to access a storage device even if a specific path isunavailable. Multiple paths can also share the I/O traffic to a storagedevice.

PowerPath/VE is particularly beneficial in highly availableenvironments because it can prevent operational interruptions anddowntime. The PowerPath/VE path failover capability avoids hostfailure by maintaining uninterrupted application support on the hostin the event of a path failure (as long as another path is available).

Dynamic path failoverPowerPath/VE enhances application availability by eliminating theI/O path as a point of failure. With the proper hardwareconfiguration, PowerPath/VE can compensate for the failure of anycomponents in the I/O path.

If a path fails, PowerPath/VE stops sending I/O to the failed pathand checks for an active alternate path. If an active path is available,PowerPath/VE dynamically redirects I/O along that path. If noactive paths are available, alternate, standby paths (if available) arebrought into service, and I/O is routed along the alternate paths. Onactive-passive storage systems, all paths to the active SP are usedbefore any paths to the passive SP. This eliminates loss of data andapplication downtime. Failovers are transparent and nondisruptiveto applications.

This path failover and failure recovery process is transparent toapplications. (Occasionally, however, there is a short delay.)

Proactive path testing and automatic path recoveryPowerPath/VE tests live and dead paths periodically. If a live pathfails the test, PowerPath/VE marks it dead. Live testing allowsPowerPath/VE to detect path problems quickly, avoiding delays thatwould otherwise result from trying to use a defective path when I/O

PowerPath/VE features 21

22


is started on the logical device. If a failed path passes the test, it isrestored automatically, and PowerPath/VE resumes sending I/O toit.

The periodic testing of live and dead paths is low priority. Periodicautorestore is low priority as well. It is not designed to restore a pathimmediately after it is repaired, but rather to restore the path within areasonable time after it is repaired.

The time it takes for all paths to be restored varies greatly. In lightlyloaded or small configurations, paths typically are restored within anhour after they are repaired (on average, much sooner). In heavilyloaded or large configurations, it may take several hours for all pathsto be restored after they are repaired, because periodic autorestorewill be pre-empted by higher priority tasks. The storage system, host,and application remain available while the path is being restored.

The fastest way to restore paths is to use rpowermt restore.“rpowermt set mode: Set paths to active or standby mode” onpage 144 provides more information.

Dynamic multipath load balancingPowerPath/VE distributes I/O requests to a logical device across allavailable paths, thus improving I/O performance and reducingmanagement time and downtime by eliminating the need toconfigure paths statically across logical devices.

“rpowermt set policy: Set the load-balancing and failover policy” onpage 151 describes the policies supported by PowerPath/VE forVMware vSphere.

Storage system typesPowerPath/VE supports three types of storage systems:

◆ Active-active — Symmetrix, Invista, supported Celerra devices,and supported third-party array systems (see the PowerPath/VE5.4 and Service Pack Releases for VMware vSphere Release Notes andthe Interoperability Navigator, available on Powerlink, forsupported third-party arrays).

◆ Active-passive — VNX OE and CLARiiON systems.

◆ ALUA (asymmetric logical unit access) — VNX OE andCLARiiON CX systems with FLARE® version 03.26 and later).



Active-active In an active-active storage system, if there are multiple interfaces to alogical device, they all provide equal access to the logical device.Active-active means that all interfaces to a device are activesimultaneously.

In an active-active system, once PowerPath/VE creates a path set, itcan use any path in the set to service an I/O request. If a path fails,PowerPath/VE can redirect an I/O request from that path to anyother viable path in the set. This redirection is transparent to theapplication, which does not receive an error.

Active-passive In an active-passive storage system, if there are multiple interfaces toa logical device, one of them is designated as the primary route to thedevice; the device is assigned to that interface card. Typically, assigneddevices are distributed equally among interface cards. I/O is notdirected to paths connected to a nonassigned interface.

Normal access to a device through any interface card other than itsassigned one is either impossible (for example, on VNX OE andCLARiiON systems) or possible but much slower than accessthrough the assigned interface card.

In the event of a failure—of an interface card or all paths to aninterface card—logical devices must be moved to another interface. Ifan interface card fails, logical devices are reassigned from the brokeninterface to another interface. This reassignment is initiated by theother, functioning interface. If all paths from a host to an interface fail,logical devices accessed on those paths are reassigned to anotherinterface, with which the host can still communicate. Thisreassignment is initiated by PowerPath/VE, which instructs thestorage system to make the reassignment.

The VNX OE and CLARiiON term for these reassignments istrespassing.

Reassignment can take several seconds to complete; however, I/Osdo not fail during it. After devices are reassigned, PowerPath/VEdetects the changes and seamlessly routes data via the new route.

After a reassignment, logical devices can be reassigned (trespassedback, in VNX OE and CLARiiON terminology) to their originallyassigned interface. This occurs automatically if the PowerPath/VEperiodic autorestore feature is enabled. It occurs manually ifrpowermt restore is run; this is the faster approach. Periodicautorestore reassigns logical devices only when restoring paths from


24


a failed state. If paths to the default interface are not marked dead,you must use rpowermt restore.

Active-passive means that only one interface to a device is active at atime, and any others are passive with respect to that device andwaiting to take over if needed.

In an active-passive system, path sets are divided into twoload-balancing groups. The active group contains all paths to theinterface to which the target logical device is assigned; the othergroup contains all paths to the other, nonassigned interface. Only oneload balancing group processes I/O requests at a time, andPowerPath/VE load balances I/O across all paths in the active group.If a path in the active load balancing group fails, PowerPath/VEredirects the I/O request to another path in the active group. If allpaths in the active load balancing group fail, PowerPath/VEreassigns the logical device to the other interface, and then redirectsthe I/O request to a path in the newly activated group.

ALUA ALUA (asymmetric logical unit access) is an array failover modeavailable on VNX OE and CLARiiON systems with FLARE version03.26 or later. In an ALUA storage system, one array controller isdesignated as the active/optimized controller and the other arraycontroller is designated as the active/non-optimized controller. As longas the active/optimized controller is viable, I/O is directed to thiscontroller. Should the active/optimized array controller becomeunavailable or fail, I/O is directed to the active/non-optimized arraycontroller until a trespass occurs.

Note: On an ESX host, if you create a VMFS volume on aPowerPath/VE-managed device using VNX OE and CLARiiON systemswith a FLARE version previous to 4.28.000.6.701 and subsequently removePowerPath/VE, the VMFS volume will not be recognized until the ALUAmode is set to Passive Not Ready (PNR).

Note: When you want to change the Failover mode from PNR to ALUAthrough NAVI on VNX OE and CLARiiON systems with FLARE version03.26 and later, you must reboot the host for ALUA mode to be effective.



PowerPath/VE and NMP coexistenceBoth NMP and PowerPath/VE can be loaded on the same ESX hostand manage storage visible to it. NMP and PowerPath/VE cannotmanage the same storage device. Claim rules are used to assignstorage devices to either NMP or to PowerPath/VE.

When an ESX host boots or performs a rescan, the ESX host discoversall physical paths to the storage devices visible to the host. By default,PowerPath/VE claims all PowerPath-manageable devices. You canchange the devices claimed by changing the claim rules.

“Managing claim rules” on page 66 provides more information.


26



2

This chapter describes how to install PowerPath/VE for VMwarevSphere. The chapter covers the following topics:

◆ Installation components and workflow.......................................... 28◆ Task 1: Installing PowerPath/VE for VMware vSphere .............. 33◆ Task 2: Installing the PowerPath/VE remote CLI (rpowermt) ... 49◆ Task 3: Installing and configuring PowerPath/VE licenses ........ 53

InstallingPowerPath/VE

Installing PowerPath/VE 27

28

Installing PowerPath/VE

Installation components and workflowFigure 3 on page 28 depicts the required and optional components forinstallation of PowerPath/VE for VMware:

Figure 3 PowerPath/VE and VMware installation components



The PowerPath/VE 5.4 and Service Pack Releases for VMware vSphereRelease Notes provides information on supported PowerPath/VE andVMware vSphere installation configurations with VUM and vCLI.

Table 1 on page 29 shows a high-level workflow of the installationprocess for unserved licenses:

Table 1 PowerPath/VE for VMware vSphere installation workflow—unservedlicenses

Task Software Action Notes

1 PowerPath/VE Install on ESX host using vCLIor VUM.

vCLI: You must have vCLI already installed on aremote server. The remote server can be a Linux orWindows host. You use vCLI to install PowerPath/VEremotely on the ESX host.VUM: You must have the VUM plugin, VUM server,vSphere client, and vCenter server already installedon the host where the vSphere Client is installed.

2 PowerPath remoteCLI (rpowermt)

Install on remote rpowermtserver.

The server can be a VM or a physical host.Theserver can be a Windows or a Linux host. It canco-exist with vCLI and the ELM license server.

3 PowerPath/VEunserved licenses

Install and configurePowerPath/VE unservedlicenses on the rpowermtserver.

The license files must be stored with the .lic fileextension.

Installation components and workflow 29

30


Table 2 on page 30 shows a high-level workflow of the installationprocess for served licenses:

Table 2 PowerPath/VE for VMware vSphere Installation workflow—servedlicenses

Task Software Action Notes

1 ELM license server Install and configurethird-party ELM software on aremote server.

You must have already activated a served license fileat the Powerlink Licensing portal. See thePowerPath/VE for VMware vSphere Licensing Guidefor more information on activating served license filesand ELM license server installation procedures.

2 PowerPath/VEserved licenses

Install and configurePowerPath/VE for VMwarevSphere served license fileson ELM license server.

The license files must be stored with the .lic fileextension.

3 PowerPath/VE Install on ESX host using vCLIor VUM

vCLI: You must have vCLI already installed on aremote server. The remote server can be a Linux orWindows host. You use vCLI to install PowerPath/VEremotely on the ESX host.VUM: You must have the VUM plugin, VUM server,vSphere client, and vCenter server already installedon the host where the vSphere Client is installed.

4 PowerPath remoteCLI (rpowermt)

Install on remote rpowermtserver.

The server can be a VM or a physical host. Theserver can be a Windows or a Linux host. It canco-exist with vCLI and the ELM license server.



Figure 4 on page 31 illustrates the PowerPath/VE configuration andinstallation workflow for unserved licenses.

Figure 4 PowerPath/VE configuration and installation workflow (unservedlicenses)

Installation components and workflow 31

32


Figure 5 on page 32 illustrates the PowerPath/VE configuration andinstallation workflow for served licenses.

Figure 5 PowerPath/VE configuration and installation workflow (served licenses)



Task 1: Installing PowerPath/VE for VMware vSphereThis section discusses the installation process for PowerPath/VE forVMware vSphere.

Supported PowerPath/VE installation scenariosYou can install PowerPath/VE for VMware vSphere:

◆ on ESX and ESXi hosts.

◆ on hosts in a VMware HA cluster environment. We recommendusing cluster functions (for example, VMotion) to move activeVMs to a node in the cluster on which PowerPath/VE is not beinginstalled. By installing PowerPath/VE on one host at a time, youensure that other cluster nodes are not impacted by theinstallation.

◆ on hosts in a live VMware DRS cluster environment withoutinterrupting cluster service. “The compliance pane turns greenwhen the remediation process is complete.” on page 44 providesmore information.

◆ in a boot-from-SAN environment. See “Installing PowerPath/VEusing VMware vCLI” on page 37.

Note: PowerPath/VE supports coexistence with NMP boot from SAN.

PowerPath/VE pre-installation tasksComplete the following steps before installing PowerPath/VE forVMware vSphere:

◆ Ensure that the base ESX version operating system is installed onthe host. Check the “Environment and system requirements’section of the PowerPath/VE 5.4 and Service Pack Releases forVMware vSphere Release Notes to verify that the ESX versioninstalled on the host is supported by PowerPath/VE.

◆ If you are installing PowerPath/VE 5.4 and service packs on anESXi host, disable Lockdown mode to ensure that registeredlicenses are persisted upon reboot of ESXi host after completingPowerPath/VE license installation. This is of or pertaining toissue 333654.

Task 1: Installing PowerPath/VE for VMware vSphere 33

34


◆ Ensure that the ESX host has SAN connectivity and that attachedSAN devices are working properly.

◆ Ensure that the ESX host has IP network connectivity.

◆ Ensure that no previous PowerPath/VE installation of thePowerPath/VE release is installed on the ESX host. UsingvSphere CLI, type the following command:

• On Linux, type:

# vihostupdate --query --server <IP address or host name>

• On Windows, go to the installation path C:\ProgramFiles\VMware\VMware vSphere CLI\bin> and type:

# vihostupdate.pl --query --server <IP address or host name>

The output should be a list of packages installed on the host.

If PowerPath/VE is installed, remove it following the instructionsin Chapter 4, “Removing PowerPath/VE.”

◆ The SPC-2 flag information must be enabled for Symmetrixdevices. PowerPath/VE supports Symmetrix devices with theSPC-2 flag enabled only.

Use the esxcfg-scsidevs -l command to determine if theSPC-2 flag is enabled on Symmetrix devices.

When the SPC-2 flag is enabled, a Symmetrix device has thefollowing form:

naa.<WWN_of_the_Symmetrix_device>

When the SPC-2 flag is disabled, a Symmetrix device has thefollowing form:

mpx.vmhba#:C#:T#:L#

and PowerPath/VE configures it as:

symm.<sid>_<dev id>

◆ Ensure that VMotion is properly configured to allow fornon-disruptive installation of PowerPath/VE.

◆ Ensure that the /root partition is not full. If the /root partition isfull PowerPath/VE may be only partially installed, andrpowermt returns PowerPath Not Found error message.



Pre-installation tasks: Installation using vCLIvCLI is VMware’s remote CLI tool. The VMware documentationprovides vCLI package installation procedures. Complete thefollowing steps before installing PowerPath/VE using vCLI:

◆ Complete the tasks in “PowerPath/VE pre-installation tasks” onpage 33

◆ Ensure that the vCLI package is already installed on a separateserver from the ESX host.

◆ Ensure that the vCLI version and ESX version match/arecompatible. VMware documentation, available on the supportsection of the website, www.vmware.com, provides informationon vCLI and ESX version compatibility.

Proceed to “Installing PowerPath/VE using VMware vCLI” onpage 37

Pre-installation tasks: Installation using VMware vCenter Update ManagerVMware vCenter Update Manager (VUM) is an optional module forVMware vCenter that manages and automates patch managementand tracking of VMware ESX hosts and some VMs. “InstallingPowerPath/VE using VUM” on page 39 provides the installation andupdate procedures for supported versions of PowerPath/VE usingVUM. The VMware Update Manager Administration Guide providesinformation on the VUM plugin and installation procedures.

You can use VUM with an online depot format or with an offlinebundle format for storing PowerPath/VE packages. The depot is theproxy URL that the VUM server uses to download ESX hostpatches. The offline bundle format uses zip or VIB files.

The PowerPath/VE 5.4 and Service Pack Releases for VMware vSphereRelease Notes provides information on supported PowerPath/VE andVMware ESX versions with VUM installation.

To install PowerPath/VE using VUM offline bundle format, followthe procedure provided in “Installing PowerPath/VE using VUMoffline bundle format” on page 39.

Complete the following steps before installing PowerPath/VE usingVUM:

◆ Complete the tasks in “PowerPath/VE pre-installation tasks” onpage 33.


36


◆ Ensure that the VUM plugin is already installed on the hostwhere the vSphere Client is installed. The VMware UpdateManager Administration Guide provides information on the VUMplugin and installation procedures.

VUM offline bundle pre-installation tasksComplete the following steps before installing PowerPath/VE usingVUM with the offline bundle format:

1. Download the PowerPath/VE package, PowerPath/VE forVMware vSphere - Install SW Bundle.zip, from Powerlink. Fromthe Support menu, select Software Downloads and locate thesoftware for PowerPath for VMware.

2. Unzip the packageEMCPower.VMWARE.5.4.x.bxxx.zip and save on a local orshared network drive.

3. Ensure that you have Upload File privileges in your vCenterClient and vCenter Server system.

4. Ensure that your vCenter Client and vCenter Server system isconfigured correctly for importing offline bundles as zip files forinstallation. The VMware vCenter Update ManagerAdministration Guide provides information on vCenter Clientand vCenter Server requirements.

Proceed to “Installing PowerPath/VE using VUM offline bundleformat” on page 39.

VUM online depot pre-installation tasksComplete the following steps before installing PowerPath/VE usingVUM with the online depot format:

1. Create the PowerPath/VE depot on a web server:

Note: The web server can be a Linux or Windows host.

a. Download the PowerPath/VE package, PowerPath/VE forVMware vSphere - Install SW Bundle.zip, from Powerlink.From the Support menu, select Software Downloads andlocate the software for PowerPath for VMware.

b. Unzip the packageEMCPower.VMWARE.5.4.x.bxxx.ONLINE.zip and theresulting directory on any web server.

c. Copy the depot folder and its contents to the web server.



d. For Windows Internet Information Services (IIS), ensure thatdirectory browsing is enabled. This is disabled by default onWindows IIS.

e. For Windows IIS, add .vib to the list of known MME typesusing the IIS Manager. This makes the files visible to VUM.

Proceed to “Installing PowerPath/VE using VUM online depotformat” on page 40.

Installing PowerPath/VE using VMware vCLIYou can install PowerPath/VE using the VMware vCLI tool. Theboot-from-SAN installation procedure is the same as a regularinstallation.

Note: Note that PowerPath/VE 5.4 SP1 and PowerPath/VE 5.4 SP2 are a fullpackage install. The following procedure is used for both a fresh install aswell as for an upgrade from PowerPath/VE 5.4 to the latest PowerPath/VEpackage, as applicable. The PowerPath/VE 5.4 and Service Pack Releases forVMware vSphere Release Notes provides information on supportedPowerPath/VE and VMware ESX configurations. Both the root and non-rootuser can install PowerPath/VE.

1. Download the PowerPath/VE package from Powerlink. From theSupport menu, select Software Downloads and locate thesoftware for PowerPath for VMware.

2. Copy the PowerPath/VE packageEMCPower.VMWARE.5.4.x.<bxxx>.zip onto the remote hostfrom which you are running the VMware vCLI tool.

3. Bring the ESX host to maintenance mode.

4. On the remote host running vCLI, type the following to install thePowerPath/VE package:

• On Linux, type:

# vihostupdate --server <IP address or host name> --install --bundle=<full pathto file location>/EMCPower.VMWARE.5.4.x.<bxxx>.zip

• On Windows, go to the installation path C:\ProgramFiles\VMware\VMware vSphere CLI\bin> and type:

C:\Program Files\VMware\VMware vSphere CLI\bin>vihostupdate.pl --server <IPaddress or host name> --install --bundle=<full path to filelocation>\EMCPower.VMWARE.5.4.x.<bxxx>.zip


38


where IP address or Fully Qualified Domain Name identifies the ESXhost onto which you are installing PowerPath/VE, location is thedirectory where PowerPath/VE is installed on the remote host,and bxxx is the PowerPath/VE build number.

5. Type the username and password when prompted.

The installation may take a few minutes.

The following output appears:

The update completed successfully, but the system needsto be rebooted for the changes to be effective.

6. Reboot the ESX host onto which you are installingPowerPath/VE.

For VMware ESX, reboot the ESX host from the service console orfrom the vSphere client.

For VMware ESXi, reboot the ESX host from the vSphere client.

7. Exit maintenance mode for the ESX host with PowerPath/VEnow installed.

8. Complete the following steps to verify that PowerPath/VE isinstalled:

a. Type the following query command:


b. Type the username and password when prompted.

Output such as the following appears:

------Bulletin ID------ ------- ----Installed---- ------Summary------------EMC-PP5.4.1.00.00-bbb <Timestamp> PowerPath 5.4.SP1 for ESX

where <bbb> is the PowerPath/VE build number.

If PowerPath/VE is not correctly installed, see Chapter 5,“Troubleshooting PowerPath/VE Installation,” for instructionson correcting the installation.

Installing PowerPath/VE non-disruptively using VMware vCLIThe migration capability built into the ESX hosts allows members of aDRS cluster to have PowerPath/VE installed without disruptingactive VMs. If the ESX hosts are part of a DRS cluster with VMotion



enabled, placing the ESX host into maintenance mode forces theactive VMs to fail over to other cluster members using VMotion.PowerPath/VE installation is supported with the maintenance mode.This makes the non-disruptive installation of PowerPath/VEpossible.

To install PowerPath/VE 5.4 for VMware ESX in a live VMware DRScluster environment without interrupting cluster service:

1. Bring the ESX host to maintenance mode.

If you bring the ESX host to maintenance mode, the running VMstransfer to another host in the cluster if VMotion is correctlyconfigured.

2. Complete the procedure “Installing PowerPath/VE usingVMware vCLI” on page 37 on the ESX host in maintenance mode.

3. Complete the procedure “PowerPath/VE rpowermt installationprocedure” on page 50.

4. Complete all the procedures in Task “Task 3: Installing andconfiguring PowerPath/VE licenses” on page 53.

5. Exit maintenance mode for the ESX host with PowerPath/VEnow installed.

6. Repeat step 1 through step 5 except step 3 for other nodes in theDRS cluster.

Note that step 3 need not be repeated because the rpowermtpackage, RTOOLS, need only be installed once on the host.

Installing PowerPath/VE using VUMThe PowerPath/VE 5.4 and Service Pack Releases for VMware vSphereRelease Notes provides information on supported PowerPath/VE andVMware ESX configurations for VUM installation. Ensure that yourPowerPath/VE and VMware ESX configuration is compatible withand supported on VUM installation before proceeding with VUMinstallation.

Installing PowerPath/VE using VUM offline bundle formatThis is a multistep process that includes:

◆ Downloading the PowerPath/VE package.


40


◆ Unzipping and saving the PowerPath/VE package as a zip file ona local or shared directory. Note that the files must be in zipformat.

◆ Importing the PowerPath/VE offline bundle.

Before beginning this procedure, ensure that you have completed thesteps in “VUM offline bundle pre-installation tasks” on page 36.

1. In vSphere Client, select a host and go to UpdateManager >Admin View > Configuration > Patch Download Settings.

2. In the Patch Download Sources pane, click Import Patches.

3. On the Select Patches page of the Import Patches wizard, browseto location where you saved the offline .zip PowerPath/VEpackage EMCPower.VMWARE.5.4.x.<bxxx>.zip and select it.

4. Click Next and wait until the file upload completes successfully.

In case of upload failure, check that the structure of the zip file iscorrect or that the VUM network settings are correct.

5. Click Next.

6. On the Confirm Import page of the Import Patches wizard, verifythe PowerPath/VE package that you imported into the VUMrepository, and click Finish.

7. Proceed to step 3 on page 42 of “Installing PowerPath/VE usingVUM online depot format” on page 40 to complete theinstallation.

Installing PowerPath/VE using VUM online depot formatThis is a multistep process that includes:

◆ Downloading the PowerPath/VE package.

◆ Creating the PowerPath/VE baseline, which is the set ofPowerPath updates and patches, in VUM.

◆ Attaching the PowerPath/VE baseline to the ESX hosts ontowhich you’d like to install the PowerPath/VE package.

◆ Staging, which is the process of pushing the PowerPath/VEpackage from the VUM server onto the ESX hosts.



◆ Remediating, which is the process of applying the stagedPowerPath/VE package from the baseline onto the desired ESXhost. Both installing and upgrading packages using VUM, as wellas installing patches and hotfixes, are carried out through theremediation stage.

Before beginning this procedure, ensure that you have completed“VUM online depot pre-installation tasks” on page 36.

To install supported versions of PowerPath/VE using VUM:

1. Download the PowerPath/VE package on VUM:

a. From Update Manager, go to Admin View > Configurationtab > Patch Download Settings > Add Patch Source.

The Add Patch Source window appears.

b. In Source URL field, type the full http path to thePowerPath/VE package index.xml file on the Web server.

c. Select Validate URL.

The text Connected indicates that the patch source URL isvalidated.

d. Click OK.

e. Click Apply and Download Now.

The task bar indicates when the download is complete.

2. Create the PowerPath/VE baseline:

a. From Update Manager, go to Baselines and Groups tab >Extension Baselines tab > Create.

The New Baseline Wizard appears, Baseline Name and Typewindow.

a. In the Baseline Name and Description box, Name field,type a name for the baseline you are creating. For example,PP 5.4.1 package.

b. In the Baseline Type box, the Host Extension radio buttonis selected by default. Do not alter the default.

c. Click Next.

The Extensions window appears.


42


b. In the Baseline Name and Type Extensions box, select thecode version that will be the baseline for the PowerPath/VEpackage and click the down arrow.

c. Click Next.

d. Click Finish.

The PowerPath/VE baseline is created.

3. Attach the PowerPath/VE baseline to the desired ESX hosts. Youcan attach the PowerPath/VE baseline to individually selectedESX hosts or to multiple hosts at a time. To attach thePowerPath/VE baseline:

a. Go to Compliance view.

• To attach the PowerPath/VE baseline to an individual ESXhost, highlight the desired host at the left of the vSphere Clientpane.

• To attach the PowerPath/VE baseline to several ESX hosts,select, for example, a folder, cluster or datacenter at the left ofthe vSphere Client pane.

b. Click Attach.

The Attach Baseline or Group window appears.

c. Under Name, select the PowerPath/VE package baseline thatyou created. It should appear under Extension Baselines.Click Attach.

In the Host Compliance box to the top-right of the screen, thecircle is blue if this is the first time you are attaching the baselineto the ESX host. If you have attached baselines to the ESX host onprevious occasions and remediated them, the circle is green andshows the text 100% Compliant. This indicates that the extensionis already installed.

4. Stage the baseline. Staging is the process of pushing thePowerPath/VE package onto individual ESX hosts from the VUMserver.

a. From the Update Manager tab, look at the Attached Baselinespane in the middle of the screen. Highlight the PowerPath/VEpackage baseline that you created and click Stage.

The Stage Wizard appears, Baseline Selection window.



Under the Name column in the Baselines box thePowerPath/VE package baseline that you created is selectedby default.

b. Do not alter the default Name selection.

Under the Host column, all the ESX hosts to which youattached the PowerPath/VE package baseline are selected bydefault.

c. If desired, alter the default Host selection to stage the baselineto only one or some of the ESX hosts.

d. Click Next.

The Patch and Extension Exclusion window appears.

e. Verify the information and click Next.

The Ready to Complete window appears.

f. Verify the information and click Finish.

The task bar indicates that staging is In Progress. Staging can takeseveral minutes.

In the Host Compliance pane to the upper-right of the screen, thecompliance circle is color-coded to show status. The colors are:

• Blue — unknown

• Red — complete non-compliant. This means that the stage hasbeen successful but it has not yet been remediated.

• Green — compliant. This means that both the stage and theremediation have been successful and are complete.

To see the status of each ESX host, select it in the vSphere Clientpane.

5. Remediate the PowerPath/VE baseline. Both installing andupgrading PowerPath/VE packages using VUM are carried outthrough the remediation stage.

a. Highlight the ESX host you would like to remediate.

b. Look at the Attached Baselines box in the middle of the screen.Highlight the PowerPath/VE package baseline that youcreated and click Remediate.

The Remediate window appears, Remediation Selection box.


44


In the Baseline Groups and Types pane the ExtensionBaselines radio button is selected by default. In the Baselinespane the PowerPath/VE package baseline that you created isselected by default.

c. Do not alter the default Baseline Groups and Types andExtension Baselines default selections.

Under the Host column, all the ESX hosts to which you stagedthe PowerPath/VE package baseline are selected by default.

d. If desired, alter the default Host selection to remediate thebaseline to only one or some of the ESX hosts.

e. Click Next.

The Patches and Extensions box appears.

f. Verify the information and click Next.

The Host Remediation Options box appears.

1. In the Task Name field, type a task name. For example,PowerPath/VE install

2. In the Task Description field, type a description. Forexample, PP/VE 5.4.1 install.

3. Change or maintain remediation time and failure optionsvalues in the Remediation Time and Failure Optionsboxes as needed to suit your environment.

g. Click Next.

The Ready to Complete window appears with yourremediation selections.

h. Verify the information and click Finish.

The task bar shows the progress of each task, including puttingany VMs in any DRS clusters into maintenance mode, removal,installation, automatic reboot, and exit of maintenance mode. Theentire remediation process may take several minutes.

The compliance pane turns green when the remediation processis complete.



Upgrading PowerPath/VETo upgrade to the latest package of PowerPath/VE, follow theprocedure in “Installing PowerPath/VE using VMware vCLI” onpage 37 or “Installing PowerPath/VE using VUM” on page 39.

Because of a VMware vSphere problem, powered-off VMs do notautomatically migrate to a live ESX host in DRS cluster during aPowerPath/VE upgrade using VUM. The VMware PR number forthis issue is 498494.

Note: If you have a valid PowerPath/VE license, licensing information ispreserved when upgrading.

Upgrading PowerPath/VE in a PowerPath/VE-controlled boot-from-SAN setupTo upgrade PowerPath/VE for VMware in anPowerPath/VE-controlled boot-from-SAN setup:

1. Add a low-number claim rule to put all the devices on the samearray class of the boot device under NMP control. Type:

esxcli --server <IP addr of ESX server> corestorage claimrule add --plugin="NMP"--type=vendor --vendor=<vendor name> --model=* --rule=<rule number>

The new claim rule number must not already be in use and mustbe higher than 201 and lower than the lower of the twoPowerPath/VE defaults (which is 250).

The vendor name is the vendor name of the array that from whichthe host is booted.

2. Reboot the ESX host.

For VMware ESX hosts, reboot the ESX host from the serviceconsole or from the vSphere client.

For VMware ESXi hosts, reboot the ESXi host from the vSphereclient.

At this point, all the devices from that array including the bootdevice are under NMP control.

3. Upgrade PowerPath/VE following the procedure in “UpgradingPowerPath/VE” on page 45.

4. Delete the rule you added in step 1 on page 45 . Type:


46


esxcli --server <IP addr of ESX server> corestorage claimrule delete--rule=<rule number>




“Boot-from-SAN configuration management” on page 92provides information on managing PowerPath/VEboot-from-SAN configurations.

Problems with PowerPath/VE upgradeUpgrade of the PowerPath/VE package 5.4 to PowerPath/VE 5.4 SP1using VUM or vCLI, could fail, causing two PowerPath packages toexist simultaneously on the ESX server. Follow the procedure belowto clean the setup and re-install the new PowerPath/VE package.This is of or pertaining to issue 319961. The PowerPath/VE 5.4 andService Pack Releases for VMware vSphere Release Notes provides moreinformation.

The PowerPath/VE 5.4 and Service Pack Releases for VMware vSphereRelease Notes provides information on supported upgrades.

Cleaning up a setup after a failed PowerPath/VE upgradeTo clean up a failed PowerPath/VE upgrade:

1. Remove all PowerPath specific claim rules between 250 and 330by running the command

# esxcli --server <IP addr of ESX server> corestorage claimrule delete--rule=<claim rule #>

2. Reboot the ESX server.

3. Query the packages installed on the ESX server by running thefollowing command:

# vihostupdate --query --server <ESX server hostname/IP address>--username root

4. Remove the PowerPath packages installed by running thefollowing command:

# vihostupdate –server <ESX server hostname/IP address> --usernameroot --remove –bulletin <PowerPath bulletin from Query command>



5. For an ESX setup, login to the service console and confirm that allPowerPath packages have been removed. Type:

# rpm -qa | grep powerpathpowerpath.plugin.esx-5.4.x.00.00-bxxxpowerpath.cim.esx-5.4.x.00.00-bxxx

If PowerPath packages are still present, remove them manuallyby typing the command:

# rpm -e powerpath.cim.esx-5.4.x.00.00-bxxx# rpm -e powerpath.plugin.esx-5.4.x.00.00-bxxx

Note that the CIM package must be removed before thePowerPath plugin package.

6. Reboot the ESX server.

7. Re-install the PowerPath/VE package using vCLI or VUM. Thiswill involve entering maintenance mode and rebooting the ESXserver after PowerPath/VE installation. “InstallingPowerPath/VE using VMware vCLI” on page 37 and “InstallingPowerPath/VE using VUM” on page 39 provide moreinformation.

Upgrade order for PowerPath/VE and VMware ESXTable 3 on page 47 provides supported versions and procedures forupgrade of PowerPath/VE and VMware ESX:

Table 3 PowerPath/VE and ESX upgrade versions and order

PowerPath/VEversion on host

ESX version onhost Step 1 Step2 Step 3 Step 4

5.4 4.0 Upgrade to ESX4.0.1/4.0.2

Upgrade toPowerPath/VE 5.4SP1

Upgrade to 4.1 Upgrade toPowerPath/VE 5.4SP2 (VUM or vCLI)

5.4 4.0 Upgrade to ESX4.1

Upgrade toPowerPath/VE 5.4SP1 (vCLI only)

Upgrade toPowerPath 5.4 SP2(VUM or vCLI)

N/A

5.4 4.0.1/4.0.2 Upgrade to ESX4.1

Upgrade toPowerPath/VE 5.4SP1 (vCLI only)

Upgrade toPowerPath 5.4 SP2(VUM or vCLI)

N/A


48


You do not need to remove the PowerPath/VE online or offlinekernel packages before upgrading VMware ESX versions.“PowerPath/VE rpowermt upgrade procedure” on page 51 providesinformation about removing RTOOLS packages when upgrading.

PowerPath/VE post-installation tasksThere are no post-installation tasks.

5.4 SP1 4.0 Upgrade to ESX4.1


N/A N/A

5.4 SP1 4.0.1/4.0.2 Upgrade to ESX4.1


N/A N/A

5.4 SP1 4.1 Upgrade toPowerPath/VE 5.4SP2

N/A N/A N/A

Table 3 PowerPath/VE and ESX upgrade versions and order

PowerPath/VEversion on host

ESX version onhost Step 1 Step2 Step 3 Step 4



Task 2: Installing the PowerPath/VE remote CLI (rpowermt)This section discusses installation of the PowerPath/VE rpowermt(or RTOOLS) package.

PowerPath/VE rpowermt installation requirementsThe rpowermt server can be any one of the following:

◆ The vCLI remote server◆ A Virtual Machine◆ The ELM license server

The PowerPath/VE 5.4 and Service Pack Releases for VMware vSphereRelease Notes provides information on supported operating systemsfor the rpowermt server.

TCP port for ESX host and rpowermt serverThe TCP port between the ESX host and rpowermt server is fixed tonumber 5989. This is a non-dynamic number.

For more information on VMware ports, go to VMware KnowledgeBase (http://kb.vmware.com/kb), and search for article 1012382.

PowerPath/VE rpowermt pre-installation tasksBefore you install the rpowermt package, RTOOLS:

❑ On Windows, ensure that you are have administrative privilegesbefore installing RTOOLS. PowerPath for Windows restrictsaccess to the PowerPath folder for users without administrativeprivileges.

❑ On Linux, ensure that you have root privileges before installingRTOOLS. You cannot install the RTOOLS package without rootprivileges.

The PowerPath/VE 5.4 and Service Pack Releases for VMware vSphereRelease Notes provides information on required user privileges forRTOOLS installation.

Task 2: Installing the PowerPath/VE remote CLI (rpowermt) 49

http://kb.vmware.com/kb

50


PowerPath/VE rpowermt installation procedureThis section provides procedures for installing rpowermt, which isincluded in the PowerPath/VE RTOOLS package.

Installing PowerPath/VE rpowermt on LinuxTo install PowerPath/VE rpowermt on a Linux host:

1. Download the RTOOLS .rpm package from Powerlink. From theSupport menu, select Software Downloads and locate thesoftware for PowerPath/VE for VMware.

2. Run the -i option to install the PowerPath/VE remote CLI. Type:

-ivh EMCPower.RTOOLS-<version>-<bxxx>.RHEL5.x86_64.rpm

where <bxxx> is the build number. For example,

-ivh EMCPower.RTOOLS-5.4.1.00.00-268.RHEL5.x86_64.rpm


Preparing... ########################################### [100%]1:EMCPower.RTOOLS ########################################### [100%]

Installing PowerPath/VE rpowermt on WindowsTo install PowerPath/VE rpowermt on a Windows host:

1. Download the Windows RTOOLS package .zip file fromPowerlink. From the Support menu, select Software Downloadsand locate the software for PowerPath/VE for VMware.

2. Unzip the Windows package file. It contains an .exe file named

EMCPower.RTOOLS.<version>.Windows.x86_32_bxxx.exe

where <version> is the PowerPath/VE version, and bxxx is thebuild number. For example,

EMCPower.RTOOLS.5.4.SP1.Windows.x86_32.b033.zip

3. Double-click on the .exe file.

An Install Shield wizard, which is used for the PowerPath/VEremote CLI installation, appears.

4. Various screens appear during the Install Shield wizardinstallation. For all fields, click Next for the default values.



5. The installation location screen appears. The Install Shield wizardlists a default folder.

• If you want the RTOOLS package installed in the defaultlocation, click Next.

• If you want to install the RTOOLS package in a location otherthan the default location:

a. Click Change located at the right.

b. Browse to the desired installation location.

c. Click Next.

6. Click Install.

7. The PowerPath/VE remote CLI is installed at the location youspecified.

8. Log out and log back in to the remote session to include therpowermt directory in the default path.

For Windows the default path is:c:\Program Files\EMC\PowerPath\rpowermt

PowerPath/VE rpowermt upgrade procedureThis section provides procedures for upgrading rpowermt (RTOOLSpackage).

Upgrading PowerPath/VE 5.4 and service packs rpowermt on LinuxThis section provides procedures for upgrading the PowerPath/VE5.4 and 5.4 SP1 Linux RTOOLS package to PowerPath/VE 5.4 SP2Linux RTOOLS package.

To upgrade PowerPath/VE rpowermt on a Linux host:

1. Run the rpowermt version command to verify that thePowerPath/VE 5.4 SP1 RTOOLS package is installed on the Linuxhost.


EMC rpowermt for PowerPath (c) client Version 5.4 SP 1 (build 33)

2. Use the -Uvh option to upgrade to the PowerPath/VE 5.4 SP2RTOOLS package. For example:

# rpm -Uvh EMCPower.RTOOLS-5.4.2.00.00-299.RHEL5.x86_64.rpm

Task 2: Installing the PowerPath/VE remote CLI (rpowermt) 51

52



Preparing... ########################################### [100%]--- Installing EMCPower.RTOOLS ---

1:EMCPower.RTOOLS ########################################### [100%]

Upgrading PowerPath/VE 5.4 SP1 rpowermt on WindowsDuring the upgrade of the PowerPath/VE 5.4 SP1 RTOOLS packageto the PowerPath/VE 5.4 SP2 RTOOLS package, the following errormessage appears:

A newer version of this application is already installed.If you wish to install this version, please uninstall thenewer version first. Click OK to exit the wizard.

The “newer version” referenced in the error message is in fact theprevious version of the RTOOLS package; that is, the PowerPath/VE5.4 SP1 RTOOLS package that already existed on the host.

If you receive this error:

1. Click OK to exit the wizard.

2. Uninstall the PowerPath/VE 5.4 SP1 RTOOLS package.

3. Install PowerPath/VE 5.4 SP2 RTOOLS package.

“Task 2: Installing the PowerPath/VE remote CLI (rpowermt)” onpage 49 provides RTOOLS package installation procedures.

PowerPath/VE rpowermt post-installation tasksThere are no post-installation tasks.



Task 3: Installing and configuring PowerPath/VE licensesThis section describes installing and configuring PowerPath/VElicenses for VMware.

PowerPath/VE licenses prerequisites

Before installing and configuring unserved licenses:Ensure that you have completed the following tasks before installingand configuring unserved licenses:

❑ Generated a license file for each ESX host that will havePowerPath/VE installed on it at the Powerlink Licensing portal.PowerPath/VE for VMware vSphere licenses are available onlythrough Powerlink Licensing.

❑ Stored the license file with .lic extension on the rpowermt server.

Note that PowerPath/VE does not support storing served licenseconfiguration files and unserved licenses in the same folder on therpowermt server. You must store served license configuration filesand unserved license files in separate folders on the rpowermt server.This is of or pertaining to issue 334577.

See PowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, for instructions on carrying out these tasks.

Before installing and configuring served licenses:Ensure that you have completed the following tasks before installingand configuring served licenses:

❑ Generated and activated a license (.lic) file for the ELM licenseserver at the Powerlink Licensing portal.

❑ Downloaded and installed the third-party product used forPowerPath/VE for VMware vSphere ELM (see “Third-partyproducts used with PowerPath/VE for VMware vSphere ELM”on page 54).

Note:Only one instance of the EMC vendor daemon is allowed on a singleELM server. Therefore, all EMC products must use the same port and alllicense files use this same port. This may require modification of the TCPport number in the PowerPath/VE license file that resides on the ELMlicense server and the PowerPath/VE license configuration file that

Task 3: Installing and configuring PowerPath/VE licenses 53

54


resides on the rpowermt server. See “TCP port conflicts” on page 102 inChapter 5, “Troubleshooting PowerPath/VE Installation,” for moreinformation.

All license server components must be version 11.6 and above.

❑ Stored the license file with .lic extension in the ELM licenseserver directory.

See the PowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, for instructions on carrying out these tasks.

Third-party products used with PowerPath/VE for VMware vSphere ELMThe following third-party products are used in the PowerPath forVMware ELM environment and are available on Powerlink:

◆ Flexera (formerly Acresso) FLEXnet Publisher Licensing Toolkit(license server) — Manages electronic PowerPath license files.The Flexera software and accompanying documentation can bedownloaded from Powerlink:Support > Software Downloads and Licensing > Downloads P-R> PowerPath for VMware

See the PowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, for more information.

Installing and configuring PowerPath/VE electronic licensesThis section discusses installation and configuration of thePowerPath/VE electronic licenses. For served licenses, theinstallation and configuration procedures include steps forcompatibility with FLEXnet Publisher Licensing Toolkit.

Determining license locationTo determine the location for served and unserved licenses:

1. Decide whether you want the .lic file to be saved to the defaultdirectory or if you want to change the default license directory.

• On Linux, there are four default directories:

– /etc/emc– /etc/emc/licenses– /opt/EMCpower– /opt/EMCpower/licenses



• On Windows, the default directory isC:\Documents and Settings\<username>\MyDocuments\EMC\PowerPath\rpowermt

2. If you want:

• to save the .lic file to the default directory, you need not takeany further action.

• to a change the default license directory, set thePPMT_LIC_PATH environment variable to point to the desireddirectory. To set the environment variable PPMT_LIC_PATH:

– On Linux, type one of the following:# setenv PPMT_LIC_PATH <license file directory>

or

# export PPMT_LIC_PATH=<license file directory>

– On Windows, type the following:C:\> set PPMT_LIC_PATH=<license file directory>

• to make the non-default path persistent across host reboots:

– On Linux, add the environment variable to the shellconfiguration file (for example, the .profile file);

– On Windows, add the environment variable to theEnvironment Variables table. FromMy Computer, select System Properties >EnvironmentVariables.

Installing and configuring unserved licensesTo install and configure unserved licenses on the rpowermt server:

1. Install the license (.lic) file obtained from the Powerlink on“Before installing and configuring unserved licenses:” on page 53on the rpowermt server.

• Copy the .lic file to one of the default directories:

– /etc/emc– /etc/emc/licenses– /opt/EMCpower– /opt/EMCpower/licenses

or


56


• Copy the .lic file to the location that you created in“Determining license location” on page 54.

Note: You can store unserved license files on multiple rpowermtservers because the license is not locked to the rpowermt server.

2. Run rpowermt host=<ip address of ESX host>

check_registration to check that PowerPath is not alreadylicensed. “rpowermt check_registration: Check PowerPath/VElicense information” on page 120 describes this command.

Upon entering the rpowermt command, you are prompted tocreate an rpowermt lockbox. The rpowermt lockbox is anRSA-enabled encrypted file used to store and protect sensitiveinformation. rpowermt uses the lockbox to securely store remotehost IP addresses and their username and passwordcombinations. By storing this information, rpowermt does notrepeatedly prompt you for the username and password for theremote host. For information on setting the rpowermt lockbox,see “Setting the rpowermt lockbox” on page 89.

Note that the PowerPath/VE 5.4 SP1 version of the rpowermtlockbox is not backward compatible with the PowerPath 5.4version of the rpowermt lockbox. Once PowerPath/VE 5.4 SP1has created or converted the lockbox, the PowerPath/VE 5.4version of rpowermt will not be able to read the PowerPath 5.4SP1 version of the rpowermt lockbox. If you use incompatibleversions of the rpowermt lockbox, an error message appears.“Incompatible rpowermt lockbox versions” on page 104 providesmore information. The PowerPath/VE 5.4 and Service Pack Releasesfor VMware vSphere Release Notes provides information onPowerPath/VE rpowermt lockbox version compatibility.

3. Run rpowermt host=<ip address of ESX host> register toregister the license for the ESX host. “rpowermt register: Registera PowerPath/VE license on an ESX host” on page 141 describesthis command.


check_registration to ensure that the license is recognized.“rpowermt check_registration: Check PowerPath/VE licenseinformation” on page 120 describes this command.



Installing and configuring served licensesTo install and configure served licenses:

1. On the rpowermt server, create a license configuration file thatidentifies the license server to the rpowermt application:

a. Copy and paste the first three lines of the license file receivedfrom Powerlink Licensing portal into the license configurationfile:

SERVER [host] INTERNET=[IP address] [port]VENDOR EMCLM [vendor daemon path] [port]USE_SERVER

where

• host is the name of the license server to be used

• IP address is of the license server is preceded by the keywordINTERNET=

and

• port is the TCP port number (by default, 27010).

These should match the host, IP address, and port in thelicense file downloaded from the Powerlink Licensingportal. For example, the license configuration file would lookas follows:

SERVER lsca2147 INTERNET=172.23.168.142 27010VENDOR EMCLM /sbin/EMCLM.exe PORT=27011USE_SERVER

Note that if you modified host and port in the license file, you mustmodify them in the license configuration file to match exactly. Theseare user-configurable fields in the license configuration file andmodifying them does not affect the signature of the licenseconfiguration file.


58


The license file downloaded from the Powerlink Licensingportal would look as follows:

SERVER lsca2147 INTERNET=172.23.168.142 27010VENDOR EMCLM /sbin/EMCLM.exe PORT=27011USE_SERVERINCREMENT PowerPathMP EMCLM 5.4 permanent 1 OVERDRAFT=1 ISSUER=EMC \

ISSUED=29-May-2009 NOTICE="ACTIVATED BY 8918825 ABC Corp \15260 VENTURA BLVD SHERMAN OAKS CA US \91403" SIGN="0034 DAX7 4X13 B1XE CX23 84X3 4X00 BXED 9028 \C2EE 9AE4 88E8 CC6A"

b. If you modified the TCP port number in the license file,modify the port number in the license configuration file. Thisis a user-configurable field in the license configuration file andmodifying it does not affect the signature of the licenseconfiguration file. See “TCP port conflicts” on page 102 formore information.

2. Install the license configuration (.lic) file on the rpowermt server:Either:

• Copy the .lic file to one of the default directories in step 1 of“Determining license location” on page 54.

Or

• Copy the .lic file to the location that you created in step 2“Determining license location” on page 54.


check_registration to check that PowerPath is not alreadylicensed. “rpowermt check_registration: Check PowerPath/VElicense information” on page 120 describes this command.

Upon entering the rpowermt command, you are prompted tocreate an rpowermt lockbox. The rpowermt lockbox is anRSA-developed encrypted file used to store and protect sensitiveinformation. rpowermt uses the lockbox to securely store remotehost IP addresses and their username and passwordcombinations. By storing this information, rpowermt does notrepeatedly prompt you for the username and password for theremote host. For information on setting the rpowermt lockbox,see “Setting the rpowermt lockbox” on page 89.



4. Run rpowermt host=<ip address of ESX host> register toregister the license for the ESX host. “rpowermt register: Registera PowerPath/VE license on an ESX host” on page 141 describesthis command.


check_registration to ensure that the license is recognized.“rpowermt check_registration: Check PowerPath/VE licenseinformation” on page 120 describes this command.

PowerPath/VE licenses post-installation and configuration tasksComplete the following post-installation tasks.

Optional configuration tasksComplete the following optional configuration tasks, as desired.

Note: Both the root and non-root user can run rpowermt commands onWindows and Linux rpowermt servers.

◆ If you want to add ESX hosts to the rpowermt lockbox that wasautomatically created upon entering the first rpowermt commandin (see step 2 on page 56 of “Installing and configuring unservedlicenses,” and step 3 on page 58 of “Installing and configuringserved licenses,”), type:

rpowermt setup add_host [host_file=<filename>] host=<hostname>username=<username> [password=<password>]

Note: EMC recommends being cautious when leaving an rpowermt serverunattended after the rpowermt lockbox is enabled. Once you have enteredthe username and password to the lockbox and rpowermt is authorized,anyone can run rpowermt commands to any ESX host managed by anrpowermt server without any username and password authentication. Thelockbox is protected by native Operating System Access Controls. That is,any lockbox file access that you give to other users through changingWindows or Linux file permissions will in turn allow full access to the ESXhostnames and root passwords contained in the lockbox.

◆ The default load-balancing and failover policy is SymmetrixOptimized for Symmetrix devices, CLARiiON Optimized forCLARiiON and VNX OE devices, and Adaptive for Invista and


60


VPLEX, supported Celerra devices, and supported third-partyarrays after registering the PowerPath/VE for VMware vSpherelicense. If desired, you can change the policies. Type:

rpowermt host=<IP address of the ESX server> set policy=<policy>

“rpowermt set policy: Set the load-balancing and failover policy”on page 151 provides more information.

Note that:

• VNX devices are managed under the clariion storage classand appear as CLARiiON devices in rpowermt output.

• VNXe devices are managed under the generic storage classand appear as Celerra devices in rpowermt output.

• VPLEX devices are managed under the invista device classand appear as Invista devices in rpowermt output.

◆ If desired, enable path latency monitoring. Type:

rpowermt host=<IP address of the ESX server> set path_latency_monitor=onrpowermt host=<IP address of the ESX server> set_path_latency_threshold=<seconds>

“rpowermt set path_latency_threshold: Sets a time limit for thecompletion of I/O” on page 148 and “rpowermt setpath_latency_monitor: Enables or disables path latencymonitoring” on page 147 provide more information.

Note: PowerPath does not configure array control devices such as VNX OEand CLARiiON and Invista LUNZs, Symmetrix VCMDB/ACLX devices, andthird-party array control devices. These devices are still under NMP control.



Verify PowerPath/VE configurationComplete the following steps to verify that PowerPath/VE isproperly configured.

1. Select a storage device and examine its configuration. From therpowermt server, type:

rpowermt host=<IP address of the ESX server> display dev=0


Pseudo name=emcpower0CLARiiON ID=FCNTR073300020Standard UID=naa.60060160e1601e006df56548a2fbdc11state=alive; policy=co; queued-IOs=0Owner: default=SP B, current=SP B Array failover mode: 1===========================================================================-------------- Host --------------- - Stor - --- I/O Path -- -- Stats ---### HW Path I/O Paths Interf. Mode State Q-IOs Errors===========================================================================1 vmhba2 C0:T1:L10 SP B7 active alive 0 02 vmhba1 C0:T1:L10 SP B6 active alive 0 02 vmhba1 C0:T0:L10 SP A6 active alive 0 11 vmhba2 C0:T0:L10 SP A7 active alive 0 1

Verify that the device has been assigned a PowerPath/VE pseudodevice name, the device state, and the device policy. ForCLARiiON and VNX OE devices, verify the default and currentowner, and the array failover mode.

Note that the policy is set to co. This indicates that the ESX host islicensed. On a licensed ESX host, the default policy is SymmetrixOptimized for Symmetrix devices, CLARiiON Optimized forCLARiiON and VNX OE devices, and Adaptive for Invista,VPLEX, supported Celerra devices, and supported third-partydevices. On an unlicensed ESX host, Basic Failover is the defaultinterim policy that protects against CLARiiON and VNX OE SPfailures, Symmetrix FA port failures, and back-end failures untilthe ESX host is licensed. Between the time that PowerPath/VE isinstalled on an ESX host (and that host is rebooted) and the timethat the rpowermt register command is run, the policy forSymmetrix, CLARiiON, VNX OE, Invista, and VPLEX devices isbf (Basic Failover), and for supported Celerra and supportedthird-party devices it is nr (No Redirect). This is the only timethat a device is set to Basic Failover or No Redirect. Once the ESXhost is licensed, the policy transitions to Symmetrix Optimized,CLARiiON Optimized, or Adaptive, as appropriate. I/O is notsupported when PowerPath/VE is unlicensed.


62


Note that the I/O Path Mode is active, indicating that the deviceshave been claimed by PowerPath/VE and are licensed. If the I/OPath Mode is unlic (unlicensed), that indicates that the devices havebeen claimed by PowerPath/VE but are not yet licensed.

2. If the output of step 1 on page 61 does not show a correctPowerPath configuration, verify that the storage devices havebeen claimed by PowerPath/VE. Type:

vicfg-scsidevs -l

The output for one device is shown below. The output indicatesthat a CLARiiON RAID 5 device has been claimed by thePowerPath/VE.

naa.60060160e1601e006df56548a2fbdc11Device Type: Direct-AccessSize: 10240 MBDisplay Name: RAID 5 DGC 0324Plugin: PowerPathConsole Device: /dev/sdbDevfs Path: /vmfs/devices/disks/naa.60060160e1601e006df56548a2fbdc11Vendor: DGC Model: RAID 5 Revis: 0324SCSI Level: 4 Is Pseudo: false Status: onIs RDM Capable: true Is Removable: falseOther Names:

vml.020001000060060160e1601e006df56548a2fbdc11524149442035

3. If the output of step 1 on page 61 and step 2 on page 62 do notshow a correct PowerPath configuration, or if you want to changethe devices that PowerPath is managing, see “Managing claimrules” on page 66.

4. If you are installing PowerPath/VE 5.4 and service packs on anESXi host, after completing all operations related to licensing andpolicy changes, re-enable Lockdown mode after having disabledit as indicated in “PowerPath/VE pre-installation tasks” onpage 33.

• On PowerPath/VE 5.4 and 5.4 SP1 on an ESXi 4.0 host, ensurethat Lockdown mode is disabled for any managementoperations related to licensing and other persistent changes.

• On PowerPath/VE 5.4 SP2 on ESXi 4.1 host, ensure thatLockdown mode is disabled before running any rpowermtcommand.

This is of or pertaining to issue 333654. The PowerPath/VE 5.4 andService Pack Releases for VMware vSphere Release Notes providesinformation on issue 333654.



Removing licenses To remove licenses:

1. Run rpowermt unregister host=<ESX host IP address> tounregister the license for the ESX host. See “rpowermt unregister:Remove a PowerPath/VE license from an ESX host” on page 160.

2. Remove the .lic file from the rpowermt server.

3. Reboot the ESX host, if desired. This is an optional step.

Note that if you do not reboot the ESX host after unregistering aPowerPath/VE license, PowerPath multipathing functionalitycontinues to be provided to any storage device previouslyclaimed by PowerPath/VE. Any new devices added to the ESXhost and then claimed by PowerPath, however, will be in theunlicensed state.

Note that for served licenses, occasionally if the ELM licenseserver cannot be contacted during the unregister process, theunregister succeeds although the license available count does notget incremented. This is a known issue (see the PowerPath/VE 5.4and Service Pack Releases for VMware vSphere Release Notes for moreinformation).

Note: See “Behavior in different licensed states” on page 18 forinformation on unlicensed ESX host and I/O support in differentlicensed states.

Re-installing and configuring after converting license typesIf you work with your EMC Customer Support Representative toconvert your license type, for example, from unserved license toserved license, or vice-versa; or to convert your ELM license serverconfiguration, the original license must be removed from the serveron which it resided and the new license must be installed andconfigured on the server on which it resides. Complete the followingprocedure:

1. For the original license that you are converting from, completethe procedure for “Removing licenses” on page 63.

2. For the new license that you are converting to, complete theprocedure:


64


• “Installing and configuring served licenses” on page 57 if youare converting to served licenses

or

• “Installing and configuring unserved licenses” on page 55 ifyou are converting to unserved licenses.

The PowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, provides for more information on licensing.

Rehosting Transferring a PowerPath/VE license from one ESX host to another iscalled rehosting. If you need to rehost, see the PowerPath/VE forVMware vSphere Licensing Guide, available on Powerlink, for theprocedure.


3

This chapter contains information and procedures on managingPowerPath/VE. Topics include:

◆ Managing claim rules ........................................................................ 66◆ Adding new paths to a PowerPath/VE logical device................. 82◆ Adding new logical devices to a configuration............................. 83◆ Removing paths or logical devices from a configuration ............ 85◆ Managing PowerPath/VE using VMware commands ................ 87◆ Managing PowerPath/VE using rpowermt................................... 88

ManagingPowerPath/VE

Managing PowerPath/VE 65

66

Managing PowerPath/VE

Managing claim rulesDuring PowerPath/VE installation, PowerPath/VE claims allSymmetrix, CLARiiON, VNX OE, and supported third-party arraydevices by default. If you do not want all these devices underPowerPath/VE control, but rather want some Symmetrix,CLARiiON, VNX OE, or third-party array devices under NMPcontrol, you must change the claim rules.

Note: EMC recommends becoming very familiar with the VMware vStorageAPIs for Multipathing framework and careful consideration before changingclaim rules. The VMware vSphere Configuration Guide provides information onclaim rules.

Table 4 lists the VMware esxcli corestorage commands used to defineand manage claim rules.

Refer to the VMware Infrastructure Command Line Interface Installationand Reference Guide for detailed information.

Table 4 esxcli corestorage commands

Command Description

esxcli corestorage claiming reclaim Attempts to unclaim all paths to a device and to then run the loadedclaim rules on each of the unclaimed paths to reclaim those paths.

esxcli corestorage claiming unclaim Unclaims a path or set of paths, and disassociates those pathsfrom an MPP plugin. You can only unclaim active paths with no I/O.

esxcli corestorage claimrule add Adds a claim rule to the set of claim rules on the host.

esxcli corestorage claimrule delete Deletes a claim rule from the set of claim rules on the host.

esxcli corestorage claimrule list Lists all claim rules on the host.

esxcli corestorage claimrule load Loads claim rules from the esx.conf configuration file into theVMkernel.

esxcli corestorage claimrule move Moves a claim rule from one rule ID to another.

esxcli corestorage claimrule run Executes path claiming rules.



The PowerPath/VE installation program defines claim rules thatassign all CLARiiON, VNX OE, and Symmetrix devices toPowerPath/VE. The following table lists the claim rules:

Note: CLARiiON devices have DGC as vendor and the model is all as denotedby the asterisk (*). Symmetrix, Invista, VPLEX, and supported Celerradevices have EMC as vendor and the model is Symmetrix, Invista, and Celerra,respectively.

Note the following considerations for the VNX, VNXe, and VPLEXdevices:

◆ VNX devices:

• are managed under the clariion storage class.• appear as CLARiiON devices in rpowermt output.• use claim rule 250.

◆ VNXe devices:

• are managed under the generic storage class.• appear as Celerra devices in rpowermt output.• use claim rule 330.

◆ VPLEX devices:

Storage system claim rules

Storage system Claim rule

CLARiiON 250 file vendor PowerPath vendor=DGC model=*

Symmetrix 260 file vendor PowerPath vendor=EMC model=SYMMETRIX

Invista 270 file vendor PowerPath vendor=EMC model=Invista

Hitachi HDS 280 file vendor PowerPath vendor=HITACHI model=OPEN

HPxP 290 file vendor PowerPath vendor=HP model=OPEN

Compaq EVA 5000 300 file vendor PowerPath vendor=COMPAQ model=HSV111

HP EVA 5000 310 file vendor PowerPath vendor=HP model=HSV111

HP EVA 8000 320 file vendor PowerPath vendor=HP model=HSV210

Celerra 330 file vendor PowerPath vendor=EMC model=Celerra

Hitachi 340 file vendor PowerPath vendor=HITACHI model=DF600F

IBM 350 file vendor PowerPath vendor=IBM model=2107900

Managing claim rules 67

68


• are managed under the invista device class.• appear as Invista devices in rpowermt output.• use claim rule 270.

If you want PowerPath/VE to manage all supported array devices,then you do not need to define any further claim rules.

You do need to define claim rules if you want NMP to manage someof the CLARiiON, VNX, Symmetrix, supported Celerra devices,Invista, or supported third-party array devices.

Listing claim rulesTo list the current set of claim rules, run the following command atthe SSH or console prompt:

esxcli --server <IP addr of ESX server> corestorage claimrule list

Below is the claim rule list as it exists after PowerPath/VEinstallation:

Rule Class Type Plugin Matches0 runtime transport NMP transport=usb1 runtime transport NMP transport=sata2 runtime transport NMP transport=ide3 runtime transport NMP transport=block4 runtime transport NMP transport=unknown101 runtime vendor MASK_PATH vendor=DELL model=Universal Xport250 runtime vendor PowerPath vendor=DGC model=*260 runtime vendor PowerPath vendor=EMC model=SYMMETRIX270 runtime vendor PowerPath vendor=EMC model=Invista280 runtime vendor PowerPath vendor=HITACHI model=OPEN290 runtime vendor PowerPath vendor=HP model=OPEN300 runtime vendor PowerPath vendor=COMPAQ model=HSV111310 runtime vendor PowerPath vendor=HP model=HSV111320 runtime vendor PowerPath vendor=HP model=HSV210330 runtime vendor PowerPath vendor=EMC model=Celerra340 runtime vendor PowerPath vendor=HITACHI model=DF600F350 runtime vendor PowerPath vendor=IBM model=210790065535 runtime vendor NMP vendor=* model=*

The column headings have the following meanings:

◆ Rule — The claim rule number that corresponds to the devicebeing managed.

◆ Class — The status of the claim rule; file means that the claim rulehas been added persistently; runtime means that the claim rulehas been fully loaded into ESX memory.



◆ Type — The type of claim or unclaim operation to perform. In theclaim rules above the types are transport and vendor. Other validvalues are location, path, driver, device, and plugin.

◆ Plugin — The plugin that is managing the device; in the claimrules shown above the plugins are NMP and PowerPath.

◆ Matches — The criteria that are selected to be applied the devicesby the claim rule.

“Storage system claim rules” on page 67 lists the PowerPath/VEclaim rules and the corresponding devices that these claim rulesmanage. All of the claim rules listed in the table are added byPowerPath/VE installation.

As mentioned in “Managing claim rules” on page 66, duringPowerPath/VE installation PowerPath/VE claims all Symmetrix,CLARiiON, VNX OE, Invista, VPLEX, supported Celerra devices,and supported third-party array devices by default. In the case ofthird-party array devices, the claim rules claim all third-party vendormodels, including some models that may not be supported byPowerPath/VE 5.4 for VMware vSphere. This is the case with claimrules 270 and 280. “Preventing PowerPath/VE from claimingdevices” on page 79 provides the procedure to manipulate the claimrules at HBA/device/target level so that the unsupported third-partyarray models are not claimed by PowerPath/VE.

Adding claim rulesClaim rules are numbered 0–65535. For NMP to claim a Symmetrix,CLARiiON, VNX OE, or third-party array device, you must add anew claim rule. The new claim rule number must not already be inuse and must be higher than 201 and lower than the PowerPath/VEdefaults (which are 250 and 260). The number of rules you must addfor each device depends on the number of HBAs in the ESX host andthe array type.

Procedure for adding claim rulesTo add claim rules:

1. Identify the device you want to for which you want to add theclaim rule by running the following command:

esxcfg-mpath --server <IP addr of ESX server> -L

2. Add the claim rule for the device on each adapter/targetcombination:


70


esxcli --server <IP addr of ESX server> corestorage claimrule add--type="location" --rule=<number> --plugin="NMP" --adapter=vmhba<#>--channel=<channel #> --target=<target #>--lun=<lun #>

3. Verify that the claim rule has been added:


The claim rules you have added should appear among the list ofclaim rules.

4. Load the claim rule:

esxcli --server <IP addr of ESX server> corestorage claimrule load

Loading the claim rule will ensure that the runtime class is createdwhen you run the claimrule list.

5. Verify that the claim rule has been loaded:


The claim rules you have added should appear among the list ofclaim rules twice: once on one line as it does after Step 3 onpage 70, indicating the claim rule as a file in the class column, andthen directly underneath the first appearance indicating the claimrule as runtime in the class column. This indicates that the claimrules have loaded correctly.

6. Unclaim the device from PowerPath/VE control using the devicenumber:

a. Identify the device number associated with the device youwant to place under NMP control:


b. Unclaim the device by the device number:

esxcli --server <IP addr of ESX server> corestorage claiming unclaim --type=device--device=<device number>

7. Run the claim rule:

esxcli --server <IP addr of ESX server> corestorage claimrule run

8. Verify that the device is under NMP control:




Example 1This example describes how to claim LUN 57 for NMP on an ESXhost with two HBAs to a DMX array with two storage ports (similarto a CX array running ALUA mode) using claim rule numbers 202and 203:

1. Identify the device you want to place under NMP control:


Output like the following appears:

vmhba1:C0:T0:L57 state:active naa.6006016029a11e0048d2fa3437a4dd11 vmhba1 0 0 57PowerPath active san fc.2000001b3210fe3d:2100001b3210fe3dfc.50060160c1e0a2c5:5006016841e0a2c5

vmhba3:C0:T0:L57 state:active naa.6006016029a11e0048d2fa3437a4dd11 vmhba3 0 0 57PowerPath active san fc.2000001b3210553a:2100001b3210553afc.50060160c1e0a2c5:5006016041e0a2c5

2. Add the claim rule to claim the device for NMP:

esxcli --server <IP addr of ESX server> corestorage claimrule add--type="location" --rule=202 --plugin="NMP" --adapter=vmhba1 --channel=0--target=0 --lun=57


3. List the claim rules to verify that the claim rules have been added:



72



Rule Class Type Plugin Matches0 runtime transport NMP transport=usb1 runtime transport NMP transport=sata2 runtime transport NMP transport=ide3 runtime transport NMP transport=block4 runtime transport NMP transport=unknown101 runtime vendor MASK_PATH vendor=DELL model=Universal Xport202 file location NMP adapter=vmhba1 channel=0 target=0 lun=57203 file location NMP adapter=vmhba3 channel=0 target=0 lun=57250 runtime vendor PowerPath vendor=DGC model=*250 file vendor PowerPath vendor=DGC model=*260 runtime vendor PowerPath vendor=EMC model=SYMMETRIX260 file vendor PowerPath vendor=EMC model=SYMMETRIX270 runtime vendor PowerPath vendor=EMC model=Invista270 file vendor PowerPath vendor=EMC model=Invista280 runtime vendor PowerPath vendor=HITACHI model=OPEN280 file vendor PowerPath vendor=HITACHI model=OPEN290 runtime vendor PowerPath vendor=HP model=OPEN290 file vendor PowerPath vendor=HP model=OPEN300 runtime vendor PowerPath vendor=COMPAQ model=HSV111300 file vendor PowerPath vendor=COMPAQ model=HSV111310 runtime vendor PowerPath vendor=HP model=HSV111310 file vendor PowerPath vendor=HP model=HSV111320 runtime vendor PowerPath vendor=HP model=HSV210320 file vendor PowerPath vendor=HP model=HSV210330 runtime vendor PowerPath vendor=EMC model=Celerra330 file vendor PowerPath vendor=EMC model=Celerra340 runtime vendor PowerPath vendor=HITACHI model=DF600F340 file vendor PowerPath vendor=HITACHI model=DF600F350 runtime vendor PowerPath vendor=IBM model=2107900350 file vendor PowerPath vendor=IBM model=210790065535 runtime vendor NMP vendor=* model=*



5. List the claim rules to verify that the claim rules have beenloaded:





Rule Class Type Plugin Matches0 runtime transport NMP transport=usb1 runtime transport NMP transport=sata2 runtime transport NMP transport=ide3 runtime transport NMP transport=block4 runtime transport NMP transport=unknown101 runtime vendor MASK_PATH vendor=DELL model=Universal Xport202 runtime location NMP adapter=vmhba1 channel=0 target=0 lun=57202 file location NMP adapter=vmhba1 channel=0 target=0 lun=57203 runtime location NMP adapter=vmhba3 channel=0 target=0 lun=57203 file location NMP adapter=vmhba3 channel=0 target=0 lun=57250 runtime vendor PowerPath vendor=DGC model=*250 file vendor PowerPath vendor=DGC model=*260 runtime vendor PowerPath vendor=EMC model=SYMMETRIX260 file vendor PowerPath vendor=EMC model=SYMMETRIX270 runtime vendor PowerPath vendor=EMC model=Invista270 file vendor PowerPath vendor=EMC model=Invista280 runtime vendor PowerPath vendor=HITACHI model=OPEN280 file vendor PowerPath vendor=HITACHI model=OPEN290 runtime vendor PowerPath vendor=HP model=OPEN290 file vendor PowerPath vendor=HP model=OPEN300 runtime vendor PowerPath vendor=COMPAQ model=HSV111300 file vendor PowerPath vendor=COMPAQ model=HSV111310 runtime vendor PowerPath vendor=HP model=HSV111310 file vendor PowerPath vendor=HP model=HSV111320 runtime vendor PowerPath vendor=HP model=HSV210320 file vendor PowerPath vendor=HP model=HSV210330 runtime vendor PowerPath vendor=EMC model=Celerra330 file vendor PowerPath vendor=EMC model=Celerra340 runtime vendor PowerPath vendor=HITACHI model=DF600F340 file vendor PowerPath vendor=HITACHI model=DF600F350 runtime vendor PowerPath vendor=IBM model=2107900350 file vendor PowerPath vendor=IBM model=210790065535 runtime vendor NMP vendor=* model=*

The presence of a second line with the same rule number with theword runtime in the class column indicates that the claim ruleshave loaded correctly.

6. Unclaim the device from PowerPath/VE control using the devicenumber that corresponds to the device:

a. Identify the device number associated with the device:





74



The device number starts with naa (in this casenaa.6006016029a11e0047d2fa3437a4dd11).

b. Unclaim the device using the device number:

esxcli --server <IP addr of ESX server> corestorage claiming unclaim --type=device--device=naa.6006016029a11e0047d2fa3437a4dd11






vmhba1:C0:T0:L57 state:active naa.6006016029a11e0047d2fa3437a4dd11 vmhba1 0 0 57NMP active san fc.2000001b3210553a:2100001b3210553afc.50060160c1e0a2c5:5006016041e0a2c5


Note that the device is indicated as NMP active.

Example 2This example describes how to claim LUN 55 for NMP array on anESX host with two HBAs to a CX with four storage ports (non-ALUAmode) using claim rule numbers 204–207:

1. Identify the device you want to place under NMP control:









2. Add claim rules to claim the device for NMP:





3. Verify that the claim rules have been added:



76



Rule Class Type Plugin Matches0 runtime transport NMP transport=usb1 runtime transport NMP transport=sata2 runtime transport NMP transport=ide3 runtime transport NMP transport=block4 runtime transport NMP transport=unknown101 runtime vendor MASK_PATH vendor=DELL model=Universal Xport204 file location NMP adapter=vmhba1 channel=0 target=0 lun=55205 file location NMP adapter=vmhba1 channel=0 target=1 lun=55206 file location NMP adapter=vmhba3 channel=0 target=0 lun=55207 file location NMP adapter=vmhba3 channel=0 target=1 lun=55250 runtime vendor PowerPath vendor=DGC model=*250 file vendor PowerPath vendor=DGC model=*260 runtime vendor PowerPath vendor=EMC model=SYMMETRIX260 file vendor PowerPath vendor=EMC model=SYMMETRIX270 runtime vendor PowerPath vendor=EMC model=Invista270 file vendor PowerPath vendor=EMC model=Invista280 runtime vendor PowerPath vendor=HITACHI model=OPEN280 file vendor PowerPath vendor=HITACHI model=OPEN290 runtime vendor PowerPath vendor=HP model=OPEN290 file vendor PowerPath vendor=HP model=OPEN300 runtime vendor PowerPath vendor=COMPAQ model=HSV111300 file vendor PowerPath vendor=COMPAQ model=HSV111310 runtime vendor PowerPath vendor=HP model=HSV111310 file vendor PowerPath vendor=HP model=HSV111320 runtime vendor PowerPath vendor=HP model=HSV210320 file vendor PowerPath vendor=HP model=HSV210330 runtime vendor PowerPath vendor=EMC model=Celerra330 file vendor PowerPath vendor=EMC model=Celerra340 runtime vendor PowerPath vendor=HITACHI model=DF600F340 file vendor PowerPath vendor=HITACHI model=DF600F350 runtime vendor PowerPath vendor=IBM model=2107900350 file vendor PowerPath vendor=IBM model=210790065535 runtime vendor NMP vendor=* model=*



5. Run the claim rules to verify that the claim rules have beenloaded:





Rule Class Type Plugin Matches0 runtime transport NMP transport=usb1 runtime transport NMP transport=sata2 runtime transport NMP transport=ide3 runtime transport NMP transport=block4 runtime transport NMP transport=unknown101 runtime vendor MASK_PATH vendor=DELL model=Universal Xport204 runtime location NMP adapter=vmhba1 channel=0 target=0 lun=55204 file location NMP adapter=vmhba1 channel=0 target=0 lun=55205 runtime location NMP adapter=vmhba1 channel=0 target=1 lun=55205 file location NMP adapter=vmhba1 channel=0 target=1 lun=55206 runtime location NMP adapter=vmhba3 channel=0 target=0 lun=55206 file location NMP adapter=vmhba3 channel=0 target=0 lun=55207 runtime location NMP adapter=vmhba3 channel=0 target=1 lun=55207 file location NMP adapter=vmhba3 channel=0 target=1 lun=55250 runtime vendor PowerPath vendor=DGC model=*250 file vendor PowerPath vendor=DGC model=*260 runtime vendor PowerPath vendor=EMC model=SYMMETRIX260 file vendor PowerPath vendor=EMC model=SYMMETRIX270 runtime vendor PowerPath vendor=EMC model=Invista270 file vendor PowerPath vendor=EMC model=Invista280 runtime vendor PowerPath vendor=HITACHI model=OPEN280 file vendor PowerPath vendor=HITACHI model=OPEN290 runtime vendor PowerPath vendor=HP model=OPEN290 file vendor PowerPath vendor=HP model=OPEN300 runtime vendor PowerPath vendor=COMPAQ model=HSV111300 file vendor PowerPath vendor=COMPAQ model=HSV111310 runtime vendor PowerPath vendor=HP model=HSV111310 file vendor PowerPath vendor=HP model=HSV111320 runtime vendor PowerPath vendor=HP model=HSV210320 file vendor PowerPath vendor=HP model=HSV210330 runtime vendor PowerPath vendor=EMC model=Celerra330 file vendor PowerPath vendor=EMC model=Celerra340 runtime vendor PowerPath vendor=HITACHI model=DF600F340 file vendor PowerPath vendor=HITACHI model=DF600F350 runtime vendor PowerPath vendor=IBM model=2107900350 file vendor PowerPath vendor=IBM model=210790065535 runtime vendor NMP vendor=* model=*

The presence of a second line with the same rule number with theword runtime in the class column indicates the correct loading of theclaim rules.

6. Unclaim the device from PowerPath/VE control using the devicenumber that corresponds to the device number:

a. Identify the device number associated with the device:



78







The device number is the data that starts with naa, in this casenaa.6006016029a11e0046d2fa3437a4dd11.






vmhba1:C0:T0:L55 state:active naa.6006016029a11e0046d2fa3437a4dd11 vmhba1 0 0 55NMP active san fc.2000001b3210fe3d:2100001b3210fe3dfc.50060160c1e0a2c5:5006016841e0a2c5

vmhba1:C0:T1:L55 state:active naa.6006016029a11e0046d2fa3437a4dd11 vmhba1 0 1 55NMP active san fc.2000001b3210fe3d:2100001b3210fe3dfc.50060160c1e0a2c5:5006016141e0a2c5



Note that the device is indicated as NMP active.



Preventing PowerPath/VE from claiming devicesFollow the procedure below to prevent an adapter, channel, target,and/or LUN from being claimed by PowerPath/VE:

1. Type the following command:

esxcli --server <IP addr of ESX server> corestorage claimrule add --plugin=NMP--rule <rule#> --type=location [--adapter <adapter>] [--channel <channel>][--target <target>] [--lun <lun#>]

where <rule#> is a value smaller than 250, which is the number ofthe first PowerPath/VE claim rule.

Use the output of esxcfg-mpath -L to determine the applicablevalues for the other options.

For additional information on the esxcli interface, refer toVMware documentation on Command Line Interface Installationand Reference.

2. After adding the appropriate claim rule(s), run the followingcommand:


3. Reboot the host.

Renumbering claim rulesAs discussed in “Managing claim rules” on page 66, in order forNMP to claim a device, you must add new claim rules. The new claimrule number must not already be in use and must be higher than 201and lower than the PowerPath/VE default numbers, which are 250and 260.

If the number of devices to be placed under NMP control and theirassociated paths exceeds 148, then you must modify the defaultPowerPath/VE claim rule using the VMware commandesxcli corestorage claimrule move. The move command moves aclaim rule from one rule ID to another.

The PowerPath/VE claim rule numbers cannot exceed 9999.

Procedure for renumbering claim rulesTo renumber claim rules:


80


1. List the claim rules:


2. Run the claimrule move command. For example, the movecommand below shows that the PowerPath/VE claim rules aremoved from 250 and 260 to 450 and 460, respectively.

Run the following commands:

esxcli --server <IP addr of ESX server> corestorage claimrule move -r 250 -n 450esxcli --server <IP addr of ESX server> corestorage claimrule move -r 260 -n 460esxcli --server <IP addr of ESX server> corestorage claimrule loadesxcli --server <IP addr of ESX server> corestorage claimrule run


[root@lsca0110 ~]# esxcli corestorage claimrule listRule Class Type Plugin Matches0 runtime transport NMP transport=usb1 runtime transport NMP transport=sata2 runtime transport NMP transport=ide3 runtime transport NMP transport=block4 runtime transport NMP transport=unknown101 runtime vendor MASK_PATH vendor=DELL model=Universal Xport270 runtime vendor PowerPath vendor=EMC model=Invista270 file vendor PowerPath vendor=EMC model=Invista280 runtime vendor PowerPath vendor=HITACHI model=OPEN280 file vendor PowerPath vendor=HITACHI model=OPEN290 runtime vendor PowerPath vendor=HP model=OPEN290 file vendor PowerPath vendor=HP model=OPEN300 runtime vendor PowerPath vendor=COMPAQ model=HSV111300 file vendor PowerPath vendor=COMPAQ model=HSV111310 runtime vendor PowerPath vendor=HP model=HSV111310 file vendor PowerPath vendor=HP model=HSV111320 runtime vendor PowerPath vendor=HP model=HSV210320 file vendor PowerPath vendor=HP model=HSV210330 runtime vendor PowerPath vendor=EMC model=Celerra330 file vendor PowerPath vendor=EMC model=Celerra340 runtime vendor PowerPath vendor=HITACHI model=DF600F340 file vendor PowerPath vendor=HITACHI model=DF600F350 runtime vendor PowerPath vendor=IBM model=2107900350 file vendor PowerPath vendor=IBM model=2107900450 runtime vendor PowerPath vendor=DGC model=*450 file vendor PowerPath vendor=DGC model=*460 runtime vendor PowerPath vendor=EMC model=SYMMETRIX460 file vendor PowerPath vendor=EMC model=SYMMETRIX65535 runtime vendor NMP vendor=* model=*

The presence of a second line with the same rule number with theword runtime in the class column indicates the correct loading of theclaim rules.



Configuration changes in PowerPath/VEYou must reconfigure PowerPath/VE after making configurationchanges that affect host-to-storage-system connectivity or logicaldevice identification; for example:

◆ Fibre Channel switch zone changes◆ Adding or removing Fibre Channel switches◆ Adding or removing HBAs or storage-system ports◆ Adding or removing logical devices◆ Redeploying existing logical devices

Note: If you do not reconfigure PowerPath/VE after making configurationchanges, many of these changes are treated as unintentional, andPowerPath/VE tries to prevent them from affecting applications.

Some operations fail if the affected logical devices are in use (eitherjust marked alive or with I/O in progress). This is noted in theappropriate places in the procedures that follow.

After any reconfiguration, you must monitor the outcome ofindividual reconfiguration steps and confirm that the resultingchanges are as expected before relying on the new configuration.Otherwise, some paths may not be as expected.

Configuration changes in PowerPath/VE 81

82


Adding new paths to a PowerPath/VE logical deviceThis procedure adds new paths to a logical device already configured(with at least one path) in PowerPath/VE.

This procedure can be done without interruption to runningapplications on ESX hosts. Once the reconfiguration is completedsuccessfully, applications can be run on the newly added devicepaths.

CAUTION!All operations must succeed for the reconfiguration to besuccessful. If any step fails, resolve that issue before proceeding.Do not use the new configuration until the entire procedurecompletes successfully.

Path limit Do not configure more than 32 paths per logical device.

Procedure To add paths to a PowerPath/VE logical device:

1. Run rpowermt host=<IP address of the ESX server>

display to confirm the current configuration, and run rpowermt

host=<IP address of the ESX server display dev=all toconfirm the configuration of the logical device(s) to which newpaths will be added. Ensure that the number of logical devices,hardware paths, and I/O paths are as expected. (See rpowermtdisplay in Appendix B for a description of hardware paths andI/O paths.) The path state should be alive for known good pathsand dead for known bad paths.

2. If there is a problem, correct it before proceeding.

3. Make physical path additions as required:

a. Map the logical device to additional storage-system ports.

b. Add new HBAs. For details, refer to your HBA vendordocumentation.

c. Attach cables.

d. Rezone Fibre Channel switches.

4. If using SAN Manager, Volume Logix, or Access Logix, make newpaths available to the host using those tools.



5. For VMware to recognize new paths so that PowerPath/VE canthen recognize the new paths, carry out one of the following:

• Run

esxcfg-rescan --server <IP addr of ESX server> vmhba#

Or

• In the vSphere Client, go to Configuration tab > Storageadapters link, and click Rescan.

6. Inspect the new PowerPath/VE configuration:

a. Run rpowermt host=<IP address of the ESX server>

display dev=all.

The new path(s) should be displayed with a state of alive.

b. Run rpowermt host=<IP address of the ESX server>

restore to test all paths.

c. Scan operating system error logs to ensure no errors arelogged against the new paths.

7. Correct any issues detected.

Adding new logical devices to a configurationThis procedure adds new logical devices (with one or more paths)that will be managed by PowerPath/VE.

On ESX hosts, logical devices can be added without interruption ofservice, since no existing application can be using a logical devicethat is not yet available.

Once the reconfiguration is completed successfully, new pseudodevices can be used.

CAUTION!All operations must succeed for the reconfiguration to besuccessful. If any step fails, resolve that issue before proceeding.Do not use the new configuration until the entire procedurecompletes successfully.

Adding new logical devices to a configuration 83

84


Procedure To add new logical devices to a configuration:


display to confirm the current configuration. Ensure that thenumber of logical devices, hardware paths, and I/O paths are asexpected. (See rpowermt display in Appendix B for a descriptionof hardware paths and I/O paths.) The path state should be alivefor known good paths and dead for known bad paths.

2. Make logical device and physical path changes as required:

a. Create new logical devices.

b. Map logical devices to one or more storage-system ports.

c. Add new HBAs. For details, refer to your HBA vendordocumentation.

d. Rezone Fibre Channel switches.

e. Attach cables.

3. If using SAN Manager, Volume Logix, or Access Logix, make newpaths available to the host using those tools.

4. For VMware to recognize a new logical device so thatPowerPath/VE can then recognize a new logical device, carry outone of the following:

• Run


Or

• In the vSphere Client, go to Configuration tab > Storageadapters link, and click Rescan.


a. Run rpowermt dev=all host=<IP address of the ESX

server> display.

All paths associated with the new logical devices should bedisplayed with a state of alive.


restore to test all paths to the new logical device.

c. Scan operating system error logs to ensure no errors arelogged against the new paths and logical device.



6. Correct any issues detected, before saving the PowerPath/VEconfiguration or using the new logical device.

7. Set PowerPath/VE-specific options for the new logical devices,such as load-balancing and failover policy.

Removing paths or logical devices from a configurationThis section describes how to remove the following from aPowerPath/VE configuration:

◆ Entire HBAs

◆ Logical devices

Note that because it is not possible to have mixed paths under bothNMP and PowerPath/VE control, you cannot use the claim rule toadd paths under NMP and then remove them from NMP. For thissame reason, it is not supported to remove specified paths to logicaldevices.

CAUTION!Failure to follow this procedure could cause unexpected behaviorwhen you later try to add devices to PowerPath/VE.

Procedure to remove an entire HBATo remove entire HBAs:

1. Disconnect the HBA cable.

I/O then fails to the other HBA after all the paths on that HBAshow as dead.

2. Run


to the HBA to remove the dead paths.

Alternatively:

In the vSphere Client, go to Configuration tab > Storageadapters link, and click Rescan to remove the dead paths fromPowerPath/VE.

3. Remove the physical HBA.

Removing paths or logical devices from a configuration 85

86


Procedure to remove logical devicesTo remove logical devices:

1. Stop I/O to those devices you want to remove.

2. Remove the devices from the storage group or Volume Logix.

In the case of Symmetrix, paths to other devices assigned to thesame directors go dead momentarily. In the case of CLARiiONand VNX OE, paths belonging to the same storage group go deadmomentarily. In both cases, this is a known issue and does notimpact I/O on other remaining devices. Then the paths to theremoved devices show as dead.

3. Run esxcfg-rescan vmhba# to all HBAs to remove the deadpaths.

Alternatively:

In the vSphere Client, go to Configuration tab > Storageadapters link, and click Rescan to remove the dead paths fromNMP.

An alternate procedure for removing logical devices is:

1. Use the procedure in “Managing claim rules” on page 66 to putthe devices that you want to remove to be under NMP control.

2. Remove the devices from the storage group or Volume Logix.

3. Run esxcfg-rescan vmhba# to all HBAs to remove the deadpaths from NMP.

Alternatively:

In the vSphere Client, go to Configuration tab > Storageadapters link, and click Rescan to remove the dead paths fromNMP.


a. Run rpowermt host=<IP address of the ESX server>

display. The output should show fewer total paths thanbefore. All paths should have a state of optimal.


display dev=all. All remaining paths associated with theaffected logical devices should be displayed with a state ofalive.



5. Correct any issues detected above before saving thePowerPath/VE configuration or using the new logical devices.

Managing PowerPath/VE using VMware commandsTable 5 lists the VMware vSphere CLI commands that you can use tomanage PowerPath/VE and paths under its control on ESX hosts.

Refer to VMware documentation for more information on thesecommands.

Table 5 VMware vSphere CLI commands for managing PowerPath/VE

Command Description

esxcfg-mpath Manages the multipathing modules loaded on an ESX host.

esxcfg-scsidevs Prints the mappings between vmhba names and /dev names.

esxcfg-rescan Rescan paths on an HBA to discover new storage devices.

esxcli corestorage Defines claim rules, claims and unclaims storage devices. Chapter4, “Managing path claim rules” provides more information.

resxtop Displays the I/O load on each HBA and path.

Managing PowerPath/VE using VMware commands 87

88


Managing PowerPath/VE using rpowermtTable 6 lists the PowerPath/VE commands that you can use tomanage PowerPath/VE and paths under its control on ESX hosts.

Table 6 PowerPath/VE commands (page 1 of 2)

Command Description

rpowermt check registration Checks the state of the PowerPath/VE license on a host.

rpowermt config Retrieves the latest user-assignable LUN names for devices on aCLARiiON and VNX OE system.

rpowermt display Displays the state of HBAs and devices configured for PowerPath/VE.

rpowermt display latency Displays storage-system path information, including statistics on pathlatency monitoring.

rpowermt display paths Display all paths between the HBAs on a host and the storage-systemports that the HBAs can access.

rpowermt display options Displays option settings for storage-system classes.

rpowermt help Displays rpowermt command syntax.

rpowermt restore Tests and restores specified paths. Paths that pass the test remain inservice or are returned to service. Paths that fail the test remain out ofservice or are taken out of service.

rpowermt set mode Sets paths to either active or standby mode.

rpowermt set policy Changes the load-balancing and failover policy for devices configured inPowerPath/VE.

rpowermt setpath_latency_monitor

Enables or disables path latency monitoring on all paths.

rpowermt setpath_latency_threshold

Sets a time limit for the completion of I/O on all paths.

rpowermt setperiodic_autorestore

Enables or disables the PowerPath/VE periodic autorestore facility.

rpowermt setstreamio_threshold

Sets the I/O threshold values when the StreamIO (si) load-balancing andfailover policy is in effect on PowerPath/VE.

rpowermt setup add_host Adds an ESX host and its username/password entry to the lockbox.

rpowermt setup list_hosts Lists each ESX host that has a username/password entry in the lockbox.



See “rpowermt Command Line interface” on page 109 for referenceinformation for the PowerPath/VE rpowermt commands.

Setting the rpowermt lockboxUpon entering the first rpowermt command, you are prompted tocreate an rpowermt lockbox. The rpowermt lockbox is anRSA-developed encrypted file used to store and protect sensitiveinformation. rpowermt uses the lockbox to securely store remotehost IP addresses and their username and passwordcombinations. By storing this information, rpowermt does notrepeatedly prompt you for the username and password for theremote host.


Enter lockbox passphrase:

1. Type a passphrase for the rpowermt lockbox. The passphraserequires a minimum of eight characters and must contain alowercase, uppercase, numeric, and special character.


Confirm passphrase:

2. Retype the rpowermt lockbox passphrase.

The following output appears.

Enter server username:

3. Type the ESX host username.


rpowermt setup remove_host Removes an ESX host’s username/password entry from the lockbox.

rpowermt setup update_host Updates an ESX host’s username/password entry in the lockbox.

rpowermt update lun_names Retrieves the latest user-assignable LUN names for devices on aCLARiiON and VNX OE system. (CLARiiON and VNX OE storage systemsonly.)

rpowermt version Displays the version of PowerPath/VE that is installed on the host.

Table 6 PowerPath/VE commands (page 2 of 2)

Command Description

Managing PowerPath/VE using rpowermt 89

90


Enter server password:

4. Type the ESX host password.


Confirm password:

5. Retype the ESX host password.

All users added to the lockbox should have root privileges. The ESXhost may silently fail to persist any PowerPath/VE settings set bynon-root users.

If there is any doubt about PowerPath/VE persistence, you canupdate the ESX host’s username/password entry in the lockbox. Toupdate the username/password entry in the lockbox, follow theprocedure in“Updating the host username and password in thelockbox” on page 90.

Updating the host username and password in the lockboxYou use the rpowermt setup_update host command to ESX host’susername/password entry in the lockbox.

Usage:

rpowermt setup update_host [host_file=<filename>] host=<hostname>username=<username> [password=<password>]

To update the username/password on the lockbox:

1. Run

# rpowermt setup update_host host=<hostname> username=root

where <hostname> is the IP address or hostname of the ESX host.

The following output appears.

Enter server username:

2. Type the ESX host username.


Enter server password:

3. Type the ESX host password.


Confirm password:

4. Retype the ESX host password.



Configuring path latency monitoringTo configure path latency monitoring:

1. Run rpowermt set path_latency_monitor=on|off

host=<hostname>.

2. Run rpowermt display options host=<hostname> to confirmthat path latency monitoring is enabled.

3. If you want to enable path latency monitoring to monitor the timeinterval in seconds within which I/Os should complete, run

rpowermt set path_latency_threshold=<seconds>host=<hostname>

4. Run rpowermt display latency host=<hostname> to viewinformation on I/O completion times.

Managing PowerPath/VE using rpowermt 91

92


Boot-from-SAN configuration managementThis section discusses boot-from-SAN configuration management forPowerPath/VE.

Diagnostic partitionTo run ESX server successfully, the host must have a diagnosticpartition or a dump partition configured to store core dumps fordebugging and technical support. A diagnostic partition will becreated when installing ESX by selecting RecommendedPartitioning. The installer automatically creates a diagnostic partitionfor the host. If you select Advanced Partitioning and choose not tospecify the diagnostic partition during installation, you can configureit using the Add Storage wizard via vSphere Client.

Availability of such partition can be identified using theesxcfg-dumppart -l command. Output such as the following appears:

[root@libl029 ~]# esxcfg-dumppart -lVM Kernel Name Console Name Is Active Is Configuredmpx.vmhba32:C0:T0:L0:7 /dev/sda7 no nonaa.600601601471190020af515d7cd9de11:2 /dev/sdaa2 yes yesnaa.60060160c1502200929d59631781dd11:2 /dev/sdt2 no no

With active Diagnostic partition available, there is no need toconfigure one more partition using vSphere Client.


4

This chapter discusses removing PowerPath/VE. Topics include:

◆ Task 1: Removing PowerPath/VE from a VMware ESX host..... 94◆ Task 2: Removing the PowerPath/VE rpowermt ......................... 97

RemovingPowerPath/VE

Removing PowerPath/VE 93

94

Removing PowerPath/VE

Task 1: Removing PowerPath/VE from a VMware ESX hostThis section discusses removing PowerPath/VE from a VMware ESXhost.

Pre-removal tasks Complete the following steps before removing PowerPath/VE froman ESX host:

❑ Ensure that:

❑ I/O is stopped to all devices claimed by PowerPath/VE

❑ the ESX host from which you are going to removePowerPath/VE is in maintenance mode

before removing PowerPath/VE. The PowerPath/VE uninstallfails if devices are in use. This includes any VMs on devicesclaimed by PowerPath/VE. See the PowerPath/VE 5.4 and ServicePack Releases for VMware vSphere Release Notes “Known problems”for more information.

❑ Ensure that the /root partition is not full. If the /root partition isfull PowerPath/VE may be only partially removed.

Removing PowerPath/VERemove PowerPath/VE using the vCLI tool:

Note: Both the root and non-root user can remove PowerPath/VE.

1. Check the package installed on the ESX host by typing:


2. Enter the username and password as prompted.


---------Bulletin ID--------- -----Installed---------------------Summary-----------------EMC-PP5.4.0.00.00-<bbb> <TIMESTAMP> PowerPath for ESX

where <bbb> is the PowerPath/VE build number.

3. Remove the PowerPath/VE package by typing the followingcommand:



# vihostupdate --remove --bulletin EMC-PP5.4.0.00.00-<bbb> --server <IP addressor host name>

where <bbb> is the current build number and <server number> isthe host number.

The removal may take a few minutes.

4. Enter the username and password as prompted.


The update completed successfully, but the system needsto be rebooted for the changes to be effective.


Note that removal while VMs are running fails but is reported by theremote tool with success status. This is a known issue and has beenreported to VMware. See the PowerPath/VE 5.4 and Service PackReleases for VMware vSphere Release Notes “Known problems” for moreinformation.

Removing PowerPath/VE in a PowerPath/VE-controlled boot-from-SAN setupTo remove PowerPath/VE 5.4 for VMware in anPowerPath/VE-controlled boot-from-SAN setup:

1. Add a low-number claim rule to put all the devices on the samearray class of the boot device under NMP control. Type:

esxcli corestorage claimrule add --plugin="NMP"--type=vendor --vendor=<vendor name> --model=*--rule=<rule number>

The new claim rule number must not already be in use and mustbe higher than 201 and lower than the lower of the twoPowerPath/VE defaults (which is 250).

The vendor name is the vendor name of the array that the host isboot from.




Task 1: Removing PowerPath/VE from a VMware ESX host 95

96


At this point, all the devices from that array including the bootdevice are under NMP control.

3. Remove PowerPath/VE following the procedure in “RemovingPowerPath/VE” on page 94.

4. Delete the rule you added in step 1 on page 95 . Type:

esxcli corestorage claimrule delete --rule=<rule number>




Post-removal tasksThere are no post-removal tasks.



Task 2: Removing the PowerPath/VE rpowermtThis section discusses removing the rpowermt package, calledRTOOLS, from the rpowermt server.

PowerPath/VE rpowermt pre-removal tasksThere are no pre-removal tasks.

Removing PowerPath/VE rpowermtThis section provides procedures for removing PowerPath/VErpowermt from Linux and Windows systems.

Removing PowerPath/VE rpowermt from LinuxTo remove the PowerPath/VE rpowermt package from a Linuxmachine:

1. Log in as root.

2. Display the package name. Enter:

rpm -qa | grep EMCPower.RTOOLS


EMCPower.RTOOLS-5.4.0.00.00-<bbb>

where <bbb> is the build number. The output is what you usewith the rpm -e command in step 3 .

3. Remove the software. Enter:

# rpm -e EMCPower.RTOOLS-5.4.0.00.00-<bbb>

where <bbb> is the build number.

Removing PowerPath/VE rpowermt from WindowsTo remove PowerPath/VE rpowermt package from a Windowsmachine:

◆ In Windows 2003: From the Start menu, select Settings > ControlPanel > Add or Remove Programs > EMC > PowerPath RemoteTools and click Remove.

Task 2: Removing the PowerPath/VE rpowermt 97

98


◆ In Windows 2008: From the From the Start menu, selectSettings > Control Panel > Programs and Features > EMC >PowerPath Remote Tools and click Uninstall.

Post-removal tasksThere are no post-removal tasks.


5

This chapter discusses troubleshooting the PowerPath/VEinstallation. Topics include:

◆ Collecting logs from vSphere Center Server or vSphere client . 100◆ Log files ............................................................................................. 101◆ TCP port conflicts............................................................................. 102◆ Problems when performing vihostupdate ................................... 103◆ Reporting a problem........................................................................ 105

TroubleshootingPowerPath/VE

Installation

Troubleshooting PowerPath/VE Installation 99

100

Troubleshooting PowerPath/VE Installation

Collecting logs from vSphere Center Server or vSphere clientYou can collect crash dump logs remotely from vSphere CenterServer or vSphere client.

1. To collect logs remotely from vSphere Center Server or vSphereclient:

• In vSphere Center Server:

a. Select File > Export System Logs.

The Export System Logs window appears.

b. Select the ESX host from which you want to collect the log.

• In vSphere client:

Select File > Export System Logs.

The Export System Logs window appears.

2. Choose the location where you want to save the logs:

a. Browse for the location you want.

b. When you have selected the desired file location, click OK.

It takes a few moments for task to complete.

The logs include VSphere Client and Hosts.

The log collection can be found at the location specified in step 2above.



Log filesYou can use certain log files to check trace messages from variousapplications and to debug. Table 4 on page 101 lists the file type andnames for VMware vSphere message tracing and error debugging forthe PowerPath/VE versions.

Note that the following error is seen in /var/log/vmkernel forLUNZ and VCMDB devices:

ScsiClaimrule: 709: Path vmhba1:C0:T1:L0 is claimed byplugin NMP, but current claimrule number 260 indicatesthat it should be claimed by plugin PowerPath.

You should disregard this message. This is expected behavior becausePowerPath does not manage those types of devices. The VMware PRnumber for this issue is 407152.

Table 4 VMware vSphere error log file types and locations

Type of host File type VMware version File location

ESX Installation log • 4.0• 4.0.x• 4.1

/var/log/vmware/esxupdate.log

Kernel log /var/log/vmkernel

ESXi Installation log • 4.0• 4.0.x

/locker/db/esxupdate.log

• 4.1 /var/log/messages

Kernel log • 4.0 /var/log/messages

• 4.0.x /var/log/vmkernel

• 4.1 /var/log/messages

Log files 101

102


TCP port conflictsIf you have two ELM license servers running on the same host andthey are using the same default TCP port, you receive the followingerror message:

Warning: Failed to get a license from the server..

For both ELM servers and vendor daemons to coexist on the samehost, you must change the TCP port number from the default of27010 to a number that does not conflict with another application onyour host. You must modify the TCP port number in thePowerPath/VE license file that resides on the ELM license server andthe PowerPath/VE license configuration file that resides on therpowermt server. A valid number is any unused port numberbetween 0 and 64000.

For example, the license configuration file with the default TCP portwould look like this:

SERVER lsca2147 INTERNET=172.23.168.142 27010VENDOR EMCLMUSE_SERVER

The server license file downloaded from the Powerlink Licensingportal would look like this:

SERVER lsca2147 INTERNET=172.23.168.142 27010VENDOR EMCLMUSE_SERVERINCREMENT PowerPathMP EMCLM 5.4 permanent 1 OVERDRAFT=1 ISSUER=EMC \

ISSUED=29-May-2009 NOTICE="ACTIVATED BY 8918825 ABC Corp \15260 VENTURA BLVD SHERMAN OAKS CA US \91403" SIGN="0074 DXB7 4X13 B1XE CX23 84D3 4X00 BCXD 9028 \C2XE 9AX4 88E8 XC6A"

If the TCP port number 27010 conflicts with another application, youmust modify it in the license configuration file and in the serverlicense file to a another unused port number between 0 and 64000.

For example, you might modify the TCP port number to 27050, sothat the license configuration file looked as follows:

SERVER lsca2147 INTERNET=172.23.168.142 27050VENDOR EMCLM /sbin/EMCLM.exe PORT=27011USE_SERVER



In the license server file you would also modify the TCP port numberto 27050:

SERVER lsca2147 INTERNET=172.23.168.142 27050VENDOR EMCLM /sbin/EMCLM.exe PORT=27011USE_SERVERINCREMENT PowerPathMP EMCLM 5.4 permanent 1 OVERDRAFT=1 ISSUER=EMC \

ISSUED=29-May-2009 NOTICE="ACTIVATED BY 8918825 ABC Corp \15260 VENTURA BLVD SHERMAN OAKS CA US \91403" SIGN="0074 DXB7 4X13 B1XE CX23 84D3 4X00 BXED 9028 \C2XE 9AX4 88E8 XC6A"

See the PowerPath/VE for VMware vSphere Licensing Guide, available onPowerlink, for more information.

Problems when performing vihostupdateIf when performing a vihostupdate you receive the error message

Operation not applicable: The operation is not supportedon the object.

check that you have the updated version of vSphere vCLI installed onthe host.

Problems when performing vihostupdate 103

104


Incompatible rpowermt lockbox versionsThe PowerPath/VE 5.4 SP2 and PowerPath/VE 5.4 SP1 versions ofthe rpowermt lockbox are not backward compatible with thePowerPath 5.4 version of the rpowermt lockbox. OncePowerPath/VE 5.4 SP2 or PowerPath/VE 5.4 SP1 has created orconverted the lockbox, the PowerPath/VE 5.4 version of rpowermtwill not be able to read the PowerPath 5.4 SP2 or SP1 version of therpowermt lockbox.

If you use incompatible versions, the following error messageappears:

Error: failed to open host file.

To correct this, remove the incompatible version of the rpowermtlockbox:


version to see the version of lockbox you have on your host.


<username>lsca2147.lss.abc.com -- src> rpowermt versionEMC rpowermt for PowerPath (c) client Version 5.4 SP1(build xxx)License search path:/etc/emc:/etc/emc/licenses:/opt/EMCpower:/opt/EMCpower/licensesHost file: /etc/emc/lockbox/<username>.clb

2. Delete rpowermt lockbox file. In the example output, therpowermt lockbox file is /etc/emc/lockbox/<username>.clb.

3. Set up the same version of rpowermt lockbox as PowerPath/VEthat you are running on your host following the procedure in“Setting the rpowermt lockbox” on page 89.



Reporting a problemIf you find a problem in PowerPath/VE 5.4 for VMware vSphere,collect crash dump logs from vSphere Center Server or vSphere client(see “Collecting logs from vSphere Center Server or vSphere client”on page 100) and submit the generated file with system informationand dump, along with any other supporting material, to EMCCustomer Support.

Additionally, the EMC Grab Utility 1.2 for ESX/vSphere, whichcollects system and software configuration information, supports theESX host only (ESXi host is not supported by this utility). Submit thegenerated tar file, along with any other supporting material, to EMCCustomer Support.

You can download EMC Grab from the Powerlink website. SelectSupport and locate the section for Product and Diagnostic Tools andGrab Utilities. Refer to the emcgrab ReadMe.txt file on Powerlink forprerequisite information and instructions on using the utility.

If you encounter a problem with Powerlink Licensing via Powerlink,contact [email protected].

Reporting a problem 105

[email protected]

106



A

This appendix lists files that are created or modified by thePowerPath/VE installation.

Files Installed andModified by

PowerPath/VE

Files Installed and Modified by PowerPath/VE 107

108

Files Installed and Modified by PowerPath/VE

Files installed and modified by the PowerPath/VE installationThe following files are installed and modified when PowerPath/VEis installed on an ESX host:

◆ /etc/cim/emc/EMCProvider.conf◆ /etc/cim/emc/db◆ /etc/cim/emc/log◆ /etc/vmware/init/init.d/20.PowerPath◆ /etc/vmware/init/manifests/EMC-emcp.mf◆ /lib/cim/libEmcOsls4Api.so◆ /lib/cim/libEmcOsls4Apisehost.so◆ /lib/cim/libEmcOsls4CMPI.so◆ /lib/cim/libEmcOsls4sehost.so◆ /lib/cim/libemcp_mpapi_rtl.so◆ /lib/cim/libsm_clsapi.so◆ /lib/cim/libsm_ofl.so◆ /lib/cim/libsm_sms.so◆ /lib/cim/libsyminqapi.so◆ /usr/lib/vmware/vmkmod/emcp◆ /usr/sbin/powermt


B

This appendix describes the rpowermt command line interface.

rpowermt CommandLine interface

rpowermt Command Line interface 109

110

rpowermt Command Line interface

rpowermt command summaryTable 5 lists the rpowermt commands that you can use to managePowerPath/VE and paths under its control on ESX hosts.

Table 5 rpowermt commands (page 1 of 2)

Command DescriptionWheredescribed

rpowermt check registration Checks the state of the PowerPath/VE license for an ESX host. page 120

rpowermt config Retrieves the latest user-assignable LUN names for devices on a CLARiiON andVNX OE system.

page 121

rpowermt display Displays the state of HBAs and devices managed by PowerPath/VE. page 122

rpowermt display latency Displays storage-system path information, including statistics on path latencymonitoring.

page 128

rpowermt display options Displays option settings for storage-system classes. page 131

rpowermt display paths Displays path-specific information. page 132

rpowermt display ports Displays port-specific information. page 135

rpowermt help Displays rpowermt command syntax. page 139

rpowermt register Registers a PowerPath/VE license for an ESX host. page 141

rpowermt restore Tests and restores paths. page 142

rpowermt set mode Sets paths to either active or standby mode. page 144

rpowermt setpath_latency_monitor

Enables or disables path latency monitoring on all paths. page 147

rpowermt setpath_latency_threshold

Sets a time limit for the completion of I/O on all paths. page 148

rpowermt setperiodic_autorestore

Enables or disables periodic autorestore. page 150

rpowermt set policy Changes the load-balancing and failover policy for devices managed byPowerPath/VE.

page 151

rpowermt setstreamio_threshold

Sets the I/O threshold values when the StreamIO (si) load-balancing and failoverpolicy is in effect on PowerPath/VE.

page 153

rpowermt setup add_host Adds an ESX host and its username/password entry to the lockbox. page 155

rpowermt setup list_hosts Lists each ESX host that has a username/password entry in the lockbox. page 157



rpowermt setup remove_host Removes an ESX host’s username/password entry from the lockbox. page 158

rpowermt setup update_host Updates an ESX host’s username/password entry in the lockbox. page 159

rpowermt unregister Removes a PowerPath/VE license from an ESX host. page 160

rpowermt update lun_names Retrieves the latest user-assignable LUN names for devices on a CLARiiON andVNX oe system.

page 161

rpowermt version Displays the version of PowerPath/VE installed on an ESX host. page 162

Table 5 rpowermt commands (page 2 of 2)

Command DescriptionWheredescribed

rpowermt command summary 111

112


rpowermt syntax summaryThe rpowermt command has the following syntax:

rpowermt <command>[class=all|symm|clariion|invista|hitachi|hpxp|ess|generic][dev=<device>|all] host=<hostname>

rpowermt help [<command>]rpowermt check_registration host=<hostname>rpowermt config host=<hostname>rpowermt display [dev=<devices>|all] [class=<classname>|all]

[every=<#seconds>] [width=<#col>] host=<hostname>rpowermt display latency [dev=<devices|all>] [class=<classname>|all]

[every=<#seconds>] [width=<#col>] host=<hostname>rpowermt display optionsrpowermt display paths [class=<classname>|all] [every=<#seconds>]

[width=<#col>] host=<hostname>rpowermt display ports [dev=<devices>|all] [class=<classname>|all]

[every=<#seconds>] [width=<#col>] host=<hostname>rpowermt register [force] host=<hostname> [port=<port#>]rpowermt restore [hba=<hba#>|all] [dev=<device>|all]

[class=<classname>|all] host=<hostname>rpowermt set path_latency_monitor=on|off [force] host=<hostname>rpowermt set path_latency_threshold=<#seconds> [force]

host=<hostname>rpowermt set periodic_autorestore=on|off [class=<classname>|all]

[force] host=<hostname>rpowermt set policy=<ad|co|lb|li|rr|so|si>

[dev=<devices>|all] [class=<classname>|all][force] host=<hostname>

rpowermt set streamio_threshold=<threshold_count>[dev=<device|all>] [class=<class|all>]

rpowermt set mode=[active|standby][hba=<hba#>|all] [dev=<device>|all] [class=<classname>|all][force] host=<hostname>

rpowermt setup add_host [host_file=<filename>] host=<hostname>username=<username> [password=<password>]

rpowermt setup update_host [host_file=<filename>] host=<hostname>username=<username> [password=<password>]

rpowermt setup remove_host [host_file=<filename>] host=<hostname>rpowermt setup list_hosts [host_file=<filename>]rpowermt unregister [force] host=<hostname>rpowermt update lun_names host=<hostname>rpowermt version host=<hostname>



rpowermt error and warning messagesTable 6 describes the error messages returned by the rpowermtcommand.

Table 6 rpowermt error messages (page 1 of 3)

Message Error

ERROR: Bad dev value emcpower<invalid value>, or notunder Powerpath control.

An invalid device number was specified.

ERROR: Bad hba value <invalid value>, or not underPowerpath control.

An invalid HBA number was specified.

ERROR: <array_type> device(s) not found. The rpowermt set policy command specified a policy that isvalid for only one class, but no devices are configured for that class.

ERROR: <class> device(s) not found. Both the class and dev arguments were used, but the specifiedvalues do not match a currently configured PowerPath/VE volumeand device path.

A management command (for example, a set command) was issuedfor an ESX host, but no devices were configured for the specifiedclass.

ERROR: Device(s) not found. A management command (for example, a set command) was issued,but no devices are being managed by PowerPath/VE.

ERROR: "every" value <invalid value> is not in range (1 -86400).

An invalid interval was specified.

ERROR: Failed to store host information into host file. An encryption-specific error occurred. Such errors, while extremelyrare, may prevent rpowermt from properly managing the lockbox.

ERROR: Failed to retrieve host information from host file.

ERROR: Failed to remove host information from host file.

ERROR: Failed to store host list into host file.

ERROR: Failed to retrieve host list from host file.

ERROR: Failed to find IP address for <unresolvedhostname>.

An rpowermt command was not able to resolve the IP for thehostname supplied with the host argument.

ERROR: Failed to access host file directory. The specified lockbox location included an invalid directory.

ERROR: Host not found. A supplied hostname or IP address cannot be reached, or there is aTCP port conflict.To troubleshoot TCP port conflicts, see “TCP port conflicts” onpage 102 in “Troubleshooting PowerPath/VE Installation.”

rpowermt error and warning messages 113

114


ERROR: Failed to open host file. User does not have the necessary privileges/permissions to accessthe specified lockbox.

An rpowermt setup command specified a lockbox that doesnot exist.

ERROR: Invalid options: <invalid_value> An invalid argument was specified.

ERROR: Invalid passphrase during host file creation. The passphrase entered for a lockbox did not meet the passphraserequirements (that is, a minimum of 8 characters and must contain alowercase, uppercase, numeric, and special character).

ERROR: <invalid_policy> is not a valid policy for <class>. The specified policy is not valid for the specified class.

ERROR: License authorization failed. The rpowermt register command located the license file,but the file did not contain a key for the specified ESX host.

ERROR: License checkout failed. The rpowermt register command located the license file,but the attribute value within the license file is corrupted.

ERROR: License file does not exist or not found. The rpowermt register command could not locate a licensefile.

ERROR: License file format error or misspelling. The license file is incorrectly formatted.

ERROR: License key or feature not found. The rpowermt register command located the license file,but the feature tag within the license file is corrupted.

ERROR: License not installed. A management command (for example, a set command) was issuedfor an ESX host that does not have a valid license registered for it.

ERROR: Missing option. An rpowermt command (other than version) was run, but thehost argument was not specified.

ERROR: Path latency threshold value <invalid_value> notin range (1 - 2147483647).

Therpowermtsetpathlatencythreshold commandspecified an invalid value.

ERROR: Permission denied. User does not the necessary privileges/permissions to create alockbox at the specified location.

ERROR: Policy must be set to StreamIO in order to setthreshold.

The load-balancing policy on the specified device is not si.

ERROR: PowerPath Not Found PowerPath is not installed on the specified host.

ERROR: Received unsupported command A valid streamio_threshold value was specified, but the version ofPowerPath/VE does not support this parameter.


Message Error



Table 7 describes the warning messages returned by the rpowermtcommand.

ERROR: Requested host cannot be found in host file. The lockbox specified with the rpowermt setupremove_host command does not have an entry for the specifiedhost.

ERROR: StreamIO threshold value <invalid value> is not inrange (64 - 2048).

An invalid streamio_threshold value was specified.

ERROR: Unsupported class name: <invalid class>.class=<symm|clariion|hitachi|hpxp|ess|invista|all>

An invalid class argument was specified.

ERROR: User Authentication Invalid. A rpowermt command (other than help or setup) was run,but the specified lockbox did not exist. rpowermt attempted tocreate a lockbox, but the supplied username and/or password wasinvalid.

ERROR: Unsupported policy: <invalid_policy> The rpowermt set policy command specified an invalidpolicy.

ERROR: "width" value <invalid value> is not in range (80 -160).

An invalid width was specified.


Message Error

Table 7 rpowermt warning messages

Message Error

Warning: PowerPath driver is already registered. The rpowermt register command was issued for an ESXhost that already has a PowerPath/VE license registered for it.

Warning: Device(s) not found. A non-management command (for example, a display command)was issued, but no devices are being managed by PowerPath/VE.

Warning: License will expire in <number_of_days> days. A registered PowerPath/VE license is set to expire within thespecified number of days.

Warning: License feature has expired. The license file was found, but it has expired.

Warning: License not installed. A non-management command (for example, a display command)was issued for an ESX host that does not have a valid licenseregistered for it.

Warning: <class> device(s) not found. A non-management command (for example, a set command) wasissued for an ESX host, but no devices were configured for thespecified class.

rpowermt error and warning messages 115

116




rpowermt server lockboxThe rpowermt command uses a lockbox to securely store ESX host IPaddresses and their corresponding username/password. Storing thisinformation eliminates the need for supplying a username/passwordfor the ESX host when running the rpowermt command.

On Linux, the default lockbox is/etc/emc/lockbox/<username>.clb, where <username> is theusername of the rpowermt user.

Note: The username is obtained by the getpwuid_r() library call. Somesystems may not support this library call. For those systems you can disablethis internal call by setting the PPMT_DISABLE_PW_LOOKUP environmentvariable. The default lockbox on these system is/etc/emc/lockbox/lockbox.clb.

The default lockbox on Windows is C:\Documents and

Settings\<username>\My

Documents\EMC\PowerPath\rpowermt\lockbox.clb.

Use the PPMT_LB_FILE environment variable to change the lockboxname and location.

For example, to change the lockbox name to /tmp/pplockbox.clb ona Linux system, enter:

PPMT_LB_FILE=/tmp/pplockbox.clb; export PPMT_LB_FILE

To change the lockbox name to C:\temp\pplockbox.clb on aWindows system, enter:

set PPMT_LB_FILE=C:\Temp\pplockbox.clb

Any rpowermt command that includes the host argument attemptsto retrieve the username/password for the specified ESX host fromthe lockbox. If the lockbox does not exist, rpowermt creates it,prompting you for the ESX host’s username/password and the newlockbox’s passphrase:

Enter server username: root

Enter lockbox passphrase:Confirm passphrase:Enter server password:Confirm password:

rpowermt server lockbox 117

118


The passphrase must meet the following requirements:

◆ 8 or more characters in length.

◆ Contain at least one numeric, one uppercase, one lowercasecharacter, and one non-alphanumeric character (such as # or !).

You will need this passphrase if the lockbox is moved to a differentrpowermt server.

If the lockbox exists but rpowermt cannot retrieve theusername/password for the ESX host (for example, no entry in thelockbox pertains to the specified host or the existingusername/password stored for the specified ESX host is no longervalid), rpowermt prompts you for the ESX host credentials of thespecified host:

Enter server username: root

Enter server password:Confirm password:

rpowermt CST libraries environment variableIf the rpowermt CST libraries have been moved, you need to set anenvironment variable to point the rpowermt server to the correctlocation. This issue may occur if you have more than one EMCproduct installed on the rpowermt server and those otherapplications use different versions of the lockbox libraries.

The optional environment variable to point the rpowermt server tothe correct directory is PP_LB_LIB.

On Windows, enter:

PP_LB_LIB=C:\program files\EMC\PowerPath\rpowermt

On Linux, enter:

export PP_LB_LIB=/usr/lib

On Linux hosts, if you receive an error message like the following:

ERROR: Failed to open host file.The corrective action is to set the environment variable. Enter:

PP_LB_LIB=<location of CST libraries>



Alternatively, remove the environment variable definition forLD_LIBRARY_PATH. Enter:

unset LD_LIBRARY_PATH

This is of or pertaining to issue 339891.

rpowermt server lockbox 119

120


rpowermt check_registration: Check PowerPath/VE licenseinformation

Description rpowermt check_registration checks the state of thePowerPath/VE license on an ESX host.

Syntax rpowermt check_registration host=<hostname>

Arguments host=<hostname>IP address or hostname of the ESX host.

Example Output like the following is returned for an ESX host with a validlicense:

PowerPath License Information:------------------------------

Host ID : 4nf0a50b-d625-bae7-57e0-002219288018Type : unserved (uncounted)State : licensed

Days until expiration : (non-expiring)

License search path:/etc/emc:/etc/emc/licenses:/opt/EMCpower:/opt/EMCpower/licenses

License file(s): /etc/emc/powerpath_node.lic/etc/emc/licenses/opt/EMCpower/opt/EMCpower/licenses



rpowermt config: Retrieve CLARiiON nice names

Note: The rpowermt config command is provided only to maintaincompatibility with the versions of the powermt command used on theWindows, Linux, and UNIX platforms. rpowermt config provides nofunctionality other than retrieving CLARiiON nice names (which is the samefunctionality provided by the rpowermt update lun_names command).

Description rpowermt config retrieves the latest CLARiiON nice names. ACLARiiON nice name, also called a user-assignable LUN name, is acharacter string associated with a logical device on a CLARiiON andVNX array and assigned through Unisphere or Navisphere. Run thiscommand to perform initial discovery of CLARiiON nice names andwhenever the nice name is changed on the CLARiiON and VNXarray through Unisphere or Navisphere.

Note: CLARiiON LUN nice names are updated automatically every hour.Use this command to manually update CLARiiON nice names.

Syntax rpowermt config host=<hostname>


Example CLARiiON nice names are disabled in the following rpowermt

display dev output:

Pseudo name=emcpower2CLARiiON ID=HK190807490007 []Standard UID=naa.60060160bab01e00930a8cdde601dd11 []

CLARiiON nice names are enabled in the following rpowermt

display dev output:

Pseudo name=emcpower2CLARiiON ID=HK190807490007 [lclb164]Standard UID=naa.60060160bab01e00930a8cdde601dd11 [LUN 4]

Note the storage group name (lclb164) and the device name (LUN 4)are now displayed.

rpowermt config: Retrieve CLARiiON nice names 121

122


rpowermt display: Monitor HBAs and devicesDescription rpowermt display displays information about HBAs and/or devices

managed by PowerPath/VE.

Syntax rpowermt display [dev=<devname>|all][class=<classname>|all] [every=<#seconds>][width=<#col>] host=<hostname>

Arguments Note: The every=<#seconds> and width=<#col> arguments can be usedwith any other option.

dev=<devname>|allDisplays information about the specified devices. all specifies alldevices. If dev is not specified, a summary HBA listing isdisplayed, with one line per HBA.

The device number can either be the PowerPath/VE pseudodevice name (for example, emcpower4 or simply, 4) or thestandard UID (for example,naa.600601606f3023009a29069be840de11), which is theESX-assigned identifier for the PowerPath/VE device.

class=all|symm|clariion|invista|hitachi|hpxp|ess|genericLimits the command to the specified type of storage system. VNXdevices are managed under the clariion storage class. VNXedevices are managed under the generic storage class. VPLEXdevices are managed under the invista storage class. allspecifies all storage-system types. The default is all.

every=<#seconds>Integer in the range 1 to 86400 that specifies, in seconds, howoften the display is updated. By default, the display is notupdated.

The #seconds value is the minimum time between refreshes; theactual time is affected by the overall system load. On busysystems, display updates can be less frequent than specified.

width=<#cols>Sets the display width to a constant value within the range of 80to 160 columns. This value applies only to the specific output; it isnot persistent. You can set a persistent value for thewidth=columns option with the PP_DISPLAY_WIDTH environmentvariable. This variable changes the width setting from 80 (thedefault) to the value you specify.



If the environment variable and command line values differ, thecommand line value takes precedence.

host=<hostname>IP address or hostname of the ESX host.

Examples With no command line device argument the HBA summary displayis shown.

# rpowermt display host=111.222.222.111

Symmetrix logical device count=2CLARiiON logical device count=3==============================================================================----- Host Bus Adapters --------- ------ I/O Paths ----- ------ Stats ------### HW Path Summary Total Dead IO/Sec Q-IOs Errors==============================================================================

1 vmhba3 optimal 10 0 - 0 02 vmhba4 optimal 10 0 - 0 0

Table 8 explains the fields in the HBA display.

Table 8 Fields in the HBA display (page 1 of 2)

Field Value Description

<array> logicaldevice count

Non-negativeinteger

Total number of unique logical devices from all storage devices of a given type that areconfigured by PowerPath/VE and that this host can access.Each rpowermt display table contains information about one type of storage system,such as a Symmetrix, CLARiiON, VNX OE, or third-party array.

Host BusAdapters ###

Non-negativeinteger

PowerPath/VE number for the HBA.When the dev option is used, the output of rpowermt display identifies the HBA bythis same HBA number.This number is preserved after configuration changes but is not preserved across reboots.

Host BusAdapters HWPath

Alphanumericstring

ESX-assigned name for the HBA.

I/O PathsSummary

optimaldegradedfailedunlic

Status of the paths originating from this HBA:• optimal means all paths are alive (usable).• degraded means one or more, but not all, paths from this HBA are dead (not usable).• failed means all paths are dead and no data is passing through this HBA.• unlic indicates that the ESX host does not have a valid PowerPath/VE license registered

for it.

I/O Paths Total Non-negativeinteger

Total number of paths that originate from this HBA. The total number of paths may exceedthe number of logical devices in a complex SAN topology, due to zoning.

rpowermt display: Monitor HBAs and devices 123

124


With a device argument entered, the summary display for that deviceis shown.

# rpowermt display dev=1 host=111.222.222.111

Pseudo name=emcpower1Symmetrix ID=000190100286Logical device ID=4230Standard UID=naa.60060480000190100286533034323330state=alive; policy=SymmOpt; queued-IOs=0==============================================================================---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---### HW Path I/O Paths Interf. Mode State Q-IOs Errors==============================================================================

1 vmhba3 C0:T0:L31 FA 9aA active alive 0 02 vmhba4 C0:T1:L31 FA 9bA active alive 0 02 vmhba4 C0:T0:L31 FA 9aA active alive 0 01 vmhba3 C0:T1:L31 FA 9bA active alive 0 0

I/O Paths Dead Integer in therange 0 - I/OPaths Total

Total number of paths originating from this HBA that are dead (not usable).

Stats IO/sec Non-negativeinteger

This field is blank for rpowermt display, unless it is used with the every parameter.Subsequent rpowermt display iterations display the average number of I/Os sentacross this bus each second.

Stats Q-IOs Non-negativeinteger

Total number of outstanding I/Os on this HBA.

Stats Errors Non-negativeinteger

Total number of times this path transitioned from alive to dead. This is always equal to or lessthan the total number of HBA I/O path errors. Note that the error count may be different fordifferent LUNs using the same physical paths. This value is cleared at boot time or whenrpowermt restore executes.

Table 8 Fields in the HBA display (page 2 of 2)




Table 9 describes the fields in the devices display.

Table 9 Fields in the devices display (page 1 of 3)


Pseudo name Alphanumericstring

Platform-specific value assigned by PowerPath/VE to the device.

Array ID Hexadecimalvalue oralphanumericstring

Identification number for the storage system on which the logical device is located. Eachrpowermt display table contains information about one type of storage system, such as aSymmetrix or CLARiiON and VNX OE.

Logical deviceID

Hexadecimalvalue oralphanumericstring

Identification number for the logical device. Each logical device on each storage system has aunique ID. Each storage system, however, uses the same storage-system ID. Together,storage-system ID and logical-device ID create a unique ID for every logical device in the world.

Standard UID Alphanumericstring

ESX-assigned identifier for the device.

state alivedead

State of the PowerPath/VE device:

• PowerPath/VE marks a PowerPath/VE device dead if it returns certain types of I/O errors andis judged unusable. Once a PowerPath/VE device is marked dead (and until it is restored),PowerPath/VE returns subsequent I/O requests with a failure status, without forwarding themto the associated logical device. This prevents further, unrecoverable corruption and allowsthe user to perform data recovery if needed.

• If a PowerPath/VE device is alive, either PowerPath/VE never marked it dead, or it wasmarked dead but restored with rpowermt restore or an automatic path test.

Owner default=SPx,current=SPx

Default and current owners of the logical device (CLARiiON and VNX OE systems only).

policy AdaptiveBasicFailoverCLAROptLeastBlocksLeastIOsRoundRobinSymmOptStreamIO

Current load-balancing and failover policy for the device.Use rpowermt set policy to change the policy. Refer to “rpowermt set policy: Set theload-balancing and failover policy” on page 151.

queued- IOs Non-negativeinteger

Number of I/O requests queued to this PowerPath/VE device.


126


Array failovermode

1 or 4 Failover mode set on a CLARiiON and VNX OE array. This field displays only whenPowerPath/VE is managing CLARiiON and VNX OE LUNs. You set this mode through theNavisphere and Unisphere software.

Note: The array failover mode must be the same for all paths that access a single LUN. If twopaths access the same LUN, and one path is set to PNR (passive not ready) mode and one toALUA (asymmetric logical unit access) mode, PowerPath/VE behavior is undefined for that LUN.

Modes supported with PowerPath/VE include:• 1—passive not ready (PNR) mode.• 4—asymmetric logical unit access (ALUA) mode.

Host ### Non-negativeinteger

PowerPath/VE number for the HBA.When the dev option is used, the output of rpowermt display identifies the HBA by thissame HBA number.This number is preserved after configuration changes but is not preserved across reboots.

Host HW Path Alphanumericstring


Host I/O Path Alphanumericstring

The platform-specific device name for the path.

Stor Interf. Alphanumericstring

Storage-system interface.For Symmetrix systems, this has three parts:• Interface type: Fibre Channel (FA) or SCSI (SA)• Interface address: integer in the range 1 to 16• Interface port: [abcd][AB]For CLARiiON and VNX OE systems, this has one part:• Interface port: SP [A-B][0-15]

I/O PathsMode

activestandbyunlic

Current path mode:• active indicates this path can accept I/O. Load balancing is performed for a device with more

than one active path, based on the load-balancing and failover policy set for the device. Onactive-passive storage systems, I/O is load balanced across paths to a single SP for eachlogical device.

• standby indicates this path is held in reserve. Being set to standby does not mean a path willnot be used. It only means the weight of the path is heavily adjusted to preclude its use innormal operations. A standby path still can be selected if it is the best path for a request. Forexample, if the active paths are so heavily loaded that the weighting factor in favor of theactive path is overcome, the I/O will be sent to the standby path.

• unlic indicates that the ESX host does not have a valid PowerPath/VE license registered for it.





I/O PathsState

alivedead

Current path state:• alive indicates the path is usable: PowerPath/VE can direct I/O to this path.• dead indicates the path is not usable: PowerPath/VE will not direct I/O to this path. After

marking the path dead and incrementing the Errors count, PowerPath/VE tests the path tosee whether it is usable. If the test succeeds, PowerPath/VE marks the path alive; the path isthen available for I/O. If the test fails, the path remains dead, and PowerPath/VE ignores it forsubsequent I/O operations. If all the paths to a logical device are dead, PowerPath/VE retestseach path.


Total number of I/O operations under way to this path. This is the total number of I/O requests tothis device that have not completed. The sum of in-progress I/Os for all paths should equal thenumber of in-progress I/Os for the PowerPath/VE device.

Stats Errors Non-negativenumber





128


rpowermt display latency: Monitor path latency

Description rpowermt display latency displays information on path latencymonitoring, including the last and longest I/O completion times foreach path.

Valid latency values are displayed only if path_latency_monitor is setto on. A value of 0 is displayed otherwise.

The rpowermt display latency output includes the following twocolumns:

◆ Current — Time in miliseconds it took the last I/O on the path tocomplete.

◆ Max — Time in miliseconds it took the longest I/O on the path tocomplete.

The Current and Max values can help determine an appropriatethreshold value, which you can then set with the rpowermt setpath_latency_threshold command.

To turn on path latency monitoring, use the rpowermt setpath_latency_monitoring command. To view whether path latencymonitoring is enabled, use the rpowermt display optionscommand.

Syntax rpowermt display latency [dev=<devices>|all][class=<classname>|all] [every=<#seconds>][width=<#col>] host=<hostname>

Arguments dev=<devices>|allDisplays latency information for the specified device or for alldevices.





every=<#seconds>Specifies the frequency with which path-latency-monitoringinformation displays. If a value is not specified, then theinformation is displayed once. The value can be an integer in therange of 1 to 86400.


width=<#cols>Sets the display width to a constant value within the range of 80to 160 columns. This value applies only to the specific output; it isnot persistent. You can set a persistent value for thewidth=columns option with the PP_DISPLAY_WIDTH environmentvariable. This variable changes the width setting from 80 (thedefault) to the value you specify.



rpowermt display latency: Monitor path latency 129

130


Example Sample output is shown below:

Symmetrix logical device count=2==============================================================================----- Host Bus Adapters --------- ------ Storage System ---- - Latency (us) -### HW Path ID Interface Current Max==============================================================================

1 vmhba3 000190100286 FA 9aA 3565 211211 vmhba3 000190100286 FA 9bA 3565 211212 vmhba4 000190100286 FA 9aA 5136 126952 vmhba4 000190100286 FA 9bA 5136 12695

CLARiiON logical device count=3==============================================================================----- Host Bus Adapters --------- ------ Storage System ---- - Latency (us) -### HW Path ID Interface Current Max==============================================================================

1 vmhba3 HK190807490051 SP A4 3565 211211 vmhba3 HK190807490051 SP B4 3565 211212 vmhba4 HK190807490051 SP A4 5136 126952 vmhba4 HK190807490051 SP B4 5136 12695

Table 10 describes the fields in the latency display.

Table 10 Fields in the latency display



Non-negative integer PowerPath/VE number for the HBA. This number is preserved after configurationchanges but is not preserved across reboots.


Alphanumeric string ESX-assigned name of the HBA.

Storage SystemID

Hexadecimal value oralphanumeric string

Identification number for the storage system on which the logical device is located.

Storage SystemInterface

Alphanumeric string Storage-system interface.For Symmetrix systems, this has three parts:• Interface type: Fibre Channel (FA) or SCSI (SA)• Interface address: integer in the range 1 to 16• Interface port: [abcd][AB]For CLARiiON and VNX OE systems, this has one part:• Interface port: SP [A-B][0-15]

Latency Current Non-negative integer In miliseconds, the amount of time it took the last I/O on the path to complete.

Latency Max Non-negative integer In miliseconds, the longest time it took any I/O on the path to complete.



rpowermt display options: Displays option settings forstorage-system classes

Description rpowermt display options displays the default storage-system classand the following option settings for each storage system class:

◆ Whether path latency monitoring is enabled, and the path latencythreshold, if one is set

◆ Whether CLARiiON and VNX OE user-assignable LUN namesare displayed

◆ Whether periodic autorestore is enabled

Syntax rpowermt display options host=<hostname>


Example Example output is shown below:

Default storage system class: allShow CLARiiON LUN names: true

Path Latency Monitor: On

Path Latency Threshold: 0 Sec.

StorageSystem Class Attributes------------ ----------Symmetrix periodic autorestore = on

CLARiiON periodic autorestore = on

Invista periodic autorestore = on

Hitachi periodic autorestore = on

HP xp periodic autorestore = on

Ess periodic autorestore = on

Generic periodic autorestore = on

rpowermt display options: Displays option settings for storage-system classes 131

132


rpowermt display paths: Display path information

Description rpowermt display paths displays path-specific information.

Syntax rpowermt display paths [class=<classname>|all][every=<#seconds>] [width=<#col>] host=<hostname>

Argumentsclass=all|symm|clariion|invista|hitachi|hpxp|ess|generic

Limits the command to the specified type of storage system. VNXdevices are managed under the clariion storage class. VNXedevices are managed under the generic storage class. VPLEXdevices are managed under the invista storage class. allspecifies all storage-system types. The default is all.

If class is specified along with dev, the command is applied topaths that meet all specified constraints.

every=<#seconds>Specifies the frequency with which port information displays. If avalue is not specified, then the information is displayed once. Thevalue can be an integer in the range of 1 to 86400.


width=<#col>Sets the display width to a constant value within the range of 80to 160 columns. This value applies only to the specific output; it isnot persistent. You can set a persistent value for thewidth=columns option with the PP_DISPLAY_WIDTH environmentvariable. This variable changes the width setting from 80 (thedefault) to the value you specify.





Example With no arguments, the display shows path information for eachstorage class:

# rpowermt display paths host=111.222.222.111

Symmetrix logical device count=2==============================================================================----- Host Bus Adapters --------- ------ Storage System ----- - I/O Paths -### HW Path ID Interface Total Dead==============================================================================


CLARiiON logical device count=3==============================================================================----- Host Bus Adapters --------- ------ Storage System ----- - I/O Paths -### HW Path ID Interface Total Dead==============================================================================

1 vmhba3 HK190807490051 SP A4 3 01 vmhba3 HK190807490051 SP B4 3 02 vmhba4 HK190807490051 SP A4 3 02 vmhba4 HK190807490051 SP B4 3 0

Table 11 explains the fields in the path display.

Table 11 Fields in the path display (page 1 of 2)


logical devicecount

Non-negativeinteger

Total number of unique logical devices from all storage devices of a given type that areconfigured by PowerPath/VE and that this host can access.Each rpowermt display table contains information about one type of storage system,such as Symmetrix.


Non-negativeinteger

PowerPath/VE number for the HBA.When the dev option is used, the output of rpowermt display identifies the HBA bythis same HBA number.This number is preserved after configuration changes but is not preserved across reboots.


Alphanumericstring


Storage SystemID



rpowermt display paths: Display path information 133

134


With a valid storage class, the display shows path information for thespecified storage class:

# rpowermt display paths class=symm host=111.222.222.111

Symmetrix logical device count=2==============================================================================----- Host Bus Adapters --------- ------ Storage System ----- - I/O Paths -### HW Path ID Interface Total Dead==============================================================================



Alphanumericstring










rpowermt display ports: Display port information

Description rpowermt display ports displays port-specific information.

Syntax rpowermt display ports [dev=<devname>|all][class=<classname>|all] [every=<#seconds>][width=<#col>] host=<hostname>

Arguments dev=<devname>|allDisplays port information for the specified device or for alldevices.



If class is specified along with dev, the command is applied topaths that meet all specified constraints.

every=<#seconds>Specifies the frequency with which port information displays. If avalue is not specified, then the information is displayed once. Thevalue can be an integer in the range of 1 to 86400.


width=<#cols>Sets the display width to a constant value within the range of 80to 160 columns. This value applies only to the specific output; it isnot persistent. You can set a persistent value for the width=#colsoption with the PP_DISPLAY_WIDTH environment variable. Thisvariable changes the width setting from 80 (the default) to thevalue you specify.


rpowermt display ports: Display port information 135

136





Example With no arguments, the display shows port information for eachstorage class:

# rpowermt display ports host=111.222.222.111

Storage class = Symmetrix==============================================================================----------- Storage System --------------- -- I/O Paths -- --- Stats ---

ID Interface Wt_Q Total Dead Q-IOs Errors==============================================================================

000190100286 FA 9aA 256 4 0 1 0000190100286 FA 9bA 256 4 0 0 0

Storage class = CLARiiON==============================================================================----------- Storage System --------------- -- I/O Paths -- --- Stats ---

ID Interface Wt_Q Total Dead Q-IOs Errors==============================================================================

HK190807490051 SP A4 256 6 0 0 0HK190807490051 SP B4 256 6 0 1 0

Table 12 explains the fields in the path display.



Storage SystemID




Alphanumericstring






rpowermt display ports: Display port information 137

138



Total number of I/O operations under way to this path. This is the total number of I/Orequests to this device that have not completed. The sum of in-progress I/Os for all pathsshould equal the number of in-progress I/Os for the PowerPath/VE device.

Stats Errors Non-negativenumber






rpowermt help: Display rpowermt command syntax

Description rpowermt help displays a summary of the command syntax. If aspecific command is specified, it displays the syntax for thatcommand.

The usage menu is also displayed when invalid commands or optionsare entered.

Syntax rpowermt help [command]

Arguments commandA valid rpowermt command.

Example Example output when no command is specified is shown below:

Usage:rpowermt <command>

[class=all|symm|clariion|invista|hitachi|hpxp|ess|generic]host=<hostname>rpowermt help [<command>]rpowermt check_registrationrpowermt configrpowermt display [dev=<devices>|all] [class=<classname>|all]

[every=<#seconds>] [width=<#col>]rpowermt display latency [dev=<devices>|all] [class=<classname>|all]

[every=<#seconds>] [width=<#col>]rpowermt display optionsrpowermt display paths [class=<classname>|all] [every=<#seconds>]

[width=<#col>]rpowermt display ports [dev=<devices>|all] [class=<classname>|all]

[every=<#seconds>] [width=<#col>]rpowermt register [force]rpowermt restore [hba=<hba#>|all] [dev=<device>|all]

[class=<classname>|all]rpowermt set path_latency_monitor=on|off [force]rpowermt set path_latency_threshold=<#seconds> [force]rpowermt set periodic_autorestore=on|off [class=<classname>|all]

[force]rpowermt set policy=<ad|co|lb|li|rr|so|si>

[dev=<devices>|all] [class=<classname>|all] [force]rpowermt set streamio_threshold=<threshold_count>

[dev=<device|all>] [class=<class|all>]rpowermt set mode=[active|standby][hba=<hba#>|all]

[dev=<device>|all] [class=<classname>|all] [force]rpowermt setup add_host [host_file=<filename>] host=<hostname>username=<username> [password=<password>]

rpowermt setup update_host [host_file=<filename>] host=<hostname>username=<username>

rpowermt setup remove_host [host_file=<filename>] host=<hostname>

rpowermt help: Display rpowermt command syntax 139

140


rpowermt setup list_hosts [host_file=<filename>]rpowermt update lun_namesrpowermt unregister [force]rpowermt version

License search path:/etc/emc:/etc/emc/licenses:/opt/EMCpower:/opt/EMCpower/licensesHost file: /etc/emc/lockbox/root.clb

Example output when a command (in this example, display) isspecified is shown below:

# rpowermt help display

display - Show multipathing information.

Usage:rpowermt display

[dev=<devname>|all] [class=<classname>|all][every=<#seconds>] [width=<#col>] host=<hostname>

rpowermt display latency[dev=<devname>|all] [class=<classname>|all][every=<#seconds>] [width=<#col>] host=<hostname>

rpowermt display options host=<hostname>rpowermt display paths

[class=<classname>|all] [every=<#seconds>][width=<#col>] host=<hostname>

rpowermt display ports[dev=<devname>|all>] [class=<classname>|all][every=<#seconds>] [width=<#col>] host=<hostname>

License search path:/etc/emc:/etc/emc/licenses:/opt/EMCpower:/opt/EMCpower/licensesHost file: /etc/emc/lockbox/root.clb



rpowermt register: Register a PowerPath/VE license on an ESXhost

Description rpowermt register registers a PowerPath/VE license for an ESXhost. “Task 3: Installing and configuring PowerPath/VE licenses” onpage 53 provides more information.

Syntax rpowermt register [force] host=<hostname>

Arguments forceIf a previous rpowermt unregister was not able to completesuccessfully, use the force argument to re-run the command. Theuse of the force option bypasses checking for multiple usersmanaging the ESX host from different rpowermt servers.Therefore, caution must be exercised when specifying this option.


Example For an ESX host that has a valid license available, the followingconfirmation is displayed:

PowerPath license is registered.

rpowermt register: Register a PowerPath/VE license on an ESX host 141

142


rpowermt restore: Restore paths

Description rpowermt restore tests and restores specified paths. It issues testI/Os and responds to the test results as follows:

◆ If a live path passes the test, rpowermt restore does nothing.

◆ If a dead path passes the test, rpowermt restore marks it alive;PowerPath/VE can now use it. In addition, other dead paths onthe same HBA and/or storage system port may be tested.

◆ If a live path fails the test, rpowermt restore marks it dead andprints a warning (every time the path fails the test). In addition,other paths that share the same HBA and port may be markeddead, and other paths that share only the HBA or only the portmay be tested.

◆ If a dead path fails the test, rpowermt restore prints a warning(every time the path fails the test).

There may be a delay in accessing a recovered path. To avoid thisdelay, run rpowermt restore after a path is physically restored.

In addition to testing and restoring paths, rpowermt restore attemptsto resurrect dead volumes. A volume may be marked dead if writeerrors occur that could jeopardize the integrity of the data structures,and if subsequent writes could aggravate the problem.

When applied to CLARiiON and VNX OE storage systems, rpowermtrestore also relocates LUNs to their default storage processor.

Syntax rpowermt restore [hba=<hba#>|all] [dev=<device>|all][class=<class>|all] host=<hostname>



Argumentsclass=all|symm|clariion|invista|hitachi|hpxp|ess|generic


If class is specified along with hba and/or dev, the command isapplied to paths that meet all specified constraints.

hba=<hba#>|allLimits restoration to paths from the specified HBA. hba# is anumber in the Host Bus Adapters ### column of thepowermt display dev output. all specifies all HBAs underPowerPath/VE control. The default is all.

dev=<device>|allLimits restoration to the specified path, or all paths to thespecified device. all specifies all paths to all devices. The defaultis all.



Example If there are new dead paths, a warning message indicates each deadpath:

Warning: Device emcpower11 bus 1 path C0:T3:L1 is currently dead.Warning: Device emcpower12 bus 1 path C0:T3:L2 is currently dead.Warning: Device emcpower13 bus 1 path C0:T3:L0 is currently dead.

rpowermt restore returns no output if there are no new dead paths.

rpowermt restore: Restore paths 143

144


rpowermt set mode: Set paths to active or standby mode

Description rpowermt set mode sets paths to active or standby mode.

For most applications, the best performance is achieved bydesignating all paths to a PowerPath/VE device as active. By usingthis command selectively, however, the path usage of devices can becontrolled. This is helpful if you do not want I/O for one device toaffect the performance of another device.

Syntax rpowermt set mode=[active|standby] [hba=<hba#>|all][dev=<device>|all] [class=<classname>|all][force] host=<hostname>

Arguments mode=active|standby

Note: PowerPath/VE does not support the standby mode with iSCSIsoftware initiators.

Determines whether a path is available for I/O or held in reserve.In active mode, user I/O is delivered to a path. If you designatea path as active, it is continuously scheduled for I/O according tothe load-balancing and failover policy in use. In standby mode,the path is held in reserve.

Being set to standby does not mean a path will not be used. Itonly means the weight of the path is heavily adjusted to precludeits use in normal operations. A standby path still can be selected ifit is the best path for a request. For example, if the active paths areso heavily loaded that the weighting factor in favor of the activepath is overcome, the I/O will be sent to the standby path.

Mode settings are not persisted across reboot.

In PowerPath/VE 5.4 SP2, when you disable a path to aPowerPath device using vSphere Client, that path mode changesfrom active to standby in rpowermt display. On enabling the pathagain, the mode goes back to active when displayed usingrpowermt. This is expected PowerPath behavior.

class=all|symm|clariion|invista|hitachi|hpxp|ess|generic




If class is specified along with hba and/or dev, the command isapplied to paths that meet all specified constraints.

hba=<hba#>|allSets the mode for paths from the specified HBA. hba# is a numberin the Host Bus Adapters ### column of powermt display devand powermt display paths output. all specifies all HBAsunder PowerPath/VE control. The default is all.

dev=<devname>|allSets the mode for all paths to the specified device. all specifiesall paths to all devices. The default is all.


The mode can be set to an individual path specifying the uniquepath identifier. The path identifier can be found using a displaycommand (for example, vmhba2:C0:T5:L11).

forceIf a previous rpowermt set mode command was not able tocomplete successfully, use the force argument to re-run thecommand. The use of the force option bypasses checking formultiple users managing the ESX host from different rpowermtservers. Therefore, caution must be exercised when specifyingthis option.


Example

To set the mode to an individual path, determine the path that youwant to set, and then use the following command:

rpowermt set mode=<active/standby> dev=<path_identifier> host=<hostname>

rpowermt set mode: Set paths to active or standby mode 145

146


Use the display command to find the path identifier. The displayoutput for a device may appear as:

Pseudo name=emcpower6HP xp ID=25158Logical device ID=R500 00025158000582Standard UID=naa.60060e80056246000000624600000246state=alive; policy=StreamIo; streamio_threshold=128; queued-IOs=0==============================================================================--------------- Host --------------- - Stor - -- I/O Path -- -- Stats ---### HW Path I/O Paths Interf. Mode State Q-IOs Errors==============================================================================

2 vmhba7 C0:T2:L6 AF active alive 0 01 vmhba6 C0:T4:L6 9F active alive 0 02 vmhba7 C0:T3:L6 9F active alive 0 01 vmhba6 C0:T2:L6 AF active alive 0 0

To disable the first path, the unique path identifier would bevmhba7:C0:T2:L6

The resulting set mode command would be:

rpowermt set mode=<active/standby> dev= vmhba7:C0:T2:L6 host=<hostname>



rpowermt set path_latency_monitor: Enables or disables pathlatency monitoring

Description rpowermt set path_latency_monitor enables or disables pathlatency monitoring. When path latency monitoring is enabled,PowerPath/VE monitors the time it takes for I/O requests tocomplete on all paths in your environment. When you enable ordisable path latency monitoring, the setting applies globally to allpaths.

You can check whether path latency monitoring is enabled byviewing the rpowermt display options output. You can also displaythe last and longest I/O completion times for each path in yourenvironment by entering the rpowermt display latency command.

The diagnostic information reported by latency monitoring allowsyou to determine normal I/O times for your environment, which canhelp determine an appropriate path latency threshold. To set a pathlatency threshold, use the rpowermt set path_latency_thresholdcommand.

Note: To avoid extraneous log messages, EMC recommends setting athreshold value before turning on latency monitoring.

Syntax rpowermt set path_latency_monitor=on|off [force]host=<hostname>

Arguments path_latency_monitor=on|offEnables (on) or disables (off) path latency monitoring. Thedefault is off.

forceIf a previous rpowermt set path_latency_monitor commandwas not able to complete successfully, use the force argument tore-run the command. The use of the force option bypasseschecking for multiple users managing the ESX host from differentrpowermt servers. Therefore, caution must be exercised whenspecifying this option.


Example rpowermt set path_latency_monitor does not return any output.

rpowermt set path_latency_monitor: Enables or disables path latency monitoring 147

148


rpowermt set path_latency_threshold: Sets a time limit for thecompletion of I/O

Description rpowermt set path_latency_threshold sets a time interval inseconds within which I/Os should complete. When an I/O takeslonger to complete than the threshold value, PowerPath/VE sends amessage to the log file. The threshold value applies to all paths inyour environment.

For example, if you set the path_latency_threshold value to 30seconds, and an I/O request on one path takes 35 seconds tocomplete, a message describing the event is sent to the log file.

Note: To avoid extraneous log messages, EMC recommends setting athreshold value before enabling path latency monitoring. Additionally, whenchanging a threshold value, disable path latency monitoring, set the newthreshold value, and then re-enable path latency monitoring.

To enable path latency monitoring, use the rpowermt setpath_latency_monitor command. To view information on I/Ocompletion times, use the rpowermt display latency command.

Log message exampleWhen an I/O completion time exceeds the threshold value,PowerPath/VE sends to the log file a message like the following:

Mar 30 12:32:40 abc01.lss.emc.com emcp: [ID 801593 kern.notice]:Mpx:Latency Threshold exceeded by UserDev vmhba4:C0:T2:L0 Vol60060160E1601E006CF56548A2FBDC11

Syntax rpowermt set path_latency_threshold=<#seconds>[force] host=<hostname>

Arguments path_latency_threshold=<#seconds>Sets a time interval in seconds within which I/Os shouldcomplete. The value applies to all paths. The default value is 0.The range of acceptable values is 0 to 2,147,483,647.

forceIf a previous rpowermt set path_latency_threshold commandwas not able to complete successfully, use the force argument tore-run the command. The use of the force option bypasseschecking for multiple users managing the ESX host from differentrpowermt servers. Therefore, caution must be exercised whenspecifying this option.




Example rpowermt set path_latency_threshold does not return any output.

rpowermt set path_latency_threshold: Sets a time limit for the completion of I/O 149

150


rpowermt set periodic_autorestore: Enable or disable periodicautorestore

Description rpowermt set periodic_autorestore enables or disables thePowerPath/VE periodic autorestore facility.

When periodic autorestore is on, PowerPath/VE periodically testsdead paths and, if they pass the test, restores them to service. Evenwhen periodic autorestore is off, path testing continues to be doneunder certain conditions, and automatic path restoration continues tooccur based on the results of that testing.

To determine the periodic autorestore setting, use rpowermt displayoptions.

Syntax rpowermt set periodic_autorestore=on|off[class=<classname>|all] [force] host=<hostname>

Arguments periodic_autorestore=on|offEnables (on) or disables (off) periodic autorestore. The default ison.


forceIf a previous rpowermt set periodic_autorestore commandwas not able to complete successfully, use the force argument tore-run the command. The use of the force option bypasseschecking for multiple users managing the ESX host from differentrpowermt servers. Therefore, caution must be exercised whenspecifying this option.


Example rpowermt set periodic_autorestore does not return any output.



rpowermt set policy: Set the load-balancing and failover policy

Description rpowermt set policy sets the load-balancing and failover policy fordevices managed by PowerPath/VE.

Note: Between the time that PowerPath/VE is installed on an ESX host andthe time that the rpowermt register command is run to register aPowerPath/VE license for the ESX host, the policy for Symmetrix,CLARiiON, VNX OE, Invista, and VPLEX devices is bf (Basic Failover), andfor supported Celerra and supported third-party devices is nr (No Redirect).This is the only time that adevice is set to bf or nr, which are the interimpolicies in effect only while an ESX host is in a pre-licensed state. I/O is notsupported while a device is in the bf or nr state. Once a PowerPath/VElicense is registered for the ESX host, the policy transitions to so (SymmetrixOptimized), co (CLARiiON Optimized), or ad (Adaptive), depending on thedevice type. I/O is now supported.

bf and nr are not supported as an argument to the rpowermt set policycommand. You cannot manually set the policy for a device to bf or nr.

Syntax rpowermt set policy=<ad|co|lb|li|rr|so|si>[dev=<devices>|all] [class=<classname>|all][force] host=<hostname>

Arguments policy=ad|co|lb|li|rr|so|siSets the policy to one of the following values:

Policy Description

ad (Adaptive) I/O requests are assigned to paths based on an algorithm that takesinto account path load.Default for EMC Invista and VPLEX, supported Celerra devices, andsupported third-party array devices.

co (CLARiiONoptimization)

I/O requests are assigned to paths based on an algorithm that takesinto account path load.This policy is valid only for CLARiiON and VNX storage systems andis the default policy for them.

lb (Least blocks) Load balance is based on the number of blocks in pending I/Os. I/Orequests are routed to the path with the fewest queued blocks,regardless of the number of requests involved.

li (Least I/O) Load balance is based on the number of pending I/Os. I/O requestsare routed to the path with the fewest queued requests, regardless oftotal block volume.

rpowermt set policy: Set the load-balancing and failover policy 151

152


dev=<devname>|allLimits the policy change to the specified devices.


class=all|symm|clariion|invista|hitachi|hpxp|ess|genericLimits the command to the specified type of storage system. allspecifies all storage-system types. VNX devices are managedunder the clariion storage class. VNXe devices are managedunder the generic storage class. VPLEX devices are managedunder the invista storage class. The default is all. If class isspecified along with dev, the command is applied to devices thatmeet all specified constraints.

forceIf a previous rpowermt set policy command was not able tocomplete successfully, use the force argument to re-run thecommand. The use of the force option bypasses checking formultiple users managing the ESX host from different rpowermtservers. Therefore, caution must be exercised when specifyingthis option.


Example rpowermt set policy does not return any output.

rr (Round robin) I/O requests are assigned to each available path in rotation.

si (StreamIO) For each possible path for an I/O to a particular volume, this policyselects the same path as was selected for the previous I/O to thevolume, unless the volume I/O count since the last path changeexceeds the volume’s threshold value. When the threshold isexceeded, the policy selects a new path based on the adaptive policyalgorithm.The volume I/O count is rezeroed on each path change.

so (Symmetrixoptimization)

I/O requests are routed to paths based on an algorithm that takes intoaccount path load. Load is a function of the number, size, and type ofI/O queued on each path.This policy is valid only for Symmetrix storage systems and is thedefault policy for them.

Policy Description



rpowermt set streamio_threshold: Set I/O threshold countpowermt set streamio_threshold sets the I/O threshold values forwhen to switch paths on devices where the StreamIO (si)load-balancing policy is in effect on PowerPath/VE. This is onlysupported on PowerPath 5.4 SP2 running on ESX 4.1 or later.

The streamio value is accepted only if the load-balancing policy ineffect on the device is si. If you specify a device for which theload-balancing policy is not si, this will generate an error message.

The value of the streamio threshold is truncated to a power of 2.Allowable values are 64, 128, 256, 512, 1024, and 2048. If a valueoutside the range is given, an error is returned. If a value within therange is given, the value is truncated down to a power of 2 (1025would become 1024). If the PowerPath license is invalid, setting thestreamio_threshold will generate an error message.

Syntax rpowermt set streamio_threshold=<threshold_count>[dev=device|all] host=<hostname>

Arguments streamio_threshold=threshold_countInteger in the range from 64 to 2048 that specifies, in number ofI/Os, when to switch paths. The default for EMC arrays is 1024and the default for third-party arrays is 128.

dev=<devname>|allLimits the change to the specified PowerPath device. allspecifies all devices. The default is all.


rpowermt set streamio_threshold: Set I/O threshold count 153

154


Example The parameter is only visible during a display command whenthe policy is si.

Pseudo name=emcpower4CLARiiON ID=HK190807490051 []Standard UID=naa.60060160e1601e005cc251d1a1fbdc11 []state=alive; policy=StreamIo; streamio_threshold=1024; queued-IOs=0Owner: default=SP A, current=SP A Array failover mode: 4==============================================================================---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---### HW Path I/O Paths Interf. Mode State Q-IOs Errors==============================================================================

1 vmhba3 C0:T2:L2 SP A4 active alive 0 01 vmhba3 C0:T3:L2 SP B4 active alive 0 02 vmhba4 C0:T2:L2 SP A4 active alive 0 02 vmhba4 C0:T3:L2 SP B4 active alive 0 0

You can check whether streamio threshold is set by viewing therpowermt display dev output.

state=alive; policy=StreamIo;streamio_threshold=1024; queued-IOs=0



rpowermt setup add_hostrpowermt setup add_host adds an ESX host and its correspondingusername/password pair to the specified lockbox (if the host_fileargument is specified) or to the default lockbox on the rpowermtserver.

Syntax rpowermt setup add_host [host_file=<filename>]host=<hostname> username=<username>[password=<password>]

Arguments host_file=<filename>Lockbox file on the rpowermt server. If the specified lockbox doesnot exist, it is created.

If the host_file argument is omitted, the ESX host and itsusername/password pair are added to the default lockbox.


username=<username>User account on the ESX host.

password=<password>Password associated with the specified username. If thepassword is not included on the command line, you areprompted to supply it.

Example If the specified lockbox does not exist, you are prompted for the ESXhost password (if not specified on command line) and new lockboxpassphrase. The lockbox is created.

Enter lockbox passphrase:Confirm passphrase:Enter server password:Confirm password:

If the specified lockbox exists, but there is no entry for specified host,you are prompted for the ESX host’s password (if not specified oncommand line). The username/password entry is added to thelockbox.

# rpowermt setup add_host host=111.222.222.111 username=root


rpowermt setup add_host 155

156


If the specified lockbox exists and it has an entry for specified host,you are prompted for host’s password (if not specified on commandline). The username/password entry is updated.

# rpowermt setup add_host host=111.222.222.111 username=root




rpowermt setup list_hostsrpowermt setup list_hosts lists each ESX host that has ausername/password entry in the specified lockbox (if the host_fileargument is specified) or in the default lockbox on the rpowermtserver.

Syntax rpowermt setup list_hosts [host_file=<filename>]

Arguments host_file=<filename>List the ESX host in the specified lockbox on the rpowermt server.If this argument is omitted, the ESX hosts in the default lockboxare listed.

Example If the specified lockbox exists and it has stored entries, a list of hostIPs are displayed:

# rpowermt setup list_hosts

111.222.222.111111.222.222.112

rpowermt setup list_hosts 157

158


rpowermt setup remove_hostrpowermt setup remove_host removes an ESX host and itsusername/password entry from the specified lockbox (if thehost_file argument is specified) or from the default lockbox on therpowermt server.

Syntax rpowermt setup remove_host [host_file=<filename>]host=<hostname>

Arguments host_file=<filename>Remove the ESX host from the specified lockbox on the rpowermtserver. If this argument is omitted, the ESX host is removed fromthe default lockbox.


Example If the specified lockbox exists and it has an entry for specified ESXhost, no output is returned.



rpowermt setup update_hostrpowermt setup update_host updates the specified ESX host’susername/password entry in the specified lockbox (if the host_fileargument is specified) or in the default lockbox on the rpowermtserver.

Syntax rpowermt setup update_host [host_file=<filename>]host=<hostname> username=<username>[password=<password>]

Arguments host_file=<filename>Lockbox file on the rpowermt server. If the specified lockbox doesnot exist, it is created. If the host_file argument is omitted, theESX host and its username/password pair are updated in thedefault lockbox.


username=<username>User account on the ESX host.

password=<password>Password associated with the specified username. If thepassword is not included on the command line, you areprompted to supply it.

Examples If the specified lockbox exists and it has an entry for specified hostexists, you are prompted for ESX host’s password (if not specified oncommand line). The host entry is updated.

# rpowermt setup update_host host=111.222.222.111 username=root


If the specified lockbox exists but it does not have an entry forspecified ESX host, you are prompted for the ESX host’s password (ifnot specified on command line). The host entry is added.

# rpowermt setup update_host host=111.222.222.111 username=root


rpowermt setup update_host 159

160


rpowermt unregister: Remove a PowerPath/VE license from anESX host

Description rpowermt unregister removes a PowerPath/VE license for an ESXhost.

Syntax rpowermt unregister [force] host=<hostname>

Arguments forceIf a previous rpowermt unregister was not able to completesuccessfully, use the force argument to attempt to release the licenseauthorization. The use of the force option bypasses checking formultiple users managing the ESX host from different rpowermtservers. Therefore, caution must be exercised when specifying thisoption.


Example For an ESX host with a current valid license:

Do you really want to unregister the PowerPath license?yes/[no]: yes

PowerPath license is unregistered.

For an ESX host that does not have a valid license:

Warning: PowerPath driver is already unlicensed.



rpowermt update lun_names: Retrieve CLARiiON nice names

Description rpowermt update lun_names retrieves the latest CLARiiON nicenames. A CLARiiON nice name, also called a user-assignable LUNname, is a character string associated with a logical device on aCLARiiON and VNX OE array and assigned through Navisphere orUnisphere. Run this command to perform initial discovery ofCLARiiON nice names and whenever the nice name is changed onthe CLARiiON array through Navisphere or Unisphere.

Note: CLARiiON LUN nice names are updated automatically every hour.Use this command to manually update CLARiiON nice names.

Syntax rpowermt update lun_names host=<hostname>


Example CLARiiON nice names are disabled in the following rpowermt

display dev output:

Pseudo name=emcpower2CLARiiON ID=HK190807490007 []Standard UID=naa.60060160bab01e00930a8cdde601dd11 []

CLARiiON nice names are enabled in the following rpowermt

display dev output:

Pseudo name=emcpower2CLARiiON ID=HK190807490007 [lclb164]Standard UID=naa.60060160bab01e00930a8cdde601dd11 [LUN 4]

Note the storage group name (lclb164) and the device name (LUN 4)are now displayed.

rpowermt update lun_names: Retrieve CLARiiON nice names 161

162


rpowermt version: Display PowerPath/VE version number

Description rpowermt version displays the version of PowerPath/VE installedon the ESX host, the license search path, and the host file name andlocation.

Syntax rpowermt version host=<hostname>


Example Example output is shown below:

EMC rpowermt for PowerPath (c) client Version 5.4 (build 254)EMC PowerPath (c) host=lcla226 Version 5.4 (build 243)

License search path:/etc/emc:/etc/emc/licenses:/opt/EMCpower:/opt/EMCpower/licenses

Host file: /etc/emc/lockbox/root.clb


C

This appendix lists messages returned by the PowerPath/VE driver,PowerPath/VE installation process, rpowermt utility, and otherPowerPath/VE utilities, and provides suggested actions.

◆ Error log messages........................................................................... 164◆ VM kernel errors .............................................................................. 165◆ PowerPath/VE error messages...................................................... 167

PowerPath/VEMessages

PowerPath/VE Messages 163

164

PowerPath/VE Messages

Error log messagesError log messages appear when you configure PowerPath/VE forcommon logging. The error log messages capture unexpected eventsthat occur. Some error log messages convey information that appearson screen when a command fails. Some messages are logged to one ofthe following locations, depending on what you are debugging:

◆ VM kernel log file

◆ CIM log file

◆ ESX update log file

The messages in this appendix are logged to the VM kernel log file.



VM kernel errorsThis section discusses the operations associated with VM kernelerrors and the error message sources.

Operation Error messages written to the VM kernel log files can have to do withspecific operations. The Operation column of Table 13 on page 167lists the following main operation categories:

◆ Device—These messages might occur, among other reasons,ifthere are problems with device availability; for example, if adevice could not be found or opened or if it is being quiesced.

◆ Module load—These messages might occur, among otherreasons, if there is any problem with the license; for example, if itis expired, invalid, or cannot be located. These messages mightalso occur if there is insufficient memory on the system to loadPowerPath/VE.

◆ Path claim/path unclaim—These messages might occur, amongother reasons, if there is a problem with path availability forPowerPath/VE.

◆ rpowermt display—These messages might occur, among otherreasons, as a result of user initiated commands from rpowermt.

◆ rpowermt register—These messages would occur if there is anyproblem with the license; for example, if it is expired, invalid, orcannot be located.

◆ SCSI command I/O on device—These messages might occur,among other reasons, due to I/O failure on devices.

◆ Set path state—These messages might occur, among otherreasons, if there is low availability of active paths.

Message sources Error messages written to the VM kernel log files can be generated byseveral commands and utilities. The message could come from thefollowing message sources:

◆ PowerPath/VE driver—Usually there are driver-related errormessages associated with PowerPath/VE messages. For instance,PowerPath/VE marks a path dead if the path test fails, and a pathtest is done as a result of an I/O failure. These failed I/Os almostalways are logged by the HBA driver. These errors can beassociated by time and device with PowerPath/VE errors.

VM kernel errors 165

166


◆ PowerPath/VE installation—These messages can appear whilePowerPath/VE is being installed.

◆ rpowermt and other PowerPath/VE utilities—If a message isgenerated by a specific rpowermt command, that command islisted.



Op

De

De

De

De

De

De

Gefor

Ge

Ge

Inq

Mo

Inq

PowerPath/VE error messagesPowerPath/VE error messages are listed in Table 13.

Table 1 PowerPath/VE error messages (page 1 of 7)

eration Message type Message Explanation Suggested action

vice dump Error No active path to thedevice found.

No active path to the dumpdevice was found.

Check the paths to storagearray that has the dumpdevice.

vice open Error Volume is beingdestroyed. Cannot open.

Device could not be opened. None. PowerPath/VE willautomatically retry thecommand.

vice open Error Volume to be opened notfound

Device could not be opened. None. PowerPath/VE willautomatically retry thecommand.

vice probe Error Volume not found SCSI device could not befound.

None. PowerPath/VE willautomatically retry thecommand.

vice probe Information Wait for volume to beunquiesced

Device is currentlyquiesced; waiting for it to beunquiesced.


vice probe Error Probe of path path failed.Status : status

Path probe has failed. None. PowerPath/VE willautomatically retry thecommand.

neral (Possible errormost commands)

Error Volume not found Device could not be found. None. PowerPath/VE willautomatically retry thecommand.

t device name Information Device name name Print device name. None.

t path names Information Number of Paths =<number>

Print path names to adevice.

None.

uiry Information page 0 rc=status VPDP page information. None.

dule load warning Could not initialize logname: status

Initializing log handle namefailed.

PowerPath/VE load failed.Check if enough memory isavailable.

uiry Error Inquiry failed. Status:status

Inquiry failed. Check if the device isaccessible.

PowerPath/VE error messages 167

168


Mo

Mo

Mo

Mo

Mo

Mo

Mopat

Mo

Mo

Pa

Op

dule load Warning vmk_LogRegister failed:status

PowerPath/VE failed toregister log handle.


dule load Information Char device namecreated

PowerPath/VE characterdevice created.

None.

dule load Error Failed to create Chardevice name. Status :status

PowerPath/VE characterdevice could not be created.


dule load Information System Id sysID,Reservation Key key

Print system I andreservation key information.

None.

dule load Information License data does notmatch host data.PowerPath is notlicensed.

License information inpersistent string does notmatch host hardware data.

Persisted license data is notvalid on this host. Add a newPowerPath/VE License. ThePowerPath/VE for VMwareLicensing Guide providesinformation.

dule load Information Persistence of SCSI Resfail Value is 0, generatingnew one

Generating new reservationkey.

None.

dule Load/ Periodich test

Warning PowerPath is notlicensed.

PowerPath/VE is notlicensed.

Register yourPowerPath/VEhost. The PowerPath/VE forVMware Licensing Guideprovides information.

dule load/daily Warning PowerPath licenseexpires in number of daysto expire days

Warning that PowerPath/VElicense is going to expire.

Check license information;contact your EMC AccountRepresentative.

dule load/daily Warning PowerPath should bemanaged to ensurelicense serverconsistency

Warning that PowerPath/VElicense is served-linger andpermanent.

Check license information;contact your EMC AccountRepresentative.

th claim Information Claiming path path. Information about the paththat is going to be claimed.

None.





Pa

Pa

Pa

Pa

Pa

Pa

Pa

Pa

Pa

Op

th claim Error The Path has alreadybeen claimed

Path has already beenclaimed by anothermultipathing plugin.

Unclaim this path from theother multipathing pluginbefore trying to claim thepath using PowerPath/VE.The PowerPath/VE forVMware vSphereInstallation andAdministration Guideprovides information.

th claim Error MpxRecognize failed. Noresources- check systemmaximum path limits.

Maximum allowed path limithas already been reached.

Verify that the number ofdevices configured on thissetup is fewer than themaximum supporteddevice/path limit.

th claim Information No standard UID. Usingthe WWN as volume UID

Standard UUDI for devicecould not be found.

None.

th claim Error MpxRecognize failed PowerPath/VE failed toclaim path.


th claim Error Could not claim the path.Status : status

PowerPath/VE failed toclaim path.


th claim Error Could not create volDaemon

PowerPath/VE failed tocreate a daemon or toallocate memory.

Check if server is out ofmemory.

th claim Information Added new Bus ID busidfor adapter adapter

PowerPath/VE has claimedfirst path to an HBA.

None.

th claim Information PowerPath is notLicensed. Volume will beregistered when a validlicense is added.

A valid PowerPath/VElicense is not available andthe defaultpolicy forunlicensed devices,BasicFailover, is set. Thedevice settings cannot bechanged.

Register yourPowerPath/VEhost. The PowerPath/VE forVMware Licensing Guideprovides information.

th claim Error Could not allocate SCSIdevice

Maximum number of SCSIdevices, 256, has alreadybeen claimed.





170


Pa

Pa

Pa

Pa

Pa

Pa

Pa

Pa

Pa

Pa

Op

th claim Error Status:status. Failed toregister device. Usershould unclaim paths.

Paths to the device havebeen claimed but the devicecould not be registered. Theuser should unclaim pathsto this device.

Check if maximum numberof devices has already beenclaimed or if some paths tothis device are claimed byother multipathing pluginslike NMP. If so, unclaimpaths to this device.

th claim/unclaim Information Wait for nIosPendingcount (count) to become 0

Until I/O count becomeszero, PowerPath/VE will notadd or remove path. Thereis a delay in adding orremoving path.


th probe Error Path path has not beenclaimed

PowerPath/VE failed to findthe path.


th probe Information Path “path name” is pathstate.

Print path state. None.

th probe Information Path “path name" ischanging to new statefrom old state.

Path state is updated to newstate from an oldinconsistent state.

None.

th probe Information Probe failed. EmsStatus :status

PowerPath/VE path probehas failed.


th unclaim Information Unclaiming path path. Information about the paththat is going to beunclaimed.

None.

th unclaim Error The Path has not beenclaimed

Path being unclaimed hasnot been claimed byPowerPath/VE.

None. The path is alreadyunclaimed.

th unclaim Error The volume is open.Cannot unclaim last pathpath.

Cannot unclaim pathbecause it is the lastavailable path and is beingused.

Stop all I/O on the devicebefore trying to unclaim allpaths to the device.

th unclaim Error The volume is beingregistered orunregistered. Cannotunclaim last path path.

PowerPath/VE cannotunclaim path because this isthe last available path and isbeing used.






Pa

Pa

Pa

Pa

Pe

rpo

rpo

rpo

rpo

rpo

rpo

SCI/O

SCI/O

Op

th unclaim Error Could on unregister thedevice. Status : status

PowerPath/VE cannotunclaim path because this isthe last available path and isbeing used.

Stop all I/O on the devicebefore trying to unclaim allpaths to the device.

th unclaim Information Unclaimed path to volumedevice. Setting state toON.

PowerPath/VE successfullyunclaimed the path.

None.

th unclaim Information Removing Bus ID busidfor adapter adapter

PowerPath/VE is unclaiminglast path to an HBA.

None.

th unclaim Information Path path is dead or notan active BF path.

PowerPath/VE can removepath if the path is dead or isnot the last active path.

None

riodic path test Information Updating NN info forStSys device

Updating device NiceNameperiodically.

None.

wermt display Error Volume is not Registered Device is not registered. None. Wait for the device tobe registered.

wermt display Error Not valid vol or Mpxdevice Device is not valid. None.

wermt display Error Not a valid extensiondevice

Device is not valid. None.

wermt display Information Mpx: pData data size size Invalid ioctl buffer size. Check if kernel moduleversion matches mpapiversion.

wermt display Information Wait for volume to beunquiesced

Waiting for the device to beunquiesced before printingany information.


wermt register Error PowerPath license hasexpired

New license that is beingregistered has alreadyexpired.

Check your licenseinformation. Contact yourEMC AccountRepresentative.

SI commandon Device

Error Failed to Create Pirp PowerPath/VE failed tocreate I/O request .

Check if enough memory ispresent on the server.

SI commandon Device

Error Not a valid extensiondevice

I/O has failed. None. PowerPath/VE willautomatically retry thecommand.




172


SCI/O

SCI/O

SCI/O

SCI/O

SCI/O

SCI/O

SCI/O

SCI/O

SCI/O

SCDe

SCI/O

Op

SI commandon Device

Error Completed the platformIO because of thepowerpath error

I/O has failed. None. PowerPath/VE willautomatically retry thecommand.

SI commandon Device

Error Allocation of dummy<ASyncIoDaemon/TestDaemon/SyncIoDaemon> entryfailed. Not aborting it.

Taskmanagementoperationhas failed.


SI commandon Device

Information Enqueued from<ASyncIoDaemon/TestDaemon/SyncIoDaemon> to Abortqueue Cmd command

PowerPath/VE has aborteda command.


SI commandon Device

Information SyncIoDaemonCallout IOto Abort Cmd command



SI commandon Device

Information Host Error host statusDevice Error dev status



SI commandon Device

Warning SCSI-2 Translationcommand failed after themaximum number ofretries!

SCSI 2 translation has failedafter maximum number ofretries.


SI commandon Device

Error Unable to send cmd(cleanup PRI) to device.SCSI status status

PowerPath/VE is unable todetermine device state.


SI commandon Device

Error Unable to send cmd(cleanup PRO) to device.SCSI status status

PowerPath/VE is unable tounregister the device.

Check for deviceconnectivity and check thesystem log for SCSI status.

SI commandon Device

Error SCSI-2reservation/release failedwith SCSI status status

SCSI 2 reservation/releasehas failed.


SI command/I/O onvice

Error Path path : could notissue sync/async IOcommand: status

PowerPath/VE failed toissue asyncrynous I/O.


SI commandon Device

Error Not a valid sense buffer Not a valid sense buffer forI/O status.






SCI/O

SCDe

SCI/O

SCI/O

SCI/O

SCI/O

Se

Se

Op

SI commandon Device

Error Invalid Command Status:status

Invalid status. None. PowerPath/VE willhandle this error.

SI command/I/O onvice

Error Failed to alloc Pirp Insufficient memory on thesystem.


SI commandon Device

Error Failed to create Platformscsicmd

Insufficient memory on thesystem.


SI commandon Device

Error Failed to create Pirp Insufficient memory on thesystem.


SI commandon Device

Error Failed to Create Scsi Pirp Insufficient memory on thesystem.


SI commandon Device

Error Sense Info S:sc ASC:ascASCQ:ascq

Non-retryable error. SomeI/O failed

None. PowerPath/VE willhandle this error.

t path state Error ProbePath Failed path PowerPath/VE path probehas failed.

Check for path availability.

t path state Error Path path could not bedisabled. It is the lastworking path to volume.

PowerPath/VE failed tochange path state.

Ensure that the path thatyou are trying to disable isnot the last active path tothe volume. There must beat least one active path tothe volume.Retry the operation after anyexisting IO has beencompleted.




174



Glossary

This glossary contains terms related to disk storage subsystems.Many of these terms are used in this manual.

AAccess Logix A software package that lets multiple hosts share storage on certain

CLARiiON and VNX OE storage systems. Access Logix implementsstorage sharing using storage groups. See also ”Storage group.”

Active (paths) One of two modes for PowerPath I/O paths. The other mode isstandby. An active path can accept I/O. The load-balancing andfailover policy (set for the PowerPath device with the powermt setpolicy command) determines how loads are balanced over activepaths. Load balancing is done for each device with more than oneactive path. See also ”Mode.” and “Standby (paths).”

Active-active(storage systems)

A storage type for Symmetrix systems in which all interfaces to adevice are active simultaneously.

Active-passivee(storage systems)

A storage type for CLARiiON and VNX OE devices in which onlyone interface to a device is active at a time, and any others are passivewith respect to that device and waiting to take over if needed.

Adapter A circuit board that enables a computer to use external devices suchas a disk storage system or a high-speed network. See also ”Host busadapter (HBA).”


176

Glossary

Adaptive (ad) A load-balancing and failover policy for PowerPath devices in whichI/O requests are assigned to paths based on an algorithm that takesinto account path load and logical device priority.

Alive One of two states for PowerPath paths and logical devices. The otherstate is dead. A live path is usable: PowerPath can direct I/O to it. Alive logical device either was never marked dead by PowerPath orwas marked dead but restored with the powermt restore command.See also ”Dead.”

ALUA (asymmetriclogical unit access)

An array failover mode available on CLARiiON and VNX OEsystems with FLARE version 03.26 or later in which one arraycontroller is the active/optimized controller and the other arraycontroller is the active/non-optimized controller. I/O is directed to theactive/optimized controller. Should the active/optimized arraycontroller become unavailable or fail, I/O is directed to theactive/non-optimized array controller until a trespass occurs.

Arbitrated loop A Fibre Channel topology supported by PowerPath. An arbitratedloop topology requires a port to successfully negotiate to establish acircuit between itself and another port on the loop.

Assign In an active-passive storage system, the term used to indicate that aninterface card is designated to a device.

BBaseline A set of updates and patches that can be created and installed using

VMware Update Manager.

Basic failover (bf) A failover policy that protects against CLARiiON and VNX OE SPfailures, Symmetrix FA port failures, and back-end failures, and thatallows non-disruptive upgrades to work when running PowerPathwithout a license key. It does not protect against HBA failures. Loadbalancing is not in effect with basic failover. I/O routing on failure islimited to one HBA and one port on each storage system interface.This policy is valid for CLARiiON, VNX OE, Symmetrix, Invista, andVPLEX devices, and is the default policy for them on platformswithout a valid PowerPath license. This is the only time that a deviceis set to Basic failover. Basic failover is not supported as anargument to the rpowermt set policy command. You cannotmanually set the policy for a device to bf. See also ”Load balancing.”


Glossary

Boot device The device that contains a computer’s startup code. Symmetrixlogical devices managed by PowerPath can be configured as bootdevices.

Bus In a computer, a collection of signal lines that work together toconnect one or more modules; for example, a disk controller and thecentral processor. A bus also can connect two cooperating controllers,such as a SCSI host adapter and a SCSI device controller. See also”SCSI.”

CChannel A point-to-point data transport link.

Channel group PowerPath’s name for a communication channel directed to only onelogical device. Several paths make up a channel group. Channelgroups can increase system performance and redundancy bydedicating a set of paths to a critical application component (forexample, database log files), while maintaining access to a redundantset of paths to the application component, in case the first set fails

Claim rules Numbers used in the VMware vSphere environment to assign storagedevices to either NMP or to PowerPath.

CLARiiON optimization(co)

A load-balancing and failover policy for PowerPath devices, in whichI/O requests are assigned to paths based on an algorithm that takesinto account path load and the logical device priority you set withpowermt set policy. This policy is valid for CLARiiON and VNX OEstorage systems only and is the default policy for them, on platformswith a valid PowerPath license. It is listed in powermt display outputas CLAROpt. See also ”Load balancing.”

Cluster Two or more interconnected hosts sharing access to the same datastorage resources. If one host fails, another host can continue to makedata available to applications.

Consistency Group A group of Symmetrix devices specially configured to act in unison,to maintain the integrity of a database distributed across multipleSRDF® (Symmetrix Remote Data Facility) units. PowerPath canreport SRDF consistency group status. See also ”SRDF (SymmetrixRemote Data Facility).”


178

Glossary

Controller A device that controls and manages access to a part of a computer orcomputerized system. Examples include disk controllers oncomputers or similar controllers on disk storage systems.

DData availability Access to any and all user data by an application.

Data channel See ”Channel.”

Dead One of two states for paths and logical devices:

• A dead path is not usable: PowerPath will not direct user I/O tothis path. PowerPath marks a path dead when it fails a path test;it marks a path alive again when it passes a path test.

• A dead logical device returned certain types of I/O errors toPowerPath and was judged unusable. Once a logical device ismarked dead (and until it is restored with powermt restore),PowerPath returns subsequent I/O requests with a failure status,without forwarding them to the logical device. This preventsfurther, unrecoverable corruption and allows the user to performdata recovery if needed. Dead is an unusual condition for logicaldevices. HP-UX is the only platform that ever marks logicaldevices as dead.

See also ”Alive,” “Path” and “Logical device.”

Default An attribute, value, or option that is assumed when no other isspecified explicitly.

Degraded One of three statuses reported by PowerPath for an HBA. The otherstatuses are failed and optimal. Degraded means one or more (but notall) I/O paths connected to this HBA have failed. See also ”Failed.”and “Optimal.”

Depot The URL location serves online bundles of software on VMwareUpdate Manager.

Device An addressable part (physical or logical) of a host or storage device.For example, PowerPath represents a path set between a host and alogical device as a uniquely named pseudo device.

Device number The value that logically identifies a device.


Glossary

Device driver Software that permits data transfer between a computer system and adevice such as a disk. Typically, a device driver interacts directly withsystem hardware. Consequently, satisfactory operation requiresdevice driver software that is compatible with a specific operatingsystem and hardware model.

Disabled (HBA) A user-defined HBA attribute, indicating the system administratorhas made the HBA unavailable for use by PowerPath. Disabling anHBA tells PowerPath not to use any paths originating from this HBAfor I/O. Disabling an HBA is done using operating-system-specificcommands, not in PowerPath. See also ”Enabled (HBA).”

EElectronic license

management (ELM)A software licensing model in which licenses are electronic files thatare cached, stored, and maintained electronically.

Enabled (HBA) A user-defined HBA attribute, indicating the system administratorconsiders the HBA available for use by PowerPath. Enabling an HBAis done using operating system–specific commands, not inPowerPath. See also ”Disabled (HBA).”

Encryption See ”PowerPath Encryption.”

E_Port An expansion port on a Fibre Channel switch that links multipleswitches into a fabric.

emcpower device The name used by PowerPath (on some operating systems) for apseudo device. See also ”Pseudo device.”

ESN Enterprise Storage Network. An ESN can provide high availability byconfiguring multiple paths between connections, configuringalternate paths to Storage Area Network (SAN) components, anddeploying redundant SAN components.

FFabric The facilities that link multiple Fibre Channel nodes.

Failed One of three statuses reported by PowerPath for an HBA. The otherstatuses are degraded and optimal. Failed means all paths to thisHBA are dead and no data is passing through the HBA. See also”Degraded.” and “Optimal.”


180

Glossary

Failover In PowerPath, the process of detecting a failure on an active path andautomatically sending data to another available path.

FC-AL See ”Fibre Channel Arbitrated Loop (FC-AL).”

Fibre A general term for all physical media types supported by the FibreChannel specification, such as optical fiber, twisted pair, and coaxialcable.

Fibre Channel The general name of an integrated set of ANSI standards that defineprotocols for flexible information transfer. Fibre Channel is ahigh-performance serial data channel.

Fibre ChannelArbitrated Loop

(FC-AL)

A standard for a shared access loop, in which several Fibre Channeldevices are connected (as opposed to point-to-point transmissions).See also ”Arbitrated loop.”

Firmware Software, typically startup and I/O instructions, stored in an HBA’sread-only memory. PowerPath installation requirements often specifyboth an HBA and a specific revision of that HBA’s firmware.

GGUI The acronym for graphical user interface, which represents an

application with icons, menus, and dialog boxes selectable by a user.Command-line interfaces are another major means of interactingwith an application. PowerPath Administrator is a GUI that allowsyou to interactively manage PowerPath on Windows platforms.

HHost The generic name for a computer connected to a network or cluster

system.

Host bus adapter(HBA)

A device through which a host can issue I/O requests. PowerPathreports the status of paths originating from HBAs as optimal, degraded,or failed.

Hub A Fibre Channel device used to connect several devices (such ascomputer servers and storage systems) into a Fibre Channel-Arbitrated Loop (FC-AL). See also ”Fibre Channel Arbitrated Loop(FC-AL).”


Glossary

IIdentifier (ID) A sequence of bits or characters that identifies a program, device,

controller, or system.

Initiator A SCSI or Fibre Channel device (usually a host system) that requestsan operation to be performed by another device, called the target. Seealso ”Target.”

Interface For example, an FA (Fibre Adapter) on a Symmetrix storage systemor an SP (Storage Processor) on a CLARiiON and VNX OE storagesystem. An array interface port is the front-end interface thatconnects to the SAN. An interface board (or array board) consists ofthe interface ports.

I/O device An addressable input/output unit, such as a disk device.

LLAC License authorization code. A unique alphanumeric value that

corresponds to one or more EMC products that you have purchased.The LAC indicates the products that you are authorized to use andprovides instructions for activating licenses at the PowerlinkLicensing portal.

least recently usedalgorithm (LRU)

The algorithm used to identify and make available the cache space byremoving the least recently used data.

Logical volume A user-defined storage device. In the Model 5200, the user can definea physical disk device as one or two logical volumes.

Least blocks (lb) A load-balancing and failover policy for PowerPath devices, in whichload balance is based on the number of blocks in pending I/Os. I/Orequests are assigned to the path with the fewest queued blocks,regardless of the number of requests involved. See also ”Loadbalancing.”

Least IOs (li) A load-balancing and failover policy for PowerPath devices, in whichload balance is based on the number of pending I/Os. I/O requestsare assigned to the path with the fewest queued requests, regardlessof total block volume. See also ”Load balancing.”


182

Glossary

Load balancing The activity of redistributing I/O workload across multiple pathsbetween an ESX host and an external storage device, according to adefined policy. See also ”Path.” and “Policy.”

Lockbox An RSA-developed encrypted file used to securely store remote hostIP addresses and their username and password combinations for therpowermt server. By storing this information, rpowermt does notrepeatedly prompt the user for the username and password for theremote host.

Logical device The smallest addressable storage unit. A logical device is an entitymanaged and presented by a storage system, which comprises one ormore physical disks or sections of physical disks. Logical devicesaggregated and managed at a higher level by a volume manager arereferenced as logical volumes rather than logical devices.

Logical VolumeManager (LVM)

Software that manages logical storage devices. Logical volumemanagers typically reside under the computer server’s filesystem.

Logical Unit Number(LUN)

An identifier for a physical or virtual device addressable through atarget.

LUN name See ”User-assignable LUN name.”

LUNZ A CLARiiON and VNX OE device used for a management programto communicate with the storage system. A LUNZ is used to tell thestorage system that the host exists and what is the WWN of the host.(A WWN, or World Wide Name, uniquely identifies a device on aFibre Channel network.)

A LUNZ device is present when no storage has been assigned to thehost. When Access Logix is used on a CLARiiON and VNX OEsystem, an agent runs on the host and communicates with the storagesystem through either the LUNZ or a storage device. On a CLARiiONand VNX OE system, the LUNZ device is replaced by the first storagedevice assigned to the host; the agent then communicates through thestorage device. See also ”VCMDB (Volume ConfigurationManagement Database).”

MMirroring Maintaining two or more identical copies of a designated volume on

two or more disks. Each copy is updated automatically during a writeoperation. Mirroring improves data availability: if one disk device


Glossary

fails, storage devices automatically use the other disk device to accessthe data. In this way, the mirrored copies of a disk can be presented asa single, fault-tolerant, virtual disk.

Mirrored pair A logical volume with all data recorded twice, once on each of twodifferent physical devices.

Mode An attribute of a PowerPath path. Path mode can be active orstandby. See also ”Active (paths).” and “Standby (paths).”

Mulitpathing Using redundant physical path components—host based adapters(HBAs), switches, storage processors (SPs), and cables—between anESX host and an external storage device to provide fault tolerance.

Multipathing Plugins(MPPs)

Third-party multipathing functionality that runs in parallel with theNMP on the ESX host and, for specified arrays, replaces the defaultNMP behavior by taking control of the path failover andload-balancing operations.

NNative Multipathing

Plugin (NMP)The default multipathing functionality for ESX hosts.

Native device A device created by the operating system to represent and provideaccess to a logical device. Typically, a native device is path aware (asopposed to path independent) and represents a single path to alogical device. The device is native in that it is provided by theoperating system for use with applications. PowerPath supportsnative devices on all platforms except AIX. See also ”PowerPathdevice.” and “Pseudo device.”

Nice name See ”User-assignable LUN name.”

No redirect (nr) A load-balancing and failover policy for PowerPath/VE devices, inwhich neither load balancing nor failover is in effect. If nr is set on afailed path and a native device is used, I/O errors will occur whenI/O is directed to that path. If one or more paths is failed and nr is set,data I/O errors can occur. This policy is the default for supportedCelerra and supported third-party devices on platforms without avalid PowerPath/VE license. This is the only time that a device is setto No redirect. No redirect is not supported as an argument to therpowermt set policy command. You cannot manually set the policyfor a device to nr. See also ”Load balancing.”


184

Glossary

OOperating system Software that manages the use and allocation of computer resources;

for example, memory, central processing unit (CPU), disk, andprinter access. PowerPath runs on several operating systems. InPowerPath documentation, an operating system and the hardware itruns on are referred to as a platform.

Optimal One of three statuses reported by PowerPath for an HBA; the othersare degraded and failed. Optimal means all paths to this HBA arealive (usable). See also ”Degraded.” and “Failed.”

Overdraft license A license type in the served licensing model that gives you a graceperiod for peak use of PowerPath for VMware ESX, with the samePowerPath performance and functionality as your normal license, inthe case that you somehow check out more licenses than you havepurchased.

PParameter A value given to a command variable. PowerPath powermt

commands have parameters that users can specify to tailor the effectsof the commands.

Path Any route between nodes in a network. In PowerPath, a path refers tothe route travelled by PowerPath data between a host and a logicaldevice. A path comprises an HBA, cables, a switch or hub (FibreChannel only), an interface and port, and a logical device.

Path failover The process of detecting a failed path and switching to another path.

Path Selection Plugin(PSP)

The vSphere plugin that performs load balancing and determineswhich physical path is used to issue an I/O request to an externalstorage device.

Path set In PowerPath, the group of all paths that read data from and writedata to the same logical device.

physical ID Physical identification number of the Symmetrix director for EREPusage. This value automatically increments by one for each directorinstalled in Symmetrix. This number must be unique in themainframe system. It should be an even number. This number isreferred to as the SCU_ID.


Glossary

Physical volume An addressable disk on the SCSI bus. See also ”SCSI.”

Physical volumeidentifier (PVID)

The number that the Logical Volume Manager uses to identifyspecific disks. See also ”Logical Volume Manager (LVM).”

Platform In PowerPath documentation, an operating system and the hardwareit runs on.

Policy A load-balancing and failover algorithm for PowerPath devices. Thiscan be changed with powermt set policy. See also ”Adaptive (ad),”“CLARiiON optimization (co),” “Least blocks (lb),” “Least IOs (li),”“Round robin (rr),” “StreamI/O (si),” and “Symmetrix optimization(so).” Basic failover is not supported as an argument to the rpowermtset policy command. You cannot manually set the policy for a deviceto bf. See also ”Basic failover (bf).”

Port (1) An access point for data entry or exit.(2) A receptacle on a device, to which a cable for another device isattached.

PowerPath device A device created by PowerPath for each logical device PowerPathdiscovers. There is a one-to-one relationship between PowerPathdevices and logical devices. PowerPath presents PowerPath devicesdifferently, depending on the platform. Much of this difference is dueto the design of the host operating system. Depending on theplatform, PowerPath may present PowerPath devices as nativedevices and/or pseudo devices. See also ”Logical device,” “Nativedevice,” and “Pseudo device.”

PowerPath driver PowerPath software that resides that within the ESX kernel thatcontains the PowerPath multipathing and load-balancingfunctionalities and caches the PowerPath license keys.

PowerPath Encryption PowerPath Encryption with RSA is host-based software that usesstrong encryption to safeguard sensitive data on disk devices.PowerPath Encryption assures the confidentiality of data on a diskdrive that is physically removed from a data center, and it preventsanyone who gains unauthorized access to the disk from reading orusing the data on that device.

PowerPath licenselibrary

A license library that resides in the ESX host that supports aninterface to query, set, and revoke licenses and maintains licensestates.


186

Glossary

Priority An integer in the range 0 to 10, which indicates the I/O priority forPowerPath devices. I/O to lower-priority logical devices tends to bedelivered to paths not used by higher-priority logical devices. Thiscan be changed with powermt set priority.

Priority is relevant only for PowerPath devices whose load-balancingand failover policy is set to Symmetrix optimization. Althoughpriority can be set for other PowerPath devices, it has no effect onload balancing on those devices.

Pseudo device A special kind of device (operating system object used to accessdevices) created by PowerPath. It is path independent, as are nativedevices, once PowerPath is installed. When a pseudo device iscreated, there is one (and only one) per path set. See also ”emcpowerdevice,” “Native device,” “Path set,” and “PowerPath device.”

RRaw device mapping

(RDM)A mechanism for a virtual machine to have direct access to a storagedevice that can be created on PowerPath-managed devices.

Reassignment On an active-passive storage system, movement of logical devicesfrom one storage system interface card to another. This occurs in theevent of a failure — of a storage system interface card or all paths toan interface card. If an interface card fails, logical devices arereassigned from the broken interface to another interface. Thisreassignment is initiated by the other, functioning interface. If allpaths from a host to an interface fail, logical devices accessed on thosepaths are reassigned to another interface, with which the host can stillcommunicate. This reassignment is initiated by PowerPath, whichinstructs the storage system to make the reassignment.

Reassignment can take several seconds to complete; however, I/Osdo not fail during it. After devices are reassigned, PowerPath detectsthe changes and seamlessly routes data via the new route.

The CLARiiON and VNX OE term for reassignment is trespassing.

Redundant path An independent communication channel between a host and a logicaldevice that already share at least one channel. PowerPath allows youto create redundant paths to promote failover. See also ”Failover.”

Rehosting Transferring a PowerPath/VE license from one ESX host to another.


Glossary

Remediation The process of applying the staged, or installed, software packagefrom the baseline onto the desired ESX host using VMware UpdateManager. Both installing and upgrading software packages usingVUM, as well as installing patches and hotfixes, are carried outthrough the remediation stage.

Remote CLI server The server that allows you to manage PowerPath for VMwarevSphere using the PowerPath CLI. It is also responsible for acquiringserved license files from an ELM server or for storing unservedlicense files. Also called rpowermt server.

Round robin (rr) A load-balancing and failover policy for PowerPath devices, in whichI/O requests are assigned to each available path in rotation. See also”Load balancing.”

rpowermt CIM client The functionality that enables communication between the VMwarevSphere host and the rpowermt host for management of PowerPathusing the PowerPath remote CLI. In the case of served licenses, therpowermt CIM client also carries out licensing queries.

rpowermt server The server that allows you to manage PowerPath for VMwarevSphere using the PowerPath CLI. It is also responsible for acquiringserved license files from an ELM server or for storing unservedlicense files. Also called remote CLI server.

rpowermt The PowerPath remote CLI commands used to configure and managePowerPath and paths under its control on ESX hosts in thePowerPath for VMware vSphere environment.

RTOOLS The package name of PowerPath remote CLI, or rpowermt.

SSAN Storage Area Network. See also ”ESN.”

SCSI The acronym for Small Computer System Interface, theANSI-standard set of protocols that defines connections betweenpersonal and other small computers and peripheral devices such asprinters and disks. PowerPath supports SCSI standards. Specificrequirements apply to each supported operating system.

SCSI device An HBA, peripheral controller, or intelligent peripheral that canattach to a SCSI bus.


188

Glossary

Single point of failure(SPOF)

A hardware or software design or configuration that depends on onecomponent for successful operation: If that component fails, theentire application fails. High-availability design tries to eliminate orminimize single points of failure through redundancy, recovery,and/or failover.

Staging The process of pushing a software package from the VMware UpdateManager server onto ESX hosts.

Standby (paths) One of two modes for I/O paths. A standby path is held in reserve.Being set to standby does not mean a path will not be used. Rather, itmeans that the weight of the path is heavily adjusted to preclude itsuse in normal operations. A standby path still can be selected if it isthe best path for a request. Path mode is set with powermt set mode.See also ”Active (paths).”

Storage group One or more LUNs within a storage system that is reserved for one ormore hosts and is inaccessible to other hosts. Access Logix enforcesthe host-to-storage group permissions and runs in the storage-systemSPs.

StreamI/O (si) A load-balancing and failover policy for PowerPath devices in which,for each possible path for an I/O to a particular volume, the selectedpath is the same as was selected for the previous I/O to the volume,unless the pending I/O count since the last path change exceeds thevolume’s threshold value. When the threshold is exceeded, the policyselects a new path based on the adaptive policy algorithm (See also”Adaptive (ad).”

String A series of connected disk devices sharing the same disk director.

Striping Segmenting logically sequential data and writing the segments tomultiple physical disks. Placing data across multiple disks improvesperformance, by aggregating the I/O performance of several disks. Italso improves availability, as the combined striped data can bepresented as a single, fault-tolerant, virtual disk.

Storage Array TypePlugin (SATP)

The vSphere plugin that handles array-specific path failoveroperations. The SATP monitors the health of each physical pathbetween the ESX host and the external storage device and performsarray-specific actions necessary for path failover should a path fail.


Glossary

Storage device A physical device that can attach to a SCSI device, which in turnconnects to the SCSI bus.

Switch A Fibre Channel device used to connect other devices (for example,computer servers and storage systems) into a Fibre Channel fabric. Ina switched topology, HBAs may be zoned to share storage-systemports. See also ”Fibre Channel.”

Symmetrixoptimization (so)

A load-balancing and failover policy for PowerPath devices, in whichI/O requests are routed to paths based on an algorithm that takes intoaccount path load and the logical device priority you set withpowermt set policy. Load is a function of the number, size, priority,and type of I/O queued on each path. This policy is valid forSymmetrix storage systems only and is the default policy for them, onplatforms with a valid PowerPath license. It is listed in powermt

display output as SymmOpt. See also ”Load balancing.”

SRDF (SymmetrixRemote Data Facility)

The microcode and hardware required to support Symmetrix remotemirroring. See also ”Mirroring.”

TTarget A SCSI or Fibre Channel device that performs an I/O process

requested by another device, called the initiator. See also ”Initiator.”

Trespassing The CLARiiON and VNX OE term for reassignment. See also”Reassignment.”

UUNIX An interactive, multitasking, multiuser operating system supported

by PowerPath.

User-assignable LUNname

A character string that a user or system manager associates with alogical device on a CLARiiON and VNX OE array and assignsthrough Navisphere and Unisphere.

VVMware vCenter

ConverterA feature in the VMware vSphere environment that automates andsimplifies physical to virtual machine conversions as well asconversions between virtual machine formats


190

Glossary

VMware DistributedResource Scheduler

(DRS)

A feature in the VMware vSphere environment that dynamicallyallocates and balances computing capacity across the clustered hosts.DRS monitors the workload of the running virtual machines and canmigrate virtual machines to ESX hosts with spare capacity.

VMware vSphere CLIcommands

The command line commands that are used to manage PowerPathand paths under its control on ESX hosts in the PowerPath forVMware vSphere environment.

VMware ESX host The server where the PowerPath driver, ESX kernel, the PowerPathlicense library, Solutions Enabler (SE) multipath provider, and therpowermt CIM client reside.

VMware vCenter FileSystem (VMFS)

A clustered file system that leverages shared storage to allowmultiple ESX hosts to read and write to the same storagesimultaneously. You can create, expand, and delete VMFS volumeson storage devices claimed by PowerPath.

VMware HighAvailability (HA)

A feature in the VMware vSphere environment that provides highavailability for applications running in virtual machines. If an ESXhost fails, affected virtual machines are automatically restarted onother host that have spare capacity.

VMware HighAvailability (HA) andDistributed Resource

Scheduler (DRS)

A VMware vSphere cluster consisting of the aggregate computingand memory resources of a group of ESX hosts sharing the samenetworks and storage arrays. ESX clusters provide high availabilityand load balancing of virtual machines.

VMware vCenter SiteRecovery Manager

(SRM)

A feature in the VMware vSphere environment that providessimplified, automated disaster recovery.

VMware StorageVMotion

A feature in the VMware vSphere environment that enables themigration of virtual machine files from one datastore to anotherwithout service interruption. You can use Storage VMotion to migratedatastores built on PowerPath-managed devices.

VMware VMotion A feature in the VMware vSphere environment that enables the liveand uninterrupted migration of running virtual machines from oneESX host to another You can use VMotion to migrate VMs built ondatastores on PowerPath-managed devices.


Glossary

VCMDB (VolumeConfigurationManagement

Database)

A Symmetrix device used for a management program tocommunicate with the storage system. A VCBDM is used to tell thestorage system that the host exists and what is the WWN of the host.(A WWN, or World Wide Name, uniquely identifies a device on aFibre Channel network.)

A VCMDB is present when using Volume Logix to perform LUNmasking. When storage is assigned to the host, the storage appears inaddition to the VCMDB. See also ”LUNZ.”

Volume An abstracted, logical disk device. Volumes read and write data likeother disk devices, but typically they do not support otheroperations. A Symmetrix volume may comprise storage on one ormore Symmetrix devices, but it is presented to hosts as a single diskdevice.

A volume can be a single disk partition or multiple disk partitions onone or more physical drives. A volume can coincide with a logicaldevice, include multiple logical devices, or contain only a piece of alogical device. Applications that use volumes do not need to be awareof the underlying physical structure; software handles the mappingof virtual addresses to physical addresses.

See also ”Logical device.”

Volume group A group of physical volumes.

Volume manager Software that creates and manages logical volumes that span multiplephysical disks, allowing greater flexibility and reliability for storingdata.

vSphere Center Server A server in the VMware vSphere environment that provides a centralpoint for configuring, provisioning, and managing VMware datacenters.

ZZone A set of devices that can access one another. All devices connected to

a Fibre Channel connectivity product (such as the ED-1032 Director)may be configured into one or more zones. Devices in the same zonecan “see” each other, while those in different zones cannot. Zoningallows an administrator to group several devices by function orlocation. See also ”Fibre Channel.”


192

Glossary


Index

AAccess Logix 175active path mode 144Active-active storage 23Active-active storage systems 22Active-passive storage 23Adapter 175Adaptive (ad) 176Adaptive policy 151Adding

logical devices 83ALUA 24Arbitrated loop 176assign 23

BBaseline 40Boot device 177Bus 177

CChannel 177Claim rules 25, 66, 69CLARiiON LUN names 121, 131, 161CLARiiON optimization policy 151CLARiiON user-assignable LUN names 121, 131,

161Cluster 177comments 11Consistency group 177Controller 178

DData availability 178Data channel 178Default 178Degraded HBA path status 123Device

definition 178display, field explanations 125driver 179installation troubleshooting 85number 178

Disabled HBA status 179documentation, related 7Dynamic multipath load balancing 22

EE_Port 179ELM license server 16ELM license server configuration 63emcpower device 179Enabled

HBA status 179error messages

rpowermt 112ESN 179

FFabric 179Failed HBA path status 123Failover policy. See Load balancing and failover

policyFC-AL 180


194

Index

Fibre 180Fibre Channel

definition 180Fibre Channel Arbitrated Loop 180Firmware 180

GGUI 180

HHost bus adapter (HBA)

disabled status 179display field explanations 123, 133, 137enabled status 179path status 123

Hub 180

II/O priority. See priorityIdentifier (ID) 181Initiator 181Interface 181Interface card 23

LLAC 18Latency (path)

monitoring 147Least blocks policy 151Least I/Os policy 151load balancing 14Load balancing and failover policy

Adaptive 151CLARiiON optimization 151least blocks 151least I/Os 151round robin 152setting 151Stream IO 152Symmetrix optimization 152

load-balancing groups 24lockbox 56, 58, 89, 117Logical devices

adding 83

Logical Unit Number (LUN) 182Logical Volume Manager (LVM) 182LUN names 121, 131, 161LUNZ 182

Mmessage sources

multipathing 165Mirrored pair 183Mirroring 182Multipathing 14Multipathing Plugins (MPPs) 14

NNative Multipathing Plugin (NMP) 14Nice names. See CLARiiON LUN namesNMP 25

OOperating system 184Optimal HBA path status 123Overdraft licenses 17

PParameter 184path failover 14Paths

HBA status 123latency monitoring on 147removing 85restoration 142

Periodic autorestore 22Physical volume 185Physical Volume Identifier (PVID) 185Platform 185Policy. See Load balancing and failover policyPowerlink website 8PowerPath default claim rules 45, 69, 95PowerPath driver 15PowerPath messages

message sources 165PowerPath messages 167

priority, setting 153Proactive path testing 21


Index

RReassignment 186Redundant path 186Rehosting 64, 186Remediate 41Removing

paths 85removing logical devices from PowerPath

configuration 85Restoring paths 142Round robin policy 152rpowermt

command summary 110error messages 112lockbox 112, 117syntax 112warning messages 112

rpowermt check_registration 120rpowermt CIM client 15rpowermt commands 88rpowermt config 121rpowermt display 122rpowermt display latency 128rpowermt display options 131rpowermt display paths 132rpowermt display ports 135rpowermt help 139rpowermt register 141rpowermt restore 142rpowermt server 187rpowermt set mode 144rpowermt set path_latency_monitor 147rpowermt set path_latency_threshold 148rpowermt set periodic_autorestore 150rpowermt set policy 151rpowermt set priority 153rpowermt setup add_host 155rpowermt setup list_hosts 157rpowermt setup remove_host 158rpowermt setup update_host 159rpowermt unregister 160rpowermt update lun_names 161rpowermt version 162rtools 50

SSAN 187SCSI

definition 187device 187

Single point of failure (SPOF) 188Single-server availability 17Stage 40standby path mode 144Storage 188Storage device 189Storage group 188Stream I/O (si) 188Stream IO policy 152Striping 188Switch 189Symmetrix optimization policy 152Symmetrix Remote Data Facility (SRDF) 189

TTarget 189trespass back 23Trespassing 23, 186

Uuser-assignable LUN names 121, 131, 161

VVCMDB (Volume Configuration Management

Database) 191Volume 191Volume group 191Volume manager 191

Wwarning messages

rpowermt 112

ZZone 191

195PowerPath/VE for VMware vSphere V5.4 and Service Packs Installation and Administration Guide

196

Index


300-008-703

Documents

program files emc powerpath rpowermt

emcpower2 clariion id hk190807490007

severe personal injury

service pack releases

fully qualified domain

vmware vstorage apis

developed encrypted file

complex san topology