IBM TotalStorage(TM) Enterprise Storage Server(TM)
Subsystem Device Driver User's Guide

Twelfth Edition (May 2002)

Document Number GC26-7442-03
Note

Before using this information and the product it supports, read the information in Notices.

Twelfth Edition (May 2002)

This edition replaces the Eleventh Edition of this book (form number GC26-7442-02), and it includes information that specifically applies to IBM ESS Subsystem Device Driver (SDD):

This edition also applies to all subsequent releases and modifications until otherwise indicated in new editions.

(C) Copyright International Business Machines Corporation 1999, 2002. All rights reserved.
U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp.


Contents

Figures

Tables

About this book

  • Who should use this book
  • Summary of changes
  • New information
  • Modified information
  • Publications
  • The IBM TotalStorage ESS library
  • Other IBM publications
  • Other non-IBM publications
  • Web sites
  • How to send your comments
  • Chapter 1. Overview of the Subsystem Device Driver

  • Subsystem Device Driver
  • Enhanced data availability
  • Dynamic I/O load-balancing
  • Automatic path-failover protection
  • Concurrent download of licensed internal code
  • Path-selection policies for the host system
  • Chapter 2. SDD for an AIX host system

  • Command syntax conventions
  • Highlighting conventions
  • Special characters conventions
  • Verifying the hardware and software requirements
  • Hardware
  • Software
  • Non-supported environments
  • Host system requirements
  • Preparing for SDD installation
  • Configuring the ESS
  • Installing the AIX fibre-channel device drivers
  • Uninstalling the AIX fibre-channel device drivers
  • Configuring fibre-channel attached devices
  • Removing fibre-channel attached devices
  • Verifying the Emulex adapter firmware level
  • Upgrading the Emulex adapter firmware level
  • Determining the installation package
  • Installing the Subsystem Device Driver
  • Verifying your currently installed version of SDD
  • Preparing to configure the Subsystem Device Driver
  • Configuring the Subsystem Device Driver
  • Unconfiguring Subsystem Device Drivers
  • Verifying the SDD configuration
  • Changing the path-selection policy
  • Adding paths to SDD devices of a volume group
  • Upgrading SDD for AIX 4.2.1, AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0
  • Understanding the SDD support for nondisruptive installation
  • Verifying your previously installed version of SDD
  • Upgrading SDD using a nondisruptive installation
  • Upgrading SDD manually
  • Removing SDD from an AIX host system
  • Using concurrent download of licensed internal code
  • Understanding the SDD support for High Availability Cluster Multi-Processing (HACMP)
  • SDD installation package attributes
  • Providing load-balancing and fallover protection
  • Displaying the ESS vpath device configuration
  • Configuring a volume group for fallover protection
  • Importing a volume group with SDD
  • Exporting a volume group with SDD
  • Losing fallover protection
  • Using ESS devices directly
  • Using ESS devices through AIX LVM
  • Migrating a non-SDD volume group to an ESS SDD multipath volume group in concurrent mode
  • Using the trace function
  • Error log messages
  • Chapter 3. SDD for a Windows NT host system

  • Verifying the hardware and software requirements
  • Hardware
  • Software
  • Non-support environments
  • ESS requirements
  • Host system requirements
  • Preparing for SDD installation
  • Configuring the ESS
  • Configuring fibre-channel adapters
  • Configuring SCSI adapters
  • Installing the Subsystem Device Driver
  • Configuring the Subsystem Device Driver
  • Adding paths to SDD devices
  • Upgrading the Subsystem Device Driver
  • Adding or modifying a multipath storage configuration to the ESS
  • Removing the Subsystem Device Driver
  • Displaying the current version of the Subsystem Device Driver
  • Support for Windows NT clustering
  • Special considerations in the Windows NT clustering environment
  • Configuring a Windows NT cluster with SDD
  • Chapter 4. SDD for a Windows 2000 host system

  • Verifying the hardware and software requirements
  • Non-supported environments
  • ESS requirements
  • Host system requirements
  • Preparing for SDD 1.3.2.1 installation
  • Configuring the ESS
  • Configuring fibre-channel adapters
  • Configuring SCSI adapters
  • Installing SDD 1.3.2.1
  • Displaying the current version of the Subsystem Device Driver
  • Uninstalling the Subsystem Device Driver
  • Adding paths to SDD devices
  • Activating additional paths
  • Verifying additional paths are installed correctly
  • Support for Windows 2000 clustering
  • Special considerations in the Windows 2000 clustering environment
  • Configuring a Windows 2000 cluster with SDD
  • Chapter 5. SDD for a Hewlett-Packard host system

  • Verifying the hardware and software requirements
  • Hardware
  • Software
  • Non-supported environments
  • SDD 1.3.0.5 support on HP-UX 11.0 and HP-UX 11i
  • Understanding how SDD works on an HP host system
  • Preparing for SDD installation
  • Configuring the ESS
  • Planning for installation
  • Installing the Subsystem Device Driver
  • Post-installation
  • Upgrading the Subsystem Device Driver
  • Uninstalling the Subsystem Device Driver
  • Changing an SDD hardware configuration
  • Using applications with SDD
  • Standard UNIX applications
  • Installing SDD on a Network File System file server
  • Oracle
  • Chapter 6. SDD for a Sun host system

  • Verifying the hardware and software requirements
  • Hardware
  • Software
  • Supported environments
  • Non-supported environments
  • Understanding how SDD works on a Sun host system
  • Preparing for SDD installation
  • Configuring the ESS
  • Planning for installation
  • Installing the Subsystem Device Driver
  • Verifying the SDD installation
  • Uninstalling the Subsystem Device Driver
  • Post-installation
  • Upgrading the Subsystem Device Driver
  • Changing an SDD hardware configuration
  • Using applications with SDD
  • Standard UNIX applications
  • Installing SDD on a Network File System file server
  • Oracle
  • Solstice DiskSuite
  • Chapter 7. Using the datapath commands

  • datapath query adapter command
  • datapath query adaptstats command
  • datapath query device command
  • datapath query devstats command
  • datapath set adapter command
  • datapath set device command
  • Notices

  • Trademarks
  • Electronic emission notices
  • Federal Communications Commission (FCC) statement
  • Industry Canada compliance statement
  • European community compliance statement
  • Japanese Voluntary Control Council for Interference (VCCI) class A statement
  • Korean government Ministry of Communication (MOC) statement
  • Taiwan class A compliance statement
  • IBM agreement for licensed internal code
  • Actions you must not take
  • Glossary

    Index


    Figures

    1. Multipath connections between a host system and the disk storage in an ESS
    2. Output from the Display Data Path Device Configuration SMIT panel
    3. IBMdpo Driver 32-bit
    4. IBMdpo Driver 64-bit

    Tables

    1. Publications in the ESS library
    2. Other IBM publications related to the ESS.
    3. Other IBM publications without order numbers
    4. ESS Web sites and descriptions
    5. SDD in the protocol stack
    6. Required number of successful I/O operations before SDD places a path in the open state
    7. AIX required fixes
    8. SDD 1.3.1.3 support for 32-bit and 64-bit applications on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0
    9. SDD 1.3.1.3 installation packages for non-HACMP and HACMP environments on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0
    10. Major files included in the SDD installation package
    11. List of previously installed installation packages that are supported with nondisruptive installation
    12. Software support for HACMP in concurrent mode
    13. Software support for HACMP in nonconcurrent mode
    14. Software support for HACMP in concurrent mode on AIX 5.1.0 (32-bit kernel only)
    15. Software support for HACMP in nonconcurrent mode on AIX 5.1.0 (32-bit kernel only)
    16. HACMP and supported SDD features
    17. SDD-specific SMIT panels and how to proceed
    18. SDD installation scenarios
    19. HP 11.0 patches necessary for proper operation of SDD
    20. SDD components installed for HP host systems
    21. System files updated for HP host systems
    22. SDD commands and their descriptions for HP host systems
    23. SDD installation scenarios
    24. SDD package file names
    25. Solaris patches necessary for proper operation of SDD
    26. SDD components installed for Sun host systems
    27. System files updated for Sun host systems
    28. SDD commands and their descriptions for Sun host systems
    29. Commands

    About this book

    This book provides step-by-step procedures for you to install, configure, and use the IBM(R) TotalStorage(TM) Enterprise Storage Server(TM) (ESS) Subsystem Device Driver (SDD) on IBM AIX(R), HP, Sun, Microsoft(R) Windows NT(R), and Microsoft Windows 2000 host systems.


    Who should use this book

    This book is intended for storage administrators, system programmers, and performance and capacity analysts.


    Summary of changes

    This book contains both information previously presented in the Eleventh Edition (form number GC26-7442-02) of the IBM TotalStorage Enterprise Storage Server Subsystem Device Driver User's Guide and major technical changes to that information. Technical changes are indicated by revision bars (|) in the left margin of the book. The following sections summarize those technical changes.

    Note:
    For the last-minute changes that are not included in this book, see the README file on the SDD compact disc or visit the SDD Web site at:
    www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates

    New information

    This edition includes the following new information:

    What's new in Chapter 2. SDD for an AIX host system:

    What's new in Chapter 3. SDD for a Windows NT host system:

    Modified information

    This edition includes the following modified information:

    What's modified in Chapter 1. Overview of the Subsystem Device Driver:

    What's modified in Chapter 4. SDD for a Windows 2000 host system:

    What's modified in Chapter 5. SDD for a Hewlett-Packard host system::

    What's modified in Chapter 6. SDD for a Sun host system:


    Publications

    The tables in this section list and describe the following publications:

    See Ordering ESS publications for information about how to order publications in the IBM TotalStorage ESS publication library. See How to send your comments for information about how to send comments about the publications.

    The IBM TotalStorage ESS library

    Table 1 shows the customer publications that comprise the ESS library. See The IBM publications center for information about ordering these and other IBM publications.

    Table 1. Publications in the ESS library
    Long title (short title) Description Order number
    IBM TotalStorage Enterprise Storage Server Copy Services Command-Line Interface User's Guide (ESS CLI User's Guide) This user's guide describes the commands you can use from the ESS Copy Services command-line interface (CLI). The CLI application provides a set of commands you can use to write customized scripts for a host system. The scripts initiate pre-defined tasks in an ESS Copy Services server application. You can use the CLI commands to indirectly control ESS Peer-to-Peer Remote Copy and FlashCopy configuration tasks within an ESS Copy Services server group.

    This book is not available in hardcopy. It is available in PDF format on the following Web site:

    www.storage.ibm.com/hardsoft/products/ess/refinfo.htm

    SC26-7434
    IBM TotalStorage Enterprise Storage Server Configuration Planner (ESS Configuration Planner) This guide provides work sheets for planning the logical configuration of the ESS. This book is not available in hardcopy. This guide is available on the following Web site:

    www.storage.ibm.com/hardsoft/products/ess/refinfo.htm

    SC26-7353
    IBM TotalStorage Enterprise Storage Server Host System Attachment Guide (ESS Attachment Guide) This book provides guidelines for attaching the ESS to your host system and for migrating from Small Computer System Interface (SCSI) to fibre-channel attachment. SC26-7296
    IBM TotalStorage Enterprise Storage Server DFSMS Software Support Reference (ESS DFSMS Software Support) This book gives an overview of the ESS and highlights its unique capabilities. It also describes Data Facility Storage Management Subsystems (DFSMS) software support for the ESS, including support for large volumes. SC26-7440
    IBM TotalStorage Enterprise Storage Server Introduction and Planning Guide (ESS Introduction and Planning Guide) This guide introduces the ESS product and lists the features you can order. It also provides guidelines for planning the installation and configuration of the ESS. GC26-7294
    IBM TotalStorage Enterprise Storage Server Quick Configuration Guide (ESS Quick Configuration Guide) This booklet provides flow charts for using the TotalStorage Enterprise Storage Server Specialist (ESS Specialist). The flow charts provide a high-level view of the tasks the IBM service support representative performs during initial logical configuration. You can also use the flow charts for tasks that you might perform when you are modifying the logical configuration. The hardcopy of this booklet is a 9-inch × 4-inch fanfold. SC26-7354
    IBM Enterprise Storage Server System/390 Command Reference (ESS S/390 Command Reference) This book describes the functions of the ESS and provides reference information for S/390(R) and e(logo)server zSeries hosts, such as channel commands, sense bytes, and error recovery procedures. SC26-7298
    IBM TotalStorage Safety Notices (Safety Notices) This book provides translations of the danger notices and caution notices that IBM uses in ESS publications. GC26-7229
    IBM TotalStorage Enterprise Storage Server SCSI Command Reference (ESS SCSI Command Reference) This book describes the functions of the ESS. It provides reference information for UNIX(R), Application System/400(R) (AS/400(R)), and e(logo)server iSeries 400 hosts, such as channel commands, sense bytes, and error recovery procedures. SC26-7297
    IBM TotalStorage Enterprise Storage Server User's Guide (ESS Users Guide) This guide provides instructions for setting up and operating the ESS and for analyzing problems. SC26-7295
    IBM TotalStorage Enterprise Storage Server Web Interface User's Guide (ESS Web Interface Users Guide) This guide provides instructions for using the two ESS Web interfaces, ESS Specialist and ESS Copy Services. SC26-7346

    Ordering ESS publications

    All the customer publications that are listed in The IBM TotalStorage ESS library are available on a compact disc that comes with the ESS, unless otherwise noted.

    The customer documents are also available on the following ESS Web site in PDF format:

    www.storage.ibm.com/hardsoft/products/ess/refinfo.htm

    The IBM publications center

    The publications center is a worldwide central repository for IBM product publications and marketing material.

    The IBM publications center offers customized search functions to help you find the publications that you need. A number of publications are available for you to view or download free of charge. You can also order publications. The publications center displays prices in your local currency. You can access the IBM publications center through the following Web site:

    www.ibm.com/shop/publications/order/

    Publications notification system

    The IBM publications center Web site offers you a notification system about IBM publications. Register and you can create your own profile of publications that interest you. The publications notification system sends you daily electronic mail (e-mail) notes that contain information about new or revised publications that are based on your profile.

    If you want to subscribe, you can access the publications notification system from the IBM publications center at the following Web site:

    www.ibm.com/shop/publications/order/

    Other IBM publications

    Table 2 lists and describes other IBM publications that have information related to the ESS.

    Table 2. Other IBM publications related to the ESS.
    Title Description Order number
    DFSMS/MVS(R) Version 1 Advanced Copy Services, This publication helps you to understand and use IBM Advanced Copy Services functions on an S/390 or zSeries. It describes two dynamic-copy functions and several point-in-time copy functions. These functions provide backup and recovery of data if a disaster occurs to your data center. The dynamic-copy functions are Peer-to-Peer Remote Copy and Extended Remote Copy. Collectively, these functions are known as remote copy. FlashCopy(TM) and Concurrent Copy are the point-in-time copy functions. SC35-0355
    DFSMS/MVS Version 1 Remote Copy Guide and Reference This publication provides guidelines for using remote copy functions with S/390 and zSeries hosts. SC35-0169
    Enterprise Storage Solutions Handbook This book helps you understand what comprises enterprise storage management. The concepts include the key technologies that you need to know and the IBM subsystems, software, and solutions that are available today. It also provides guidelines for implementing various enterprise storage administration tasks, so that you can establish your own enterprise storage management environment. SG24-5250
    ESS Fibre-Channel Migration Scenarios This white paper describes how to change your host system attachment to the ESS from SCSI and SAN Data Gateway to native fibre-channel attachment.

    To get the white paper, go to the following Web site:

    www.storage.ibm.com/hardsoft/products/ess/refinfo.htm

    No order number
    Enterprise Systems Architecture/390 ESCON I/O Interface This publication provides a description of the physical and logical ESA/390 I/O interface and the protocols which govern information transfer over that interface. It is intended for designers of programs and equipment associated with the ESCON I/O interface and for service personnel maintaining that equipment. However, anyone concerned with the functional details of the ESCON I/O interface will find it useful. SA22-7202
    ESS Solutions for Open Systems Storage Compaq AlphaServer, HP, and Sun This book helps you to install, tailor, and configure the ESS when you attach Compaq AlphaServer (running Tru64 UNIX), HP, and Sun hosts. This book does not cover Compaq AlphaServer running the Open VMS operating system. This book also focuses on the settings required to give optimal performance and on device driver levels. This book is for the experienced UNIX professional who has a broad understanding of storage concepts. SG24-6119
    Fibre Channel Connection (FICON) I/O Interface, Physical Layer This publication provides information to the Fiber Channel I/O Interface. This book is also available in PDF format by accessing the following Web site:

    www.ibm.com/servers/resourcelink/

    SA24-7172
    Fibre-channel Subsystem Installation Guide This publication tells you how to attach the xSeries 430 and NUMA-Q host system with fibre-channel adapters. See note.
    Fibre Transport Services (FTS) Direct Attach, Physical and Configuration Planning Guide This publication provides information about fibre-optic and ESCON-trunking systems. GA22-7234
    IBM Enterprise Storage Server This book, from the IBM International Technical Support Organization, introduces the ESS and provides an understanding of its benefits. It also describes in detail the architecture, hardware, and functions of the ESS. SG24-5465
    IBM Enterprise Storage Server Performance Monitoring and Tuning Guide This book provides guidance on the best way to configure, monitor, and manage your ESS to ensure optimum performance. SG24-5656
    IBM OS/390 Hardware Configuration Definition User's Guide This publication provides detailed information about the IODF. It also provides details about configuring parallel access volumes (PAVs). OS/390 uses the IODF. SC28-1848
    IBM SAN Fibre Channel Managed Hub 3534 Service Guide The IBM SAN Fibre Channel Managed Hub can now be upgraded to switched fabric capabilities with this Entry Switch Activation Feature. As your fibre channel SAN requirements grow, and you need to migrate from the operational characteristics of the Fibre Channel arbitrated loop (FC-AL) configuration provided by the IBM Fibre Channel Managed Hub, 35341RU, to a fabric capable switched environment, the Entry Switch Activation feature is designed to provide this upgrade capability. This upgrade is designed to allow a cost-effective, and scalable approach to developing fabric based Storage Area Networks (SANs). The Entry Switch Activation feature (P/N 19P3126) supplies the activation key necessary to convert the FC-AL based Managed Hub to fabric capability with eight fabric F_ports, one of which can be an interswitch link-capable port, an E_port, for attachment to the IBM SAN Fibre Channel Switch, or other supported switches. GC26-7391
    IBM SAN Fibre Channel Managed Hub 3534 Users Guide The IBM SAN Fibre Channel Switch 3534 is an eight-port Fibre Channel Gigabit Hub that consists of a motherboard with connectors for supporting up to eight ports, including seven fixed shortwave optic ports and one GBIC port, and an operating system for building and managing a switched loop architecture. SY27-7616
    IBM SAN Fibre Channel Switch, 2109 Model S08 Users Guide The IBM Fibre Channel Switch 2109 Model S08 Users Guide manual describes the switch and the IBM StorWatch Specialist. It provides information on the commands and how to manage the switch with Telnet and Simple Network Management Protocol (SNMP).

    To get a copy of this manual, see the Web site at:

    www.ibm.com/storage/fcswitch

    SC26-7349
    IBM SAN Fibre Channel Switch 2109 Model S16 Installation and Service Guide This publication describes how to install and maintain the IBM SAN Fibre Channel Switch 2109 Model S16. It is intended for trained service representatives and service providers who act as the primary level of field hardware service support to help solve and diagnose hardware problems.

    To get a copy of this manual, see the Web site at:

    www.ibm.com/storage/fcswitch

    SC26-7352
    IBM StorWatch Expert Hands-On Usage Guide This guide helps you to install, tailor, and configure ESS Expert, and it shows you how to use Expert. SG24-6102
    IBM TotalStorage Enterprise Storage Server Subsystem Device Driver Installation and Users Guide This book describes how to use the IBM Subsystem Device Driver on open-systems hosts to enhance performance and availability on the ESS. The Subsystem Device Driver creates redundant paths for shared logical unit numbers. The Subsystem Device Driver permits applications to run without interruption when path errors occur. It balances the workload across paths, and it transparently integrates with applications.

    For information about the Subsystem Device Driver, see the following Web site:

    www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates/

    GC26-7442
    Implementing ESS Copy Services on S/390 This publication tells you how to install, customize, and configure Copy Services on an ESS that is attached to an S/390 or zSeries host system. Copy Services functions include Peer-to-Peer Remote Copy, Extended Remote Copy, FlashCopy(TM) and, Concurrent Copy. This publication describes the functions, prerequisites, and corequisites and describes how to implement each of the functions into your environment. SG24-5680
    Implementing ESS Copy Services on UNIX and Windows NT/2000 This publication tells you how to install, customize, and configure ESS Copy Services on UNIX or Windows NT host systems. Copy Services functions include Peer-to-Peer Remote Copy, FlashCopy, Extended Remote Copy, and Concurrent Copy. Extended Remote Copy and Concurrent Copy are not available for UNIX and Windows NT host systems; they are only available on the S/390 or zSeries. This publication describes the functions and shows you how to implement each of the functions into your environment. It also shows you how to implement these solutions in ahigh-availability cluster multiprocessing (HACMP) cluster. SG24-5757
    Implementing Fibre Channel Attachment on the ESS This book helps you to install, tailor, and configure fibre-channel attachment of open-systems hosts to the ESS. It gives you a broad understanding of the procedures involved and describes the prerequisites and requirements. It also shows you how to implement fibre-channel attachment. This book also describes the steps required to migrate to direct fibre-channel attachment from native SCSI adapters and from fibre-channel attachment through the SAN Data Gateway (SDG). SG24-6113
    Implementing the IBM Enterprise Storage Server This book can help you install, tailor, and configure the ESS in your environment. SG24-5420
    NUMA-Q ESS Integration Release Notes for NUMA Systems

    This publication provides information about special procedures and limitations involved in running ESS with Copy Services on an IBM e(logo)server xSeries 430 and an IBM NUMA-Q(R) host system.

    It also provides information on how to:

    • Configure the ESS
    • Configure the IBM NUMA-Q and xSeries 430 host system
    • Manage the ESS from the IBM NUMA-Q and xSeries 430 host system with DYNIX/ptx tools

    Part number 1003-80094.
    OS/390 MVS System Messages Volume 1 (ABA - ASA) This publication lists OS/390 and zSeries MVS system messages ABA to ASA. GC28-1784
    z/Architecture Principles of Operation This publication provides, for reference purposes, a detailed definition of the z/Architecture. It is written as a reference for use primarily by assembler language programmers and describes each function at the level of detail needed to prepare an assembler language program that relies on that function; although anyone concerned with the functional details of z/Architecture will find it useful. SA22-7832

    Note:
    There is no order number for this publication. This publication is not available through IBM ordering systems. Contact your sales representative to obtain this publication.

    Other non-IBM publications

    Table 3 lists and describes other related publications that are not available through IBM ordering systems. To order, contact the sales representative at the branch office in your locality.

    Table 3. Other IBM publications without order numbers
    Title Description
    Quick Start Guide: An Example with Network File System (NFS) This publication tells you how to configure the Veritas Cluster Server. See also the companion document, Veritas Cluster Server User's Guide.
    Veritas Cluster Server Installation Guide This publication tells you how to install the Veritas Cluster Server. See also the companion document, Veritas Cluster Server Release Notes.
    Veritas Cluster Server Release Notes This publication tells you how to install the Veritas Cluster Server. See also the companion document, Veritas Cluster Server Installation Guide.
    Veritas Cluster Server User's Guide This publication tells you how to configure the Veritas Cluster Server. See also the companion document, Quick Start Guide: An Example with NFS.
    Veritas Volume Manager Hardware Notes This publication tells you how to implement dynamic multipathing.
    Veritas Volume Manager Installation Guide This publication tells you how to install VxVM. It is not available through IBM ordering systems. Contact your sales representative to obtain this document.
    Veritas Volume Manager Storage Administrators Guide This publication tells you how to administer and configure the disk volume groups.


    Web sites

    Table 4 shows Web sites that have information about the ESS and other IBM storage products.

    Table 4. ESS Web sites and descriptions
    Web site Description
    www.storage.ibm.com/ This Web site has general information about IBM storage products.
    www.storage.ibm.com/hardsoft/products/ess/ess.htm This Web site has information about the IBM Enterprise Storage Server (ESS).
    ssddom02.storage.ibm.com/disk/ess/documentation.html This Web site allows you to view and print the ESS publications.
    www.storage.ibm.com/hardsoft/products/ess/supserver.htm This Web site provides current information about the host system models, operating systems, and adapters that the ESS supports.
    ssddom01.storage.ibm.com/techsup/swtechsup.nsf/support/sddupdates/ This Web site provides information about the IBM Subsystem Device Driver.
    www.storage.ibm.com/hardsoft/products/sangateway/sangateway.htm This Web site provides information about attaching Storage Area Network or host system that uses an industry-standard, fibre-channel arbitrated loop (FC-AL) topology through the IBM 2108 Storage Area Network Data Gateway Model G07.
    www.storage.ibm.com/software/sms/sdm/sdmtech.htm This Web site provides information about the latest updates to Copy Services components including XRC, PPRC, Concurrent Copy, and FlashCopy for S/390 and zSeries.
    ssddom01.storage.ibm.com/techsup/swtechsup.nsf/support/sddcliupdates/ This Web site provides information about the IBM ESS Copy Services Command-Line Interface (CLI).


    How to send your comments

    Your feedback is important to help us provide the highest quality information. If you have any comments about this book or any other ESS documentation, you can submit them in one of the following ways:


    Chapter 1. Overview of the Subsystem Device Driver

    This chapter introduces the IBM TotalStorage Enterprise Storage Server (ESS) Subsystem Device Driver (SDD) and provides an overview of SDD functions.


    Subsystem Device Driver

    The Subsystem Device Driver is a pseudo device driver designed to support the multipath configuration environments in the IBM ESS. It resides in a host system with the native disk device driver and provides the following functions:

    As the diagrams in Table 5 show, SDD resides above the disk driver of a host system in the protocol stack and acts as a pseudo device driver. I/O operations that are sent to SDD proceed to the host disk driver after path selection. When an active path experiences a failure (such as a cable or controller failure), SDD dynamically switches to another path.

    Table 5. SDD in the protocol stack

    cmvc filename: s008996q


    cmvc filename: s008998q


    cmvc filename: s008999q


    cmvc filename: s008997q


    cmvc filename: s009024

    Each SDD device represents a unique physical device on the storage server. There can be up to 32 hdisk devices that represent up to 32 different paths to the same physical device.

    SDD devices behave almost like hdisk devices. Most operations on an hdisk device can be performed on the SDD device, including commands such as open, close, dd, or fsck.


    Enhanced data availability

    Figure 1 shows that an SDD-residing host system is attached through SCSI or fibre-channel adapters to an ESS that has internal component redundancy and multipath configuration. SDD uses this multipath configuration to enhance data availability. That is, when there is a path failure, SDD reroutes I/O operations from the failing path to an alternate operational path. This capability prevents a single failing bus adapter on the host system, SCSI or fibre-channel cable, or host-interface adapter on the ESS from disrupting data access.

    Figure 1. Multipath connections between a host system and the disk storage in an ESS

    cmvc filename: s009000q


    Dynamic I/O load-balancing

    By distributing the I/O workload over multiple active paths, SDD provides dynamic load-balancing and eliminates data-flow bottlenecks. In the event of failure in one data path, SDD automatically switches the affected I/O operations to another active data path, ensuring path-failover protection.


    Automatic path-failover protection

    The SDD failover protection system is designed to minimize any disruptions in I/O operations and recover I/O operations from a failing data path. SDD provides path-failover protection through the following process:

    SDD dynamically selects an alternate I/O path when it detects a software or hardware problem.


    Concurrent download of licensed internal code

    With SDD you can concurrently download and install the licensed internal code while applications continue running. During the download and installation process, the host adapters inside the ESS might not respond to host I/O requests for approximately 30 seconds. SDD makes this process transparent to the host system through its path-selection and retry algorithms.


    Path-selection policies for the host system

    SDD uses similar path-selection algorithms on all the host systems. There are two modes of operation:

    single-path mode
    The host system has only one path that is configured to an ESS logical unit number (LUN). SDD, in single-path mode, has the following characteristics:

    multiple-path mode
    The host system has multiple paths that are configured to an ESS LUN. SDD in multiple-path mode has the following characteristics:

    Chapter 2. SDD for an AIX host system

    This chapter provides step-by-step procedures for you to install, configure, upgrade, and remove the Subsystem Device Driver on an AIX host system that is attached to an ESS.


    Command syntax conventions

    This section describes the notational conventions that are used in this chapter.

    Highlighting conventions

    The following typefaces are used to show emphasis:

    boldface
    Text in boldface represents menu items and command names.

    italics
    Text in italics is used to emphasize a word. In command syntax, it is used for variables for which you supply actual values.

    monospace
    Text in monospace identifies the commands that you type, samples of command output, examples of program code or messages from the system, and configuration status of the paths or volumes (such as, Dead, Active, Open, Closed, Online, Offline, Invalid, Available, Defined).

    Special characters conventions

    The following special character conventions are used in this chapter:

    * asterisks
    Asterisks (*) are used as wildcard symbols to search for the beginning or remaining characters of the installation package name.

    For example, the asterisks in the beginning and at the end of Sdd characters in the lslpp -l *Sdd* command are used as wildcard symbols to search for the characters ibm... and ...rte.

    ... ellipses
    Ellipses indicate that there are more commands in the next command line.

    Verifying the hardware and software requirements

    You must install the following hardware and software components to ensure that SDD installs and operates successfully.

    Hardware

    Software

    Non-supported environments

    SDD does not support the following environments:

    SDD does not support the following activities:

    SDD 1.3.0.1 (installed from ibmSdd_421.rte installation package) does not support the following dependency:

    SDD 1.3.1.3 (installed from ibmSdd_432.rte or ibmSdd_510.rte installation package) does not support the following dependency:

    Host system requirements

    To successfully install SDD, you must have AIX 4.2.1, AIX 4.3.2, AIX 4.3.3 or AIX 5.1.0 installed on your host system along with the program temporary fixes (PTFs) in Table 7.

    Table 7. AIX required fixes
    AIX level PTF number Component name Component level
    4.2.1 IX62304


    U451711 perfagent.tools 2.2.1.4

    U453402 bos.rte.libc 4.2.1.9

    U453481 bos.adt.prof 4.2.1.11

    U458416 bos.mp 4.2.1.15

    U458478 bos.rte.tty 4.2.1.14

    U458496 bos.up 4.2.1.15

    U458505 bos.net.tcp.client 4.2.1.19

    U462492 bos.rte.lvm 4.2.1.16
    4.3.2 U461953 bos.rte.lvm 4.3.2.4

    Tip:

    You must check for and download the latest APARs, maintenance level fixes, and microcode updates at the following Web site:

    service.software.ibm.com/support/rs6000

    ESS requirements

    SCSI requirements

    To use the SDD SCSI support, ensure that your host system meets the following requirements:

    For information about the SCSI adapters that can attach to your AIX host system, go to the following Web site:

    www.storage.ibm.com/hardsoft/products/ess/supserver.htm

    Fibre requirements

    You must check for and download the latest fibre-channel device driver APARs, maintenance level fixes, and microcode updates at the following Web site:

    techsupport.services.ibm.com/server/fixes

    For information about the fibre-channel adapters that can be used on your AIX host system go to the following Web site:

    www-1.ibm.com/servers/eserver/pseries/library/hardware_docs/

    To use the SDD fibre support, ensure your host system meets the following requirements:


    Preparing for SDD installation

    Before you install SDD, you must configure the ESS to your host system and the required fibre-channel adapters are attached.

    Configuring the ESS

    Before you install SDD, configure your ESS for single-port or multiple-port access for each LUN. SDD requires a minimum of two independent paths that share the same logical unit to use the load-balancing and fallover features.

    For more information about configuring your IBM Enterprise Storage Server, see IBM TotalStorage Enterprise Storage Server Introduction and Planning Guide.

    Note:
    Ensure the ibm2105.rte installation package is installed.

    Installing the AIX fibre-channel device drivers

    You must check for the latest information on fibre-channel device driver APARs, maintenance level fixes, and microcode updates at the following Web site:

    techsupport.services.ibm.com/server/fixes

    Perform the following steps to install the AIX fibre-channel device drivers:

    1. Install the fibre-channel device drivers from the AIX 4.3.3 compact disc. The fibre-channel device drivers include the following installation packages:

      devices.pci.df1000f9
      The adapter device driver for RS/6000 with feature code 6228.

      devices.pci.df1000f7
      The adapter device driver for RS/6000 with feature code 6227.

      devices.common.IBM.fc
      The FCP and SCSI protocol driver.

      devices.fcp.disk
      The FCP disk driver.
    2. Check to see if the correct APARS are installed by issuing the instfix -i | grep IY nnnnn command (where nnnnn represents the APAR numbers). If the APARS are listed, that means that they are installed. If they are installed, go to Configuring fibre-channel attached devices. Otherwise, go to step 3.
    3. Install the APARS.

    Uninstalling the AIX fibre-channel device drivers

    The following steps describe how to uninstall the AIX fibre-channel device drivers. There are two methods for uninstalling all of your fibre-channel device drivers. You can use of the following commands:

    Using the smitty deinstall command

    Perform the following steps to use the smitty deinstall command:

    1. Type smitty deinstall at the AIX command prompt and press Enter. The Remove Installed Software panel is displayed.
    2. Press F4. All of the software that is installed is displayed.
    3. Select the file name of the fibre-channel device driver that you want to uninstall. Press Enter. The selected file name is displayed in the Software Name Field of the Remove Installed Software panel.
    4. Use the Tab key to toggle to No in the PREVIEW Only? field. Press Enter. The uninstallation process begins.

    Using the installp command

    Perform the following steps to use the installp command from the AIX command line:

    1. Type installp -ug devices.pci.df1000f9 and press Enter.
    2. Type installp -ug devices.pci.df1000f7 and press Enter.
    3. Type installp -ug devices.common.IBM.fc and press Enter.
    4. Type installp -ug devices.fcp.disk and press Enter.

    Configuring fibre-channel attached devices

    The newly installed devices must be configured before they can be used. Use one of the following commands to configure these devices:

    After the system restarts, use the lsdev -Cc disk command to check the ESS fibre-channel protocol (FCP) disk configuration. If the FCP devices are configured correctly, they should be in the Available status. If the FCP devices are configured correctly, go to Verifying the Emulex adapter firmware level to determine if the proper firmware level is installed.

    Removing fibre-channel attached devices

    To remove all fibre-channel attached devices, you must issue the rmdev -dl fcsN -R command for each installed FCP adapter, where N is the FCP adapter number. For example, if you have two installed FCP adapters (adapter 0 and adapter 1), you must issue both of the folllowing commands:

    Verifying the Emulex adapter firmware level

    You must verify that your current adapter firmware is at the latest level. If your current adapter firmware is not at the latest level, you must upgrade to a new adapter firmware (microcode).

    Tip:

    Perform the following steps to verify the firmware level that is currently installed:

    1. Issue the lscfg -vl fcsN command. The vital product data for the adapter is displayed.
    2. Look at the ZB field. The ZB field should look something like this:
      +--------------------------------------------------------------------------------+
      |(ZB).............S2F3.22A0                                                      |
      +--------------------------------------------------------------------------------+

      To verify the firmware level, ignore the second character in the ZB field. In the example, the firmware level is sf322A0.

    3. If the adapter firmware level is at the latest level, there is no need to upgrade; otherwise, the firmware level must be upgraded. To upgrade the firmware level, go to Upgrading the Emulex adapter firmware level.

    Upgrading the Emulex adapter firmware level

    Upgrading the firmware level consists of downloading the firmware (microcode) from your AIX host system to the adapter. Before you upgrade the firmware, ensure you have fibre-channel attached devices (see Configuring fibre-channel attached devices). After the devices are configured, download the firmware from the AIX host system to the FCP adapter by performing the following steps:

    1. Verify that the correct level of firmware is installed on your AIX host system. Locate the file called df1000f7.131.320.320.320.503 which should be in the /etc/microcode directory. This file was copied into the /etc/microcode directory during the installation of the fibre-channel device drivers.
    2. From the AIX command prompt, type diag and press Enter.
    3. Highlight the Task Selection option.
    4. Highlight the Download Microcode option.
    5. Select all the fibre-channel adapters to which you want to download firmware. Press F7. The Download panel is displayed with one of the selected adapters highlighted. Press Enter to continue.
    6. Type the filename for the firmware that is contained in the /etc/microcode directory and press Enter; or use the Tab key to toggle to Latest.
    7. Follow the instructions that are displayed to download the firmware, one adapter at a time.
    8. After the download is complete, issue the lscfg -v -l fcsN command to verify the firmware level on each fibre-channel adapter.

    Determining the installation package

    Before you install SDD on your AIX 4.3.2, AIX 4.3.3, or AIX 5.1.0 host system, you need to determine the installation package that is appropriate for your environment.

    Table 8 lists and describes the SDD 1.3.13 support for 32-bit and 64-bit applications on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0.

    Table 9 lists the various scenarios and the installation packages that you must use for your environment.

    Support for 32-bit and 64-bit applications on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0 host systems


    Table 8. SDD 1.3.1.3 support for 32-bit and 64-bit applications on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0
    SDD installation package names
    AIX kernel mode Application mode SDD interface
    ibmSdd_432.rte 32-bit 32-bit, 64-bit LVM, raw device
    ibmSdd_433.rte 32-bit 32-bit, 64-bit LVM, raw device
    ibmSdd_510.rte 32-bit, 64-bit 32-bit, 64-bit LVM, raw device
    ibmSdd_510nchacmp.rte 32-bit, 64-bit 32-bit, 64-bit LVM, raw device

    Support for AIX 4.2.1 host system

    The installation package for AIX 4.2.1 (ibmSdd_421.rte) will not be enhanced after version 1.3.0.1. You will not receive future fixes for this installation package.

    Switching between 32-bit and 64-bit modes on AIX 5.1.x host systems

    SDD 1.3.1.3 supports AIX 5.1.x host systems that run in both 32-bit and 64-bit kernel modes. You can use the bootinfo -K or ls -al /unix command to check the current kernel mode in which your AIX 5.1.x host system is running.

    The bootinfo -K command directly returns the kernel mode information of your host system. The ls -al /unix command displays the /unix link information. If the /unix links to /usr/lib/boot/unix_mp, your AIX 5.1.x host system runs in 32-bit mode. If the /unix links to /usr/lib/boot/unix_64, your AIX 5.1.x host system runs in 64-bit mode.

    If your host system is currently running in 32-bit mode, you can switch it to 64-bit mode by typing the following commands in the given order:

    ln -sf /usr/lib/boot/unix_64 /unix
    ln -sf /usr/lib/boot/unix_64 /usr/lib/boot/unix
    bosboot -ak /usr/lib/boot/unix_64
    shutdown -Fr
    

    The kernel mode of your AIX host system is switched to 64-bit mode after the system restarts.

    If your host system is currently running in 64-bit mode, you can switch it to 32-bit mode by typing the following commands in the given order:

    ln -sf /usr/lib/boot/unix_mp /unix
    ln -sf /usr/lib/boot/unix_mp /usr/lib/boot/unix
    bosboot -ak /usr/lib/boot/unix_mp
    shutdown -Fr
    

    The kernel mode of your AIX host system is switched to 32-bit mode after the system restarts.

    Support for non-HACMP and HACMP environments on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0 host systems


    Table 9. SDD 1.3.1.3 installation packages for non-HACMP and HACMP environments on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0
    Host system
    Scenario Environment Use installation package
    AIX 4.3.2 1 Your server is not running HACMP ibmSdd_432.rte
    AIX 4.3.3 2 Your server is not running HACMP ibmSdd_432.rte
    3 Your server is running HACMP in concurrent mode ibmSdd_432.rte
    4 Your server is running HACMP in nonconcurrent mode, or both concurrent and nonconcurrent modes (mixed environment) ibmSdd_433.rte
    AIX 5.1.0 5 Your server is not running HACMP ibmSdd_510.rte
    6 Your server is running HACMP in concurrent mode ibmSdd_510.rte
    7 Your server is running HACMP in nonconcurrent mode, or both concurrent and nonconcurrent modes (mixed environment) ibmSdd_510nchacmp.rte

    Installation of major files on your AIX host system

    The installation package installs a number of major files on your AIX system.Table 10 lists the major files that are part of the SDD installation package.

    Table 10. Major files included in the SDD installation package
    Filename Description
    defdpo Define method of the SDD pseudo parent data path optimizer (dpo)
    cfgdpo Configure method of the SDD pseudo parent dpo
    define_vp Define method of the SDD vpath devices
    addpaths The command that dynamically adds more paths to Subsystem Device Driver devices while they are in Available status.
    Note:
    This command is not supported with Subsystem Device Driver for AIX 4.2.1; It is not Available if you have the ibmSdd_421.rte installation package installed. This feature only supports Subsystem Device Driver for AIX 4.3.2 and higher.
    cfgvpath Configure method of SDD vpath devices
    cfallvpath Fast-path configure method to configure the SDD pseudo parent dpo and all vpath devices.
    vpathdd SDD
    hd2vp The SDD script that converts an ESS hdisk device volume group to a Subsystem Device Drive vpath device volume group.
    vp2hd The SDD script that converts a SDD vpath device volume group to an ESS hdisk device volume group.
    datapath The SDD driver console command tool.
    lsvpcfg The SDD driver query configuration status command.
    mkvg4vp The command that creates a SDD volume group.
    extendvg4vp The command that extends SDD devices to a SDD volume group.
    dpovgfix The command that fixes a SDD volume group that has mixed vpath and hdisk physical volumes.
    savevg4vp The command that backs-up all files belonging to a specified volume group with SDD devices.
    restvg4vp The command that restores all files belonging to a specified volume group with SDD devices.


    Installing the Subsystem Device Driver

    SDD is released as an installation image. To install SDD, use the installation package that is appropriate for your environment.

    Table 8 lists and describes the SDD 1.3.13 support for 32-bit and 64-bit applications on AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0.

    Table 9 lists the various scenarios and the installation packages that you must use for your environment.

    Note:

    Use the System Management Interface Tool (SMIT) facility to install SDD. The SMIT facility has two interfaces, nongraphical (type smitty to invoke the nongraphical user interface) or graphical (type smit to invoke the graphical user interface).

    Tip:

    The list items on the SMIT panel might be worded differently from one AIX version to another.

    Throughout this SMIT procedure, /dev/cd0 is used for the compact disc drive address. The drive address can be different in your environment. Perform the following SMIT steps to install the SDD package on your system.

    1. Log in as the root user.
    2. Load the compact disc into the CD-ROM drive.
    3. From your desktop window, type smitty install_update and press Enter to go directly to the installation panels. The Install and Update Software menu is displayed.
    4. Highlight Install Software and press Enter.
    5. Press F4 to display the INPUT Device/Directory for Software panel.
    6. Select the compact disc drive that you are using for the installation, for example, /dev/cd0; and press Enter.
    7. Press Enter again. The Install Software panel is displayed.
    8. Highlight Software to Install and press F4. The Software to Install panel is displayed.
    9. Select the installation package that is appropriate for your environment.
    10. Press Enter. The Install and Update from LATEST Available Software panel is displayed with the name of the software you selected to install.
    11. Check the default option settings to ensure that they are what you need.
    12. Press Enter to install. SMIT responds with the following message:
      +--------------------------------------------------------------------------------+
      |     ARE YOU SURE??                                                             |
      |     Continuing may delete information you may want to keep.                    |
      |     This is your last chance to stop before continuing.                        |
      +--------------------------------------------------------------------------------+
    13. Press Enter to continue. The installation process can take several minutes to complete.
    14. When the installation is complete, press F10 to exit from SMIT. Remove the compact disc.

    Tip:

    The published AIX limitation on one system is 10,000 devices. The combined number of hdisk and vpath devices should not exceed the number of devices that AIX supports. In a multipath environment, since each path to a disk creates an hdisk, the total number of disks being configured can be reduced by the number of paths to each disk.

    Verifying your currently installed version of SDD

    You can verify your previously installed version of the SDD by issuing the following command:

    lslpp -l *Sdd*

    The asterisks (*) in the beginning and end of Sdd characters are used as wildcard symbols to search for the characters "ibm..." and "...rte"

    Alternatively, you can issue one of the following commands:

    lslpp -l ibmSdd_432.rte

    lslpp -l ibmSdd.rte.432

    ...

    ...

    If you have successfully installed the package, the output from the lslpp -l *Sdd* or lslpp -l ibmSdd_432.rte command looks like this:

    +--------------------------------------------------------------------------------+
    |Fileset                      Level  	 State        Description                  |
    |------------------------------------------------------------------------------  |
    |Path: /usr/lib/objrepos                                                         |
    |  ibmSdd_432.rte             1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V432 V433 for concurrent |
    |                                                   HACMP                        |
    |                                                                                |
    |Path: /etc/objrepos                                                             |
    |  ibmSdd_432.rte             1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V432 V433 for concurrent |
    |                                                   HACMP                        |
    |                                                                                |
    +--------------------------------------------------------------------------------+

    If you have successfully installed the ibmSdd_433.rte package, the output from the lslpp -l ibmSdd_433.rte command looks like this:

    +--------------------------------------------------------------------------------+
    |Fileset                      Level  	 State        Description                  |
    |--------------------------------------------------------------------------------|
    |Path: /usr/lib/objrepos                                                         |
    |  ibmSdd_433.rte             1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V433  for nonconcurrent  |
    |                                                   HACMP                        |
    |                                                                                |
    |Path: /etc/objrepos                                                             |
    |  ibmSdd_433.rte             1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V433  for nonconcurrent  |
    |                                                   HACMP                        |
    |                                                                                |
    +--------------------------------------------------------------------------------+

    If you have successfully installed the ibmSdd_510.rte package, the output from the lslpp -l ibmSdd_510.rte command looks like this:

    +--------------------------------------------------------------------------------+
    |Fileset                      Level  	 State        Description                  |
    |--------------------------------------------------------------------------------|
    |Path: /usr/lib/objrepos                                                         |
    |  ibmSdd_510.rte             1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V510 for concurrent HACMP|
    |                                                                                |
    |Path: /etc/objrepos                                                             |
    |  ibmSdd_510.rte             1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V510 for concurrent HACMP|
    |                                                                                |
    +--------------------------------------------------------------------------------+

    If you have successfully installed the ibmSdd_510nchacmp.rte package, the output from the lslpp -l ibmSdd_510nchacmp.rte command looks like this:

    +--------------------------------------------------------------------------------+
    |Fileset                      Level    State        Description                  |
    |--------------------------------------------------------------------------------|
    |Path: /usr/lib/objrepos                                                         |
    |  ibmSdd_510nchacmp.rte      1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V510 for nonconcurrent   |
    |                                                   HACMP                        |
    |                                                                                |
    |Path: /etc/objrepos                                                             |
    |  ibmSdd_510nchacmp.rte      1.3.1.3  COMMITTED    IBM Subsystem Device Driver  |
    |                                                   AIX V510 for nonconcurrent   |
    |                                                   HACMP                        |
    |                                                                                |
    +--------------------------------------------------------------------------------+

    Preparing to configure the Subsystem Device Driver

    Before you configure SDD, ensure that:

    Configure the ESS devices before you configure the SDD. If you configure multiple paths to an ESS device, make sure that all paths (hdisks) are in Available status. Otherwise, some SDD devices will lose multiple-path capability.

    Perform the following steps:

    1. Issue the lsdev -Cc disk | grep 2105 command to check the ESS device configuration.
    2. If you have already created some ESS volume groups, vary off (deactivate) all active volume groups with ESS subsystem disks by using the varyoffvg (LVM) command.

      Attention: Before you vary off a volume group, unmount all file systems in that volume group. If some ESS devices (hdisks) are used as physical volumes of an active volume group and if there are file systems of that volume group being mounted, you must unmount all file systems and vary off (deactivate) all active volume groups with ESS SDD disks.


    Configuring the Subsystem Device Driver

    Perform the following steps to configure SDD using SMIT:

    Note:
    The list items on the SMIT panel might be worded differently from one AIX version to another.
    1. Type smitty device from your desktop window. The Devices menu is displayed.
    2. Highlight Data Path Device and press Enter. The Data Path Device panel is displayed.
    3. Highlight Define and Configure All Data Path Devices and press Enter. The configuration process begins.
    4. Check the SDD configuration status. See Displaying the ESS vpath device configuration.
    5. Enter the varyonvg command to vary on all deactivated ESS volume groups.
    6. If you want to convert the ESS hdisk volume group to SDD vpath devices, you must run the hd2vp utility. (See hd2vp and vp2hd for information about this utility.)
    7. Mount the file systems for all volume groups that were previously unmounted.

    Unconfiguring Subsystem Device Drivers

    Before you unconfigure SDD devices, ensure that:

    Then, run the vp2hd conversion script to convert the volume group from SDD devices (vpathN) to ESS devices (hdisks).

    Note:
    If you are running HACMP with ibmSdd_433.rte or ibmSdd_510nchacmp.rte installation package installed on your host system, there are special requirements regarding unconfiguring and removing SDD vpath devices. See Special requirements.

    You can use SMIT to unconfigure the SDD devices in two ways. Either you can unconfigure without deleting the device information from the Object Database Management (ODM) database, or you can delete device information from the ODM database. If you unconfigure without deleting the device information, the device remains in the Defined status. You can use either SMIT or the mkdev -l vpathN command to return the device to the Available status.

    If you delete the device information from the ODM database, that device is removed from the system. To return it, follow the procedure described in "Configuring the Subsystem Device Driver".

    Perform the following steps to unconfigure SDD devices:

    Note:
    The list items on the SMIT panel might be worded differently from one AIX version to another.
    1. Type smitty device from your desktop window. The Devices menu is displayed.
    2. Highlight Devices and press Enter.
    3. Highlight Data Path Device and press Enter. The Data Path Device panel is displayed.
    4. Highlight Remove a Data Path Device and press Enter. A list of all SDD devices and their status (either Defined or Available) is displayed.
    5. Select the device that you want to unconfigure. Select whether or not you want to delete the device information from the ODM database.
    6. Press Enter. The device is unconfigured to the status that you selected.
    7. To unconfigure more SDD devices you have to repeat steps 4 - 6 for each SDD device.

    The fast-path command to unconfigure all SDD devices from the Available status to the Defined status is: rmdev -l dpo -R. The fast-path command to remove all Subsystem Device Driver devices from your system is: rmdev -dl dpo -R.

    Verifying the SDD configuration

    To check the SDD configuration, you can use either the SMIT Display Device Configuration panel or the lsvpcfg console command.

    Perform the following steps to verify the SDD configuration on an AIX host system:

    Note:
    The list items on the SMIT panel might be worded differently from one AIX version to another.
    1. Type smitty device from your desktop window. The Devices menu is displayed.
    2. Select Data Path Device and press Enter. The Data Path Device panel is displayed.
    3. Select Display Data Path Device Configuration and press Enter to display the status (either Defined or Available) of all SDD pseudo devices and the paths to each device.

    If any device is listed as Defined, the configuration was not successful. Check the configuration procedure again. See Configuring the Subsystem Device Driver for the procedure.

    Perform the following steps to verify that multiple paths are configured for each adapter connected to an ESS port:

    1. Type smitty device from your desktop window. The Devices menu is displayed.
    2. Highlight Data Path Device and press Enter. The Data Path Device panel is displayed.
    3. Highlight Display Data Path Device Adapter Status and press Enter. All attached paths for each adapter are displayed.

    If you want to use the command-line interface to verify the configuration, type lsvpcfg.

    You should see an output similar to this:

    +--------------------------------------------------------------------------------+
    |vpath0 (Avail pv vpathvg) 018FA067 = hdisk1 (Avail )                            |
    |vpath1 (Avail ) 019FA067 = hdisk2 (Avail )                                      |
    |vpath2 (Avail ) 01AFA067 = hdisk3 (Avail )                                      |
    |vpath3 (Avail ) 01BFA067 = hdisk4 (Avail ) hdisk27 (Avail )                     |
    |vpath4 (Avail ) 01CFA067 = hdisk5 (Avail ) hdisk28 (Avail )                     |
    |vpath5 (Avail ) 01DFA067 = hdisk6 (Avail ) hdisk29 (Avail )                     |
    |vpath6 (Avail ) 01EFA067 = hdisk7 (Avail ) hdisk30 (Avail )                     |
    |vpath7 (Avail ) 01FFA067 = hdisk8 (Avail ) hdisk31 (Avail )                     |
    |vpath8 (Avail ) 020FA067 = hdisk9 (Avail ) hdisk32 (Avail )                     |
    |vpath9 (Avail pv vpathvg) 02BFA067 = hdisk20 (Avail ) hdisk44 (Avail )          |
    |vpath10 (Avail pv vpathvg) 02CFA067 = hdisk21 (Avail ) hdisk45 (Avail )         |
    |vpath11 (Avail pv vpathvg) 02DFA067 = hdisk22 (Avail ) hdisk46 (Avail )         |
    |vpath12 (Avail pv vpathvg) 02EFA067 = hdisk23 (Avail ) hdisk47 (Avail )         |
    |vpath13 (Avail pv vpathvg) 02FFA067 = hdisk24 (Avail ) hdisk48 (Avail )         |
    +--------------------------------------------------------------------------------+

    The output shows:

    Changing the path-selection policy

    SDD supports path-selection policies that increase the performance of a multipath-configured ESS and make path failures transparent to applications. The following path-selection policies are supported:

    load balancing (lb)
    The path to use for an I/O operation is chosen by estimating the load on the adapter to which each path is attached. The load is a function of the number of I/O operations currently in process. If multiple paths have the same load, a path is chosen at random from those paths.

    round robin (rr)
    The path to use for each I/O operation is chosen at random from those paths that were not used for the last I/O operation. If a device has only two paths, SDD alternates between the two.

    fallover only (fo)
    All I/O operations for the device are sent to the same (preferred) path until the path fails because of I/O errors. Then an alternate path is chosen for subsequent I/O operations.

    The path-selection policy is set at the SDD device level. The default path-selection policy for a SDD device is load balancing. You can change the policy for a SDD device with the chdev command.

    Before changing the path-selection policy, determine the active attributes for the SDD device. Type the lsattr -El vpathN command. Press Enter, where N represents the vpath number. The output should look similar to this:

    +--------------------------------------------------------------------------------+
    |pvid         0004379001b90b3f0000000000000000 Data Path Optimizer Parent False  |
    |policy       df                               Scheduling Policy          True   |
    |active_hdisk hdisk1/30C12028                  Active hdisk               False  |
    |active_hdisk hdisk5/30C12028                                                    |
    +--------------------------------------------------------------------------------+

    The path-selection policy is the only attribute of an SDD device that can be changed. The valid policies are rr, lb, fo, and df. Here are the explanations for these policy values:

    rr
    round robin

    fo
    fallover only

    lb
    load balancing

    df
    (load balancing) default policy

    Attention: By changing an SDD device's attribute, the chdev command unconfigures and then reconfigures the device. You must ensure that the device is not in use if you are going to change its attribute. Otherwise, the command fails.

    Use the following command to change the SDD path-selection policy:

    chdev -l vpath N -a policy=[rr/fo/lb/df]
    

    Adding paths to SDD devices of a volume group

    You can add more paths to SDD devices that belong to a volume group after you have initially configured SDD. This section shows you how to add paths to SDD devices from AIX 4.2.1 and AIX 4.3.2 or higher host systems.

    Adding paths from AIX 4.3.2 or higher host systems

    If your host system is AIX 4.3.2 or higher, you can issue the addpaths command to add paths to SDD devices of a volume group.

    The addpaths command allows you to dynamically add more paths to SDD devices while they are in the Available status. It also allows you to add paths to vpath devices belonging to active volume groups.

    You can issue the addpaths command to add a new path to a vpath device that has only one existing path. But the new path is not automatically in the Open status; you must change it to the Open status by closing and reopening the vpath device.

    Before you issue the addpaths command, make sure that ESS logical volume sharing is enabled for all applicable devices. You can enable ESS logical volume sharing through the ESS Specialist. See IBM TotalStorage Enterprise Storage Server Web Interface User's Guide for more information.

    Complete the following steps to add paths to SDD devices of a volume group with the addpaths command:

    1. Issue the lspv command to list the physical volumes.
    2. Identify the volume group that contains the SDD devices to which you want to add more paths.
    3. Verify that all the physical volumes belonging to the SDD volume group are SDD devices (vpathNs). If they are not, you must fix the problem before proceeding to the next step. Otherwise, the entire volume group loses the path-fallover protection.

      You can issue the dpovgfix vg-name command to ensure that all physical volumes within the SDD volume group are SDD devices.

    4. Stop all I/O operations by stopping all applications to the volume group.

      The addpaths command is designed to add paths when there are no I/O activities. The command fails if it detects active I/Os.

    5. Run the AIX configuration manager in one of the following ways to recognize all new hdisk devices. Ensure that all logical drives on the ESS are identified as hdisks before continuing.
    6. Issue the addpaths from the AIX command line to add more paths to the SDD devices.
    7. Type the lsvpcfg command from the AIX command line to verify the configuration of the SDD devices in the volume group.

      SDD devices should show two or more hdisks associated with each SDD device when the fallover protection is required.

    Adding paths from AIX 4.2.1 host systems

    Tip: This addpaths command is not supported with Subsystem Device Driver for AIX 4.2.1

    To activate additional paths to a SDD device, the related SDD devices must be unconfigured and then reconfigured. The SDD conversion scripts should be run to enable the necessary SDD associations and links between the SDD vpath (pseudo) devices and the ESS hdisk devices.

    Note:
    Ensure that logical volume sharing is enabled at the ESS for all applicable devices. Logical volume sharing is enabled using the ESS Specialist. See IBM TotalStorage Enterprise Storage Server Web Interface User's Guide for information about enabling volume sharing.

    Perform the following steps to activate additional paths to SDD devices of a volume group from your AIX 4.2.1 host system:

    1. Identify the volume groups containing the SDD devices to which you want to add additional paths. Type the following command:
      lspv
      
    2. Check if all the physical volumes belonging to that SDD volume group are SDD devices (vpathNs). If they are not, you need to fix the problem.

      Attention: You must fix this problem with the volume group before proceeding to step 3. Otherwise, the volume group loses path fallover capability. To fix the problem, type the following command:

      dpovgfix vg-name
      
      Vg-name represents the volume group.
    3. Identify the associated file systems for the selected volume group. Type the following command:
      
      
      lsvgfs vg-name
      
    4. Identify the associated mounted file systems for the selected volume group. Type the following command:
      mount
      
    5. Unmount the file systems of the selected volume group listed in step 3. Type the following command:
      
      
      umount mounted-filesystem
      
    6. Run the vp2hd volume group conversion script to convert the volume group from SDD devices to ESS hdisk devices. Type the following command to run the script:
      vp2hd vg-name
      
      When the conversion script completes, the volume group is in the Active status (varied on).
    7. Vary off the selected volume group in preparation for SDD reconfiguration. Type the following command:
      
      
      varyoffvg  vg-name
      
    8. Run the AIX configuration manager cfgmgr to recognize all new hdisk devices. You can do this in one of two ways:
      Note:
      Ensure that all logical drives on the ESS are identified as hdisks before continuing.
    9. Unconfigure affected SDD devices to the Defined status by using the rmdev -l vpathN command; where N represents the vpath-number you want to set to the Defined status N=[0,1,2,...]. This command allows you to unconfigure only SDD devices for which you are adding paths.
      Note:
      Issue the rmdev -l dpo -R command if you need to unconfigure all Subsystem Device Driver devices. SDD volume groups must be inactive before unconfiguring. This command attempts to unconfigure all SDD devices recursively.
    10. Reconfigure SDD devices by using either SMIT or the command-line interface.

      If you are using SMIT, perform the following steps:

      Note:
      The list items on the SMIT panel might be worded differently from one AIX version to another.
      1. Type Smitty device from your desktop window. The Devices menu is displayed.
      2. Highlight Data Path Devices and press Enter. The Data Path Devices menu is displayed.
      3. Highlight Define and Configure All Data Path Devices and press Enter. SMIT executes a script to define and configure all SDD devices that are in the Defined status.
      If you are using the command-line interface, type the mkdev -l vpathN command for each SDD device or type the cfallvpath command to configure all SDD devices.
    11. Verify your datapath configuration using either SMIT or the command-line interface.

      If you are using SMIT, perform the following steps:

      Note:
      The list items on the SMIT panel might be worded differently from one AIX version to another.
      1. Type Smitty device from your desktop window. The Devices menu is displayed.
      2. Highlight Data Path Devices and press Enter. The Data Path Devices menu is displayed.
      3. Highlight Display Data Path Device Configuration and press Enter.
      If you are using the command-line interface, type the lsvpcfg command to display the SDD configuration status.

      SDD devices should show two or more hdisks associated with each SDD device when fallover protection is required.

    12. Vary on the volume groups selected in step 3. Type the following command:
      
      
      varyonvg  vg-name
      
    13. Run the hd2vp script to convert the volume group from ESS hdisk devices back to SDD vpath devices. Type the following command:
      
      
      hd2vp vg-name
      
    14. Mount all file systems for the volume groups that were previously unmounted.

    Upgrading SDD for AIX 4.2.1, AIX 4.3.2, AIX 4.3.3, and AIX 5.1.0

    Understanding the SDD support for nondisruptive installation

    Attention: The nondisruptive installation requires that you:

    With the nondisruptive installation capability, all configurations are automatically updated during the system restart time. The system restart will start all the automatic configuration, which loads the new driver into the kernel of the host system. The nondisruptive installation capability is beneficial if you have a large number of SDD devices, volume groups, and file systems configured.

    Without the nondisruptive installation capability, you must manually perform the following tasks when upgrading to a newer version of SDD:

    SDD 1.3.0.1 (or later) supports nondisruptive installation if you are upgrading from any one of the installation packages listed in Table 11.

    Table 11. List of previously installed installation packages that are supported with nondisruptive installation
    Installation package name
    ibmSdd_421.rte
    ibmSdd.rte.421
    ibmSdd_432.rte
    ibmSdd.rte.432
    ibmSdd_433.rte
    ibmSdd.rte.433
    ibmSdd_510.rte
    ibmSdd_510nchacmp.rte

    If you have previously installed from any of the installation packages listed in Table 11, with SDD 1.3.0.1 (or later) you can upgrade while all of the SDD:

    Note:
    The installation package for AIX 4.2.1 (ibmSdd_421.rte) is not updated after version 1.3.0.1.

    If you are upgrading from a previous version of SDD that you installed from other installation packages, you cannot perform the nondisruptive installation. To upgrade SDD to a newer version, all the SDD installation packages must be uninstalled.

    To verify your currently installed version of SDD , go to Verifying your currently installed version of SDD.

    If the previous version of SDD is installed from one of the installation packages listed in Table 11, proceed to Upgrading SDD using a nondisruptive installation.

    If the previous version of SDD is installed from a installation package not listed in Table 11, proceed to Upgrading SDD manually.

    Note:
    The installation package for AIX 4.2.1 (ibmSdd_421.rte) is not updated after version 1.3.0.1.

    Verifying your previously installed version of SDD

    You can verify your previously installed version of the SDD by issuing the following command:

    lslpp -l *Sdd*

    (where the asterisks (*) in the beginning and end of Sdd characters are used as wildcard symbols to search for the characters "ibm..." and "...rte")

    Alternatively, you can issue the following commands:

    lslpp -l ibmSdd_432.rte

    lslpp -l ibmSdd.rte.432

    ...

    ...

    
    

    Upgrading SDD using a nondisruptive installation

    You can use SDD 1.3.0.1 (or later) using a nondisruptive installation if you are upgrading from any of the installation packages listed in Table 11. See Understanding the SDD support for nondisruptive installation for more information.

    Note:
    The installation package for AIX 4.2.1 (ibmSdd_421.rte) is not updated after version 1.3.0.1.

    Perform the following steps to upgrade SDD with a nondisruptive installation:

    1. Terminate all I/O operations to the SDD volume groups.
    2. Complete the installation instructions provided in Installing the Subsystem Device Driver .
    3. Restart your system by typing shutdown -rF
    4. Verify the SDD configuration by typing lsvpcfg
    5. Verify your currently installed version of the SDD by completing the instructions provided in Verifying your currently installed version of SDD

      Attention: If the physical volumes on an SDD volume group are mixed with hdisk devices and vpath devices, you must run the dpovgfix utility to fix this problem. Otherwise, SDD will not function properly. Issue the dpovgfix vg_name command to fix this problem.

    Upgrading SDD manually

    If you are upgrading from a previous version of the SDD that you installed with a installation package not listed in Table 11, you cannot perform a nondisruptive installation.

    Note:
    The installation package for AIX 4.2.1 (ibmSdd_421.rte) is not updated after version 1.3.0.1.

    Perform the following steps to upgrade SDD:

    1. Remove any .toc files generated during previous SDD or DPO installations. Type the following command to delete any .toc file found in the /usr/sys/inst.images directory:

      rm .toc

      Note:
      Ensure that this file is removed because it contains information about the previous version of SDD or DPO.
    2. Issue the lspv command to find out all the Subsystem Device Driver volume groups.
    3. Issue the lsvgfs command for each SDD volume group to find out which file systems are mounted. Type the following command:
      lsvgfs  vg_name
      
    4. Issue the umount command to unmount all file systems belonging to SDD volume groups. Type the following command:
      umount  filesystem_name
      
    5. Run the vp2hd script to convert the volume group from SDD devices to ESS hdisk devices.
    6. Issue the varyoffvg command to vary off the volume groups. Type the following command:
      varyoffvg  vg_name
      
    7. Remove all SDD devices. Type the following command:
      rmdev  -dl dpo -R
      
    8. Use the smitty command to uninstall SDD. Type smitty deinstall and press Enter. The uninstall process begins. Complete the uninstall process. See Removing SDD from an AIX host system for the step-by-step procedure for uninstalling SDD.
    9. Use the smitty command to install the newer version of SDD from the compact disc. Type smitty install and press Enter. The installation process begins. Go to Installing the Subsystem Device Driver to complete the installation process.
    10. Use the smitty device command to configure all the SDD devices to the Available status. See Configuring the Subsystem Device Driver for a step-by-step procedure for configuring devices.
    11. Issue the lsvpcfg command to verify the SDD configuration. Type the following command:
      lsvpcfg
      
    12. Issue the varyonvg command for each volume group that was previously varied offline. Type the following command:
      varyonvg vg_name
      
    13. Run the hd2vp script for each SDD volume group to convert the physical volumes from ESS hdisk devices back to SDD vpath devices. Type the following command:
      
      
      hd2vp  vg_name
      
    14. Issue the lspv command to verify that all physical volumes of the SDD volume groups are SDD vpath devices.

    Attention: If the physical volumes on an SDD volume group's physical volumes are mixed with hdisk devices and vpath devices, you must run the dpovgfix utility to fix this problem. Otherwise, SDD will not function properly. Issue the dpovgfix vg_name command to fix this problem.


    Removing SDD from an AIX host system

    Before you remove the SDD package from your AIX host system, all the SDD devices must be unconfigured and removed from your host system.

    Note:
    See Unconfiguring Subsystem Device Drivers.
    The fast-path rmdev -dl dpo -R command removes all the SDD devices from your system. After all SDD devices are removed, perform the following steps to remove SDD.
    1. Type smitty deinstall from your desktop window to go directly to the Remove Installed Software panel.
    2. Type one of the following installation package names in the SOFTWARE name field:

      ibmSdd_421.rte

      ibmSdd_432.rte

      ibmSdd_433.rte

      ibmSdd_510.rte

      ibmSdd_510nchacmp.rte

      Then press Enter.

      Note:
      See Verifying your currently installed version of SDD to verify your currently installed installation package or version of SDD; You can also press F4 in the Software name field to list the currently installed installation package or version of SDD.

      They could also hit F4 in the SOFTWARE name field and do a search (/) on Sdd.

    3. Press the Tab key in the PREVIEW Only? field to toggle between Yes and No. Select No to remove the software package from your AIX host system.
      Note:
      If you select Yes, the process stops at this point and previews what you are removing. The results of your pre-check are displayed without removing the software. If the status for any SDD device is either Available or Defined, the process fails.
    4. Select No for the remaining fields on this panel.
    5. Press Enter. SMIT responds with the following message:
      +--------------------------------------------------------------------------------+
      |     ARE YOU SURE??                                                             |
      |     Continuing may delete information you may want to keep.                    |
      |     This is your last chance to stop before continuing.                        |
      +--------------------------------------------------------------------------------+
    6. Press Enter to begin the removal process. This might take a few minutes.
    7. When the process is complete, the SDD software package is removed from your system.

    Using concurrent download of licensed internal code

    Concurrent download of licensed internal code is the capability to download and install licensed internal code on an ESS while applications continue to run. This capability is supported for single-path (SCSI only) and multiple-path (SCSI or FCP) access to an ESS.

    Attention: During the download of licensed internal code, the AIX error log might overflow and excessive system paging space could be consumed. When the system paging space drops too low it could cause your AIX system to hang. To avoid this problem, you can perform the following steps prior to doing the download:

    1. Save the existing error report by typing the following command from the AIX command-line interface:
      > errpt > file.save
      
    2. Delete the error log from the error log buffer by typing the following command:
      > errclear 0
      
    3. Enlarge the system paging space by using the SMIT tool.
    4. Stop the AIX error log daemon by typing the following command:
      
      
      /usr/lib/errstop
      

    Once you have completed steps 1- 4, you can perform the download of the ESS licensed internal code. After the download completes, type /usr/lib/errdemon from the command-line interface to restart the AIX error log daemon.


    Understanding the SDD support for High Availability Cluster Multi-Processing (HACMP)

    You can run the Subsystem Device Driver in concurrent and nonconcurrent multihost environments in which more than one host is attached to the same LUNs on an ESS. RS/6000 servers running HACMP in concurrent or nonconcurrent mode are supported. Different SDD releases support different kinds of environments. (See Table 12, Table 14, Table 13 and Table 15.)

    HACMP provides a reliable way for clustered IBM RS/6000 servers which share disk resources to recover from server and disk failures. In a HACMP environment, each RS/6000 server in a cluster is a node. Each node has access to shared disk resources that are accessed by other nodes. When there is a failure, HACMP transfers ownership of shared disks and other resources based on how you define the relationship among nodes in a cluster. This process is known as node fallover or node fallback. HACMP supports two modes of operation:

    nonconcurrent
    Only one node in a cluster is actively accessing shared disk resources while other nodes are standby.

    concurrent
    Multiple nodes in a cluster are actively accessing shared disk resources.

    Tip: If you use a mix of nonconcurrent and concurrent resource groups (such as, cascading and concurrent resource groups or rotating and concurrent resource groups) with HACMP, you should use the nonconcurrent version of SDD.

    HACMP running in concurrent mode is supported with the ibmSdd_432.rte installation package for SDD 1.1.4 (SCSI only). HACMP running in concurrent mode is supported with the ibmSdd_432.rte installation package for SDD 1.3.1.3 or later (SCSI and fibre) and ibmSdd_510.rte installation package for SDD 1.2.0.0 or later (SCSI and fibre). The ibmSdd_433.rte installation package for SDD 1.3.1.3 (or later) and the ibmSdd_510nchacmp.rte installation package for SDD 1.3.1.3 are for HACMP environments only. These versions support nonconcurrent modes. However, in order to make the best use of the manner in which the device reserves are made, IBM recommends that you:

    HACMP is not supported on all models of the ESS. For information about supported ESS models and required ESS microcode levels, go to the following Web site:
    www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates

    SDD supports RS/6000 servers connected to shared disks with SCSI adapters and drives as well as FCP adapters and drives. The kind of attachment support depends on the version of SDD that you have installed. Table 12 and Table 14 summarizes the software requirements to support HACMP:

    Table 12. Software support for HACMP in concurrent mode
    Subsystem Device Driver Version and Release Level HACMP 4.3.1 + APARs HACMP 4.4 + APARs
    SDD 1.1.4.0 (SCSI only)
    • IY07392
    • IY03438
    • IY11560
    • IY08933
    • IY11564
    • IY12021
    • IY12056
    • F model requires IY11110

    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • F model requires IY11480

    SDD 1.2.0.0 (SCSI/FCP)
    • IY07392
    • IY13474
    • IY03438
    • IY08933
    • IY11560
    • IY11564
    • IY12021
    • IY12056
    • F model requires IY11110

    • IY13432
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • F model requires IY11480

    SDD 1.2.2.x (SCSI/FCP)
    • IY07392
    • IY13474
    • IY03438
    • IY08933
    • IY11560
    • IY11564
    • IY12021
    • IY12056
    • F model requires IY11110

    • IY13432
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • F model requires IY11480

    SDD 1.3.1.3 (SCSI/FCP)
    • IY07392
    • IY13474
    • IY03438
    • IY08933
    • IY11560
    • IY11564
    • IY12021
    • IY12056
    • F model requires IY11110

    • IY13432
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • F model requires IY11480


    Table 13. Software support for HACMP in nonconcurrent mode
    Subsystem Device Driver Version and Release Level HACMP 4.3.1 + APARs HACMP 4.4 + APARs
    SDD 1.2.2.x (SCSI/FCP)
    • IY07392
    • IY13474
    • IY03438
    • IY08933
    • IY11560
    • IY11564
    • IY12021
    • IY12056
    • IY14682
    • F model requires IY11110

    • IY13432
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • IY14683
    • F model requires IY11480

    ibmSdd_433.rte installation package for SDD 1.3.1.3 (SCSI/FCP)
    • IY07392
    • IY13474
    • IY03438
    • IY08933
    • IY11560
    • IY11564
    • IY12021
    • IY12056
    • IY14682
    • F model requires IY11110

    • IY13432
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • IY14683
    • F model requires IY11480


    Table 14. Software support for HACMP in concurrent mode on AIX 5.1.0 (32-bit kernel only)
    Subsystem Device Driver Version and Release Level HACMP 4.4 + APARs
    ibmSdd_510.rte installation package for SDD 1.3.1.3 (SCSI/FCP)
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • IY13432
    • IY14683
    • IY17684
    • IY19089
    • IY19156
    • F model requires IY11480


    Table 15. Software support for HACMP in nonconcurrent mode on AIX 5.1.0 (32-bit kernel only)
    Subsystem Device Driver Version and Release Level HACMP 4.4 + APARs
    ibmSdd_510nchacmp.rte installation package for SDD 1.3.1.3 (SCSI/FCP)
    • IY11563
    • IY11565
    • IY12022
    • IY12057
    • IY13432
    • IY14683
    • IY17684
    • IY19089
    • IY19156
    • F model requires IY11480

    Note:
    For the most up-to-date list of required APARs go to the following Web sites:

    Even though SDD supports HACMP, certain combinations of features are not supported. Table 16 lists those combinations:

    Table 16. HACMP and supported SDD features
    Feature RS/6000 node running HACMP
    ESS concurrent code load Yes
    Subsystem Device Driver load balancing Yes
    SCSI Yes
    FCP (fibre) Yes
    Single-path fibre No
    SCSI and fibre-channel connections to the same LUN from one host (mixed environment) No

    SDD installation package attributes

    The ibmSdd_433.rte and ibmSdd_510nchacmp.rte installation packages for SDD 1.3.1.3 have different features compared with ibmSdd_432.rte and ibmSdd_510.rte installation packages for SDD 1.3.1.3. The ibmSdd_433.rte and ibmSdd_510nchacmp.rte installation packages implement the SCSI-3 Persistent Reserve command set, in order to support HACMP in nonconcurrent mode with single-point failure protection. The ibmSdd_433.rte and ibmSdd_510nchacmp.rte installation packages require the ESS G3 level microcode on the ESS to support the SCSI-3 Persistent Reserve command set. If the ESS G3 level microcode is not installed, the ibmSdd_433.rte and ibmSdd_510nchacmp.rte installation packages will switch the multipath configuration to a single-path configuration. There is no single-point failure protection for single-path configurations.

    The ibmSdd_433.rte and ibmSdd_510nchacmp.rte installation packages have a new attribute under its pseudo parent (dpo), that reflects whether the ESS supports the Persistent Reserve Command set or not. The attribute name is persistent_resv. If SDD detects that G3 level microcode is installed, the persistent_resv attribute is created in the CuAt ODM and its value is set to yes; otherwise this attribute only exists in the PdAt ODM and its value is set to no (default). You can use the following command to check the persistent_resv attribute, after the SDD device configuration is complete:

    odmget  -q  "name = dpo"  CuAt
    

    If your attached ESS has the G3 microcode, the output should look similar to this:

    			name  =  "dpo"
    			attribute =  "persistent_resv"
    			value  =  "yes"
    			generic  =  "D"
    			rep  =  "sl"
    			nls_index  =  0
    

    In order to implement the Persistent Reserve command set, each host server needs a unique 8-byte reservation key. There are 2 ways to get a unique reservation key. In HACMP environments, HACMP generates a unique key for each node in the ODM database. When SDD cannot find that key in the ODM database, it generates a unique reservation key by using the middle 8 bytes of the output from the uname -m command.

    To check the Persistent Reserve Key of a node, provided by HACMP, type the command:

    odmget  -q  "name = ioaccess" CuAt
    

    The output should look similar to this:

          name = "ioaccess"
          attribute = "perservekey"
          value = "01043792"
          type = "R"
          generic = ""
          rep = "s"
          nls_index = 0
    

    Special requirements

    There is a special requirement regarding unconfiguring and removing the ibmSdd_433.rte and ibmSdd_510nchacmp.rte installation packages for theSDD 1.3.1.3 vpath devices. You must unconfigure and remove the vpath devices before you unconfigure and remove the vpath devices' underlying ESS hdisks. Otherwise, if the ESS hdisks are unconfigured and removed first, the persistent reserve will not be released on the physical devices, even though the vpath devices have been successfully unconfigured and removed.

    SDD does not automatically create the pvid attribute in the ODM database for each vpath device. The AIX disk driver automatically creates the pvid attribute in the ODM database, if a pvid exists on the physical device; however, SDD does not. Therefore, the first time you import a new SDD volume group to a new cluster node, you must import the volume group using hdisks as physical volumes. Next, run the hd2vp conversion script (see SDD utility programs) to convert the volume group's physical volumes from ESS hdisks to vpath devices. This conversion step not only creates pvid attributes for all vpath devices which belong to that imported volume group, it also deletes the pvid attributes for these vpath devices' underlying hdisks. Later on you can import and vary on the volume group directly from the vpath devices. These special requirements apply to both concurrent and nonconcurrent volume groups.

    Under certain conditions, the status of a physical device's pvid on a system is not always as expected. So it is necessary to determine the status of a pvid as displayed by the lspv command, in order to select the appropriate import volume group action.

    There are four scenarios:

    Scenario 1. lspv displays pvid's for both hdisks and vpath:

    	>lspv
    	hdisk1		003dfc10a11904fa	None
    	hdisk2		003dfc10a11904fa	None
    	vpath0		003dfc10a11904fa	None
    

    Scenario 2. lspv displays pvid's for hdisks only:

    	>lspv
    	hdisk1		003dfc10a11904fa	None
    	hdisk2		003dfc10a11904fa	None
    	vpath0		none					None
    

    For both Scenario 1 and Scenario 2, the volume group should be imported using the hdisk names and then converted using the hd2vp command:

    	>importvg -y vg_name -V major# hdisk1
    	>hd2vp vg_name
    

    Scenario 3. lspv displays the pvid for vpath only:

    	>lspv
    	hdisk1		none					None
    	hdisk2		none					None
    	vpath0		003dfc10a11904fa	None
    

    For Scenario 3, the volume group should be imported using the vpath name:

    	>importvg -y vg_name -V major# vpath0
    

    Scenario 4. lspv does not display the pvid on the hdisks or the vpath:

    	>lspv
    	hdisk1		none			None
    	hdisk2		none			None
    	vpath0		none			None
    

    For Scenario 4, the pvid will need to be placed in the odm for the vpath devices and then the volume group can be imported using the vpath name:

    	>chdev -l vpath0 -a pv=yes
    	>importvg -y vg_name -V major# vpath0
    
    Note:
    See Importing a volume group with SDD for a detailed procedure for importing a volume group with the SDD devices.

    Recovering paths that are lost during HACMP node fallover

    Normally, when there is a node failure, HACMP transfers ownership of shared disks and other resources, through a process known as node fallover. Certain situations, such as, a loose or disconnected SCSI or fibre-adapter card, can cause your vpath devices to lose one or more underlying paths during node fallover. Perform the following steps to recover these paths:

    If your vpath devices have lost one or more underlying paths that belong to an active volume group, you can use either the Add Paths to Available Data Path Devices SMIT panel or run the addpaths command from the AIX command line to recover the lost paths. Go to Adding paths to SDD devices of a volume group for more information about the addpaths command.

    Note:
    Simply running the cfgmgr command while the vpath devices are in the Available status will not recover the lost paths; that is why you need to run the addpaths command to recover the lost paths.

    SDD does not support the addpaths command for AIX 4.2.1; it is not available if you have the ibmSdd_421.rte installation package installed (this feature only supports SDD for AIX 4.3.2 and higher.) If you have the ibmSdd_421.rte installation package installed, and if your vpath devices have lost one or more underlying paths and they belong to an active volume group, perform the following steps to recover the lost paths:

    Note:
    1. Issue the lspv command to find the volume group name for the vpath devices that have lost paths.
    2. Issue the lsvgfs vg-name command to find out the file systems for the volume group.
    3. Issue the mount command to find out if any file systems of the volume group were mounted. Issue the umount filesystem-name command to unmount any file systems that were mounted.
    4. Issue the vp2hd vg-name command to convert the volume group's physical volumes from vpath devices to ESS hdisks.
    5. Vary off the volume group. This puts the physical volumes (hdisks) in the Closed status.
    6. Issue the rmdev -l vpathN command on each vpath device that has lost a path; run the mkdev -l vpathN command on the same vpath devices to recover the paths.
    7. Issue the lsvpcfg or lsvpcfg vpathN0 vpathN1 vpathN2 command to ensure that all the paths are configured.
    8. Vary on the volume group.
    9. Issue the hd2vp vg-name command to convert the volume group's physical volumes back to SDD vpath devices.
    10. Mount all the file systems which were unmounted at 3.

    Providing load-balancing and fallover protection

    SDD provides load-balancing and fallover protection for AIX applications and for the LVM when ESS vpath devices are used. These devices must have a minimum of two paths to a physical logical unit number (LUN) for fallover protection to exist.

    Displaying the ESS vpath device configuration

    To provide fallover protection, an ESS vpath device must include a minimum of two paths. Both the SDD vpath device and the ESS hdisk devices must all be in the Available status. In the following example, vpath0, vpath1, and vpath2 all have a single path and, therefore, will not provide fallover protection because there is no alternate path to the ESS LUN. The other SDD vpath devices have two paths and, therefore, can provide fallover protection.

    To display which ESS vpath devices are available to provide fallover protection, use either the Display Data Path Device Configuration SMIT panel, or run the lsvpcfg command. Perform the following steps to use SMIT:

    Note:
    The list items on the SMIT panel might be worded differently from one AIX version to another.
    1. Type smitty device from your desktop window. The Devices panel is displayed.
    2. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    3. Select Display Data Path Device Configuration and press Enter.

    You will see an output similar to the following:

    Figure 2. Output from the Display Data Path Device Configuration SMIT panel

    +--------------------------------------------------------------------------------+
    |vpath0 (Avail pv vpathvg) 018FA067 = hdisk1 (Avail )                            |
    |vpath1 (Avail ) 019FA067= hdisk2 (Avail )                                       |
    |vpath2 (Avail ) 01AFA067 = hdisk3 (Avail )                                      |
    |vpath3 (Avail ) 01BFA067 = hdisk4 (Avail ) hdisk27 (Avail )                     |
    |vpath4 (Avail ) 01CFA067 = hdisk5 (Avail ) hdisk28 (Avail )                     |
    |vpath5 (Avail ) 01DFA067 = hdisk6 (Avail ) hdisk29 (Avail )                     |
    |vpath6 (Avail ) 01EFA067 = hdisk7 (Avail ) hdisk30 (Avail )                     |
    |vpath7 (Avail ) 01FFA067 = hdisk8 (Avail ) hdisk31 (Avail )                     |
    |vpath8 (Avail ) 020FA067 = hdisk9 (Avail ) hdisk32 (Avail )                     |
    |vpath9 (Avail pv vpathvg) 02BFA067 = hdisk20 (Avail ) hdisk44 (Avail )          |
    |vpath10 (Avail pv vpathvg) 02CFA067 = hdisk21 (Avail ) hdisk45 (Avail )         |
    |vpath11 (Avail pv vpathvg) 02DFA067 = hdisk22 (Avail ) hdisk46 (Avail )         |
    |vpath12 (Avail pv vpathvg) 02EFA067 = hdisk23 (Avail ) hdisk47 (Avail )         |
    |vpath13 (Avail pv vpathvg) 02FFA067 = hdisk24 (Avail ) hdisk48 (Avail )         |
    +--------------------------------------------------------------------------------+

    The following information is displayed:

    You can also use the datapath command to display information about an SDD vpath device. This command displays the number of paths to the device. For example, the datapath query device 10 command might produce this output:

    +--------------------------------------------------------------------------------+
    |DEV#:  10  DEVICE NAME: vpath10   TYPE: 2105B09   SERIAL: 02CFA067              |
    |==================================================================              |
    |Path#      Adapter/Hard Disk   State     Mode    Select        Errors           |
    |    0          scsi6/hdisk21    OPEN   NORMAL        44             0           |
    |    1          scsi5/hdisk45    OPEN   NORMAL        43             0           |
    +--------------------------------------------------------------------------------+

    The sample output shows that device vpath10 has two paths and both are operational.

    Configuring a volume group for fallover protection

    You can create a volume group with SDD vpath devices using the Logical Volume Groups SMIT panel. Choose the SDD vpath devices that have fallover protection for the volume group.

    It is possible to create a volume group that has only a single path (see Figure 2) and then add paths later by reconfiguring the ESS. (See Adding paths to SDD devices of a volume group for information about adding paths to a SDD device.) However, an SDD volume group does not have fallover protection if any of its physical volumes only has a single path.

    Perform the following steps to create a new volume group with SDD vpaths:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Group and press Enter. The Add Volume Group with Data Path Devices panel is displayed.
    5. Select Add Volume Group with Data Path Devices and press Enter.
      Note:
      Press F4 while highlighting the PHYSICAL VOLUME names field to list all the available SDD vpaths.

    If you use a script file to create a volume group with SDD vpath devices, you must modify your script file and replace the mkvg command with the mkvg4vp command.

    All the functions that apply to a regular volume group also apply to an SDD volume group. Use SMIT to create a logical volume group (mirrored, striped, or compressed) or a file system (mirrored, striped, or compressed) on an SDD volume group.

    Once you create the volume group, AIX creates the SDD vpath device as a physical volume (pv). In Figure 2, vpath9 through vpath13 are included in a volume group and they become physical volumes. To list all the physical volumes known to AIX, use the lspv command. Any ESS vpath devices that were created into physical volumes are included in the output similar to the following:

    +--------------------------------------------------------------------------------+
    |hdisk0         0001926922c706b2    rootvg                                       |
    |hdisk1         none                None                                         |
    |...                                                                             |
    |hdisk10        none                None                                         |
    |hdisk11        00000000e7f5c88a    None                                         |
    |...                                                                             |
    |hdisk48        none                None                                         |
    |hdisk49        00000000e7f5c88a    None                                         |
    |vpath0         00019269aa5bc858    None                                         |
    |vpath1         none                None                                         |
    |vpath2         none                None                                         |
    |vpath3         none                None                                         |
    |vpath4         none                None                                         |
    |vpath5         none                None                                         |
    |vpath6         none                None                                         |
    |vpath7         none                None                                         |
    |vpath8         none                None                                         |
    |vpath9         00019269aa5bbadd    vpathvg                                      |
    |vpath10        00019269aa5bc4dc    vpathvg                                      |
    |vpath11        00019269aa5bc670    vpathvg                                      |
    |vpath12        000192697f9fd2d3    vpathvg                                      |
    |vpath13        000192697f9fde04    vpathvg                                      |
    +--------------------------------------------------------------------------------+

    To display the devices that comprise a volume group, enter the lsvg -p vg-name command. For example, the lsvg -p vpathvg command might produce the following output:

    +--------------------------------------------------------------------------------+
    |PV_NAME           PV STATE    TOTAL PPs   FREE PPs    FREE DISTRIBUTION         |
    |vpath9            active      29          4           00..00..00..00..04        |
    |vpath10           active      29          4           00..00..00..00..04        |
    |vpath11           active      29          4           00..00..00..00..04        |
    |vpath12           active      29          4           00..00..00..00..04        |
    |vpath13           active      29          28          06..05..05..06..06        |
    +--------------------------------------------------------------------------------+

    The example output indicates that the vpathvg volume group uses physical volumes vpath9 through vpath13.

    Importing a volume group with SDD

    You can import a new volume group definition from a set of physical volumes with SDD vpath devices using the Volume Groups SMIT panel.

    Note:
    To use this feature, you must either have root user authority or be a member of the system group.

    Attention: SDD does not automatically create the pvid attribute in the ODM database for each vpath device. The AIX disk driver automatically creates the pvid attribute in the ODM database, if a pvid exists on the physical device. Therefore, the first time you import a new SDD volume group to a new cluster node, you must import the volume group using hdisks as physical volumes. Next, run the hd2vp conversion script (see SDD utility programs) to convert the volume group's physical volumes from ESS hdisks to vpath devices. This conversion step not only creates pvid attributes for all vpath devices which belong to that imported volume group, it also deletes the pvid attributes for these vpath devices' underlying hdisks. Later on you can import and vary on the volume group directly from the vpath devices. These special requirements apply to both concurrent and nonconcurrent volume groups.

    See "Special requirements" for special requirements regarding unconfiguring and removing the ibmSdd_433.rte or ibmSdd_510nchacmp.rte installation packages for SDD 1.3.1.3 vpath devices.

    Perform the following steps to import a volume group with SDD devices:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Groups and press Enter. The Volume Groups panel is displayed.
    5. Select Import a Volume Group and press Enter. The Import a Volume Group panel is displayed.
    6. In the Import a Volume Group panel, perform the following tasks:

      You can press the F4 key for a list of choices.

    Exporting a volume group with SDD

    You can export a volume group definition from a set of physical volumes with SDD vpath devices using the Volume Groups SMIT panel.

    The exportvg command removes the definition of the volume group specified by the Volume Group parameter from the system. Since all system knowledge of the volume group and its contents are removed, an exported volume group is no longer accessible. The exportvg command does not modify any user data in the volume group.

    A volume group is a nonshared resource within the system; it should not be accessed by another system until it has been explicitly exported from its current system and imported on another. The primary use of the exportvg command, coupled with the importvg command, is to allow portable volumes to be exchanged between systems. Only a complete volume group can be exported, not individual physical volumes.

    Using the exportvg command and the importvg command, you can also switch ownership of data on physical volumes shared between two systems.

    Note:
    To use this feature, you must either have root user authority or be a member of the system group.

    Perform the following steps to export a volume group with SDD devices:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Groups and press Enter. The Volume Groups panel is displayed.
    5. Select Export a Volume Group and press Enter. The Export a Volume Group panel is displayed.
    6. Type in the volume group to export and press Enter.

    You can use the F4 key to select which volume group you want to export.

    Losing fallover protection

    AIX can only create volume groups from disk (or pseudo) devices that are physical volumes. If a volume group is created using a device that is not a physical volume, AIX makes it a physical volume as part of the procedure of creating the volume group. A physical volume has a physical volume identifier (pvid) written on its sector 0 and also has a pvid attribute attached to the device attributes in the CuAt ODM. The lspv command lists all the physical volumes known to AIX. Here is a sample output from this command:

    +--------------------------------------------------------------------------------+
    |hdisk0         0001926922c706b2    rootvg                                       |
    |hdisk1         none                None                                         |
    |...                                                                             |
    |hdisk10        none                None                                         |
    |hdisk11        00000000e7f5c88a    None                                         |
    |...                                                                             |
    |hdisk48        none                None                                         |
    |hdisk49        00000000e7f5c88a    None                                         |
    |vpath0         00019269aa5bc858    None                                         |
    |vpath1         none                None                                         |
    |vpath2         none                None                                         |
    |vpath3         none                None                                         |
    |vpath4         none                None                                         |
    |vpath5         none                None                                         |
    |vpath6         none                None                                         |
    |vpath7         none                None                                         |
    |vpath8         none                None                                         |
    |vpath9         00019269aa5bbadd    vpathvg                                      |
    |vpath10        00019269aa5bc4dc    vpathvg                                      |
    |vpath11        00019269aa5bc670    vpathvg                                      |
    |vpath12        000192697f9fd2d3    vpathvg                                      |
    |vpath13        000192697f9fde04    vpathvg                                      |
    +--------------------------------------------------------------------------------+

    In some cases, access to data is not lost, but fallover protection might not be present. fallover protection can be lost in several ways:

    The following sections provide more information about the ways that fallover protection can be lost.

    Through the loss of a device path

    Due to hardware errors, SDD might remove one or more paths to a vpath pseudo device. A pseudo device loses fallover protection when it only has a single path. You can use the datapath query device command to show the status of paths to a pseudo device. You cannot use any Dead path for I/O operations.

    By creating a volume group from single-path vpath (pseudo) devices

    A volume group created using any single-path pseudo devices does not have fallover protection because there is no alternate path to the ESS LUN.

    As a side effect of running the disk change method

    It is possible to modify attributes for an hdisk device by running the chdev command. The chdev command invokes the hdisk configuration method to make the requested change. In addition, the hdisk configuration method sets the pvid attribute for an hdisk if it determines that the hdisk has a pvid written on sector 0 of the LUN. This causes the vpath pseudo device and one or more of its hdisks to have the same pvid attribute in the ODM. If the volume group containing the vpath pseudo device is activated, the LVM uses the first device it finds in the ODM with the desired pvid to activate the volume group.

    As an example, if you issue the lsvpcfg command, the following output is displayed:

    +--------------------------------------------------------------------------------+
    |vpath0 (Avail pv vpathvg) 018FA067 = hdisk1 (Avail )                            |
    |vpath1 (Avail ) 019FA067 = hdisk2 (Avail )                                      |
    |vpath2 (Avail ) 01AFA067 = hdisk3 (Avail )                                      |
    |vpath3 (Avail ) 01BFA067 = hdisk4 (Avail ) hdisk27 (Avail )                     |
    |vpath4 (Avail ) 01CFA067 = hdisk5 (Avail ) hdisk28 (Avail )                     |
    |vpath5 (Avail ) 01DFA067 = hdisk6 (Avail ) hdisk29 (Avail )                     |
    |vpath6 (Avail ) 01EFA067 = hdisk7 (Avail ) hdisk30 (Avail )                     |
    |vpath7 (Avail ) 01FFA067 = hdisk8 (Avail ) hdisk31 (Avail )                     |
    |vpath8 (Avail ) 020FA067 = hdisk9 (Avail ) hdisk32 (Avail )                     |
    |vpath9 (Avail pv vpathvg) 02BFA067 = hdisk20 (Avail ) hdisk44 (Avail )          |
    |vpath10 (Avail pv vpathvg) 02CFA067 = hdisk21 (Avail ) hdisk45 (Avail )         |
    |vpath11 (Avail pv vpathvg) 02DFA067 = hdisk22 (Avail ) hdisk46 (Avail )         |
    |vpath12 (Avail pv vpathvg) 02EFA067 = hdisk23 (Avail ) hdisk47 (Avail )         |
    |vpath13 (Avail pv vpathvg) 02FFA067 = hdisk24 (Avail ) hdisk48 (Avail )         |
    +--------------------------------------------------------------------------------+

    The following example of a chdev command could also set the pvid attribute for an hdisk:

    chdev -l hdisk46 -a queue_depth=30
    

    For this example, the output of the lsvpcfg command would look similar to this:

    +--------------------------------------------------------------------------------+
    |vpath0 (Avail pv vpathvg) 018FA067 = hdisk1 (Avail )                            |
    |vpath1 (Avail ) 019FA067 = hdisk2 (Avail )                                      |
    |vpath2 (Avail ) 01AFA067 = hdisk3 (Avail )                                      |
    |vpath3 (Avail ) 01BFA067 = hdisk4 (Avail ) hdisk27 (Avail )                     |
    |vpath4 (Avail ) 01CFA067 = hdisk5 (Avail ) hdisk28 (Avail )                     |
    |vpath5 (Avail ) 01DFA067 = hdisk6 (Avail ) hdisk29 (Avail )                     |
    |vpath6 (Avail ) 01EFA067 = hdisk7 (Avail ) hdisk30 (Avail )                     |
    |vpath7 (Avail ) 01FFA067 = hdisk8 (Avail ) hdisk31 (Avail )                     |
    |vpath8 (Avail ) 020FA067 = hdisk9 (Avail ) hdisk32 (Avail )                     |
    |vpath9 (Avail pv vpathvg) 02BFA067 = hdisk20 (Avail ) hdisk44 (Avail )          |
    |vpath10 (Avail pv vpathvg) 02CFA067 = hdisk21 (Avail ) hdisk45 (Avail )         |
    |vpath11 (Avail pv vpathvg) 02DFA067 = hdisk22 (Avail ) hdisk46 (Avail pv vpathvg|
    |vpath12 (Avail pv vpathvg) 02EFA067 = hdisk23 (Avail ) hdisk47 (Avail )         |
    |vpath13 (Avail pv vpathvg) 02FFA067 = hdisk24 (Avail ) hdisk48 (Avail )         |
    +--------------------------------------------------------------------------------+

    The output of the lsvpcfg command shows that vpath11 contains hdisk22 and hdisk46. However, hdisk46 is the one with the pv attribute set. If you run the lsvg -p vpathvg command again, you might see something like this:

    +--------------------------------------------------------------------------------+
    |vpathvg:                                                                        |
    |PV_NAME           PV STATE    TOTAL PPs   FREE PPs    FREE DISTRIBUTION         |
    |vpath10           active      29          4           00..00..00..00..04        |
    |hdisk46           active      29          4           00..00..00..00..04        |
    |vpath12           active      29          4           00..00..00..00..04        |
    |vpath13           active      29          28          06..05..05..06..06        |
    +--------------------------------------------------------------------------------+

    Notice that now device vpath11 has been replaced by hdisk46. That is because hdisk46 is one of the hdisk devices included in vpath11 and it has a pvid attribute in the ODM. In this example, the LVM used hdisk46 instead of vpath11 when it activated volume group vpathvg. The volume group is now in a mixed mode of operation because it partially uses vpath pseudo devices and partially uses hdisk devices. This is a problem that must be fixed because fallover protection is effectively disabled for the vpath11 physical volume of the vpathvg volume group.

    Note:
    The way to fix this problem with the mixed volume group is to run the dpovgfix vg-name command after running the chdev command.

    By manually deleting devices and running the configuration manager

    (cfgmgr)

    Assume that vpath3 is made up of hdisk4 and hdisk27 and that vpath3 is currently a physical volume. If the vpath3, hdisk4, and hdisk27 devices are all deleted by using the rmdev command and then cfgmgr is invoked at the command line, only one path of the original vpath3 is configured by AIX. The following commands would produce this situation:

    rmdev -dl vpath3 rmdev -dl hdisk4 rmdev -dl hdisk27
    cfgmgr
    

    The datapath query device command displays the vpath3 configuration status.

    Next, all paths to the vpath must be restored. You can restore the paths in one of the following ways:

    Recovering from mixed volume groups

    Run the dpovgfix shell script to recover a mixed volume group. The syntax is dpovgfix vg-name. The script tries to find a pseudo device corresponding to each hdisk in the volume group and replaces the hdisk with the vpath pseudo device. In order for the shell script to be executed, all mounted file systems of this volume group have to be unmounted. After successful completion of the dpovgfix shell script, mount the file systems again.

    Extending an existing Subsystem Device Driver volume group

    You can extend a volume group with SDD vpath devices using the Logical Volume Groups SMIT panel. The SDD vpath devices to be added to the volume group should be chosen from those that can provide fallover protection. It is possible to add an SDD vpath device to an SDD volume group that has only a single path (vpath0 in Figure 2) and then add paths later by reconfiguring the ESS. With a single path, fallover protection is not provided. (See Adding paths to SDD devices of a volume group for information about adding paths to an SDD device.)

    Perform the following steps to extend a volume group with SDD devices:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Group and press Enter. The Add Volume Group with Data Path Devices panel is displayed.
    5. Select Add a Data Path Volume to a Volume Group and press Enter.
    6. Type in the volume group name and physical volume name and press Enter. You can also use the F4 key to list all the available SDD devices, and you can select the devices you want to add to the volume group.

    If you use a script file to extend an existing SDD volume group, you must modify your script file and replace the extendvg command with the extendvg4vp command.

    Backing-up all files belonging to a Subsystem Device Driver volume group

    You can back up all files belonging to a specified volume group with Subsystem Device Driver vpath devices using the Volume Groups SMIT panel.

    To backup a volume group with SDD devices, go to Accessing the Back Up a Volume Group with Data Path Devices SMIT panel.

    If you use a script file to back up all files belonging to a specified SDD volume group, you must modify your script file and replace the savevg command with the savevg4vp command.

    Attention: Backing-up files (running the savevg4vp command) will result in the loss of all material previously stored on the selected output medium. Data integrity of the archive may be compromised if a file is modified during system backup. Keep system activity at a minimum during the system backup procedure.

    Restoring all files belonging to a Subsystem Device Driver volume group

    You can restore all files belonging to a specified volume group with Subsystem Device Driver vpath devices using the Volume Groups SMIT panel.

    To restore a volume group with SDD devices and go to Accessing the Remake a Volume Group with Data Path Devices SMIT panel.

    If you use a script file to restore all files belonging to a specified SDD volume group, you must modify your script file and replace the restvg command with the restvg4vp command.

    SDD-specific SMIT panels

    SDD supports several special SMIT panels. Some SMIT panels provide SDD-specific functions, while other SMIT panels provide AIX functions (but requires SDD-specific commands). For example, the Add a Volume Group with Data Path Devices function uses the SDD mkvg4vp command, instead of the AIX mkvg command. Table 17 lists the SDD-specific SMIT panels and how you can use them.


    Table 17. SDD-specific SMIT panels and how to proceed
    SMIT panels How to proceed:


    Display Data Path Device Configuration

    Go to:

    Accessing the Display Data Path Device Configuration SMIT panel

    Display Data Path Device Adapter Status Accessing the Display Data Path Device Adapter Status SMIT panel
    Display Data Path Device Adapter Status Accessing the Display Data Path Device Adapter Status SMIT panel
    Define and Configure all Data Path Devices Accessing the Define and Configure All Data Path Devices SMIT panel
    Add Paths to Available Data Path Devices
    Note:
    SDD does not support the addpaths command for AIX 4.2.1; it supports this command for AIX 4.3.2 or higher.
    Accessing the Add Paths to Available Data Path Devices SMIT panel
    Configure a Defined Data Path Device Accessing the Configure a Defined Data Path Device SMIT panel
    Remove a Data Path Device Accessing the Remove a Data Path Device SMIT panel
    Add a Volume Group with Data Path Devices Accessing the Add a Volume Group with Data Path Devices SMIT panel
    Add a Data Path Volume to a Volume Group Accessing the Add a Data Path Volume to a Volume Group SMIT panel
    Remove a Physical Volume from a Volume group Accessing the Remove a Physical Volume from a Volume group SMIT panel
    Back Up a Volume Group with Data Path Devices Accessing the Back Up a Volume Group with Data Path Devices SMIT panel
    Remake a Volume Group with Data Path Devices Accessing the Remake a Volume Group with Data Path Devices SMIT panel

    Accessing the Display Data Path Device Configuration SMIT panel

    Perform the following steps to access the Display Data Path Device Configuration panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Display Data Path Device Configuration and press Enter.

    Accessing the Display Data Path Device Adapter Status SMIT panel

    Perform the following steps to access the Display Data Path Device Adapter Status panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Display Data Path Device Adapter Status and press Enter.

    Accessing the Display Data Path Device Adapter Status SMIT panel

    Perform the following steps to access the Display Data Path Device Adapter Status panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Display Data Path Device Adapter Status and press Enter.

    Accessing the Define and Configure All Data Path Devices SMIT panel

    To access the Define and Configure All Data Path Devices panel, perform the following steps:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Define and Configure All Data Path Devices and press Enter.

    Accessing the Add Paths to Available Data Path Devices SMIT panel

    Perform the following steps to access the Add Paths to Available Data Path Devices panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Add Paths to Available Data Path Devices and press Enter.
    Note:
    This SMIT panel is not available if you have the ibmSdd_421.rte installation package installed. SDD does not support the addpaths command for AIX 4.2.1; it supports this command for AIX 4.3.2 or higher.

    Accessing the Configure a Defined Data Path Device SMIT panel

    Perform the following steps to access the Configure a Defined Data Path Device panel:

    1. Type SMITTY from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Configure a Defined Data Path Device and press Enter.

    Accessing the Remove a Data Path Device SMIT panel

    Perform the following steps to access the Remove a Data Path Device panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Remove a Data Path Device and press Enter.

    Accessing the Add a Volume Group with Data Path Devices SMIT panel

    Perform the following steps to access the Add a volume group with data path devices panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Groups and press Enter. The Add Volume Group with Data Path Devices panel is displayed.
    5. Select Add Volume Group with Data Path Devices and press Enter.
      Note:
      Press F4 while highlighting the PHYSICAL VOLUME names field to list all the available SDD vpaths.

    Accessing the Add a Data Path Volume to a Volume Group SMIT panel

    Perform the following steps to access the Add a Data Path Volume to a Volume Group panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical) and press Enter. The System Storage Management (Physical & Logical) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
    4. Select Volume Group and press Enter. The Volume Group panel is displayed.
    5. Select Add a Data Path Volume to a Volume Group and press Enter.
    6. Type the volume group name and physical volume name and press Enter. Alternately, you can use the F4 key to list all the available SDD vpath devices and use the F7 key to select the physical volumes you want to add.

    Accessing the Remove a Physical Volume from a Volume group SMIT panel

    Perform the following steps to access the Remove a Physical Volume from a Volume group panel:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Logical Volume manager and press Enter. The Logical Volume manager panel is displayed.
    3. Select Volume Groups and press Enter. The Volume Groups panel is displayed.
    4. Select Set Characteristics of a Volume Group and press Enter. The Set Characteristics of a Volume Group panel is displayed.
    5. Select Remove a Physical Volume from a Volume group and press Enter. The Remove a Physical Volume from a Volume Group panel is displayed.

    Accessing the Back Up a Volume Group with Data Path Devices SMIT panel

    Perform the following steps to access the Back Up a Volume Group with Data Path Devices panel and to backup a volume group with SDD devices:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Groups and press Enter. The Volume Groups panel is displayed.
    5. Select Back Up a Volume Group with Data Path Devices and press Enter. The Back Up a Volume Group with Data Path Devices panel is displayed.
    6. In the Back Up a Volume Group with Data Path Devices panel, perform the following steps:

      Tip: You can also use the F4 key to list all the available SDD devices, and you can select the devices or files you want to backup. Attention: Backing-up files (running the savevg4vp command) will result in the loss of all material previously stored on the selected output medium. Data integrity of the archive may be compromised if a file is modified during system backup. Keep system activity at a minimum during the system backup procedure.

    Accessing the Remake a Volume Group with Data Path Devices SMIT panel

    Perform the following steps to access the Remake a Volume Group with Data Path Devices panel and restore a volume group with SDD devices:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    3. Select Logical Volume Manager and press Enter. The Volume Group panel is displayed.
    4. Select Volume Groups and press Enter. The Volume Groups panel is displayed.
    5. Select Remake a Volume Group with Data Path Devices and press Enter. The Remake a Volume Group with Data Path Devices panel is displayed.
    6. Type in the Restore DEVICE or FILE name, and press Enter. You can also use the F4 key to list all the available SDD devices, and you can select the devices or files you want to restore.

    SDD utility programs

    addpaths

    You can use the addpaths command to dynamically add more paths to SDD devices while they are in the Available status. In addition, this command allows you to add paths to vpath devices (which are then opened) belonging to active volume groups.

    This command will open a new path (or multiple paths) automatically if the vpath is in Open status, and the original number of path of the vpath is more than one. You can use either the Add Paths to Available Data Path Devices SMIT panel, or run the addpaths command from the AIX command line.

    SDD does not support the addpaths command for AIX 4.2.1. It is not available if you have the ibmSdd_421.rte installation package installed. SDD supports the addpaths command for AIX 4.3.2 or higher.

    For more information about this command, go to Adding paths to SDD devices of a volume group.

    hd2vp and vp2hd

    SDD provides two conversion scripts, hd2vp and vp2hd. The hd2vp script converts a volume group from ESS hdisks into SDD vpaths, and the vp2hd script converts a volume group from SDD vpaths into ESS hdisks. Use the vp2hd program when you want to configure your applications back to original ESS hdisks, or when you want to remove the SDD from your AIX host system.

    Note:
    You must convert all your applications and volume groups to the original ESS hdisk device special files before removing SDD.

    The syntax for these conversion scripts is as follows:

    hd2vp  vgname
    vp2hd  vgname
    

    These two conversion programs require that a volume group contain either all original ESS hdisks or all SDD vpaths. The program fails if a volume group contains both kinds of device special files (mixed volume group).

    Tip: Always use SMIT to create a volume group of SDD devices. This avoids the problem of a mixed volume group.

    dpovgfix

    You can use the dpovgfix script tool to recover mixed volume groups.

    Performing AIX system management operations on adapters and ESS hdisk devices can cause original ESS hdisks to be contained within a SDD volume group. This is known as a mixed volume group. Mixed volume groups happen when a SDD volume group is not active (varied off), and certain AIX commands to the hdisk put the pvid attribute of hdisk back into the ODM database. The following is an example of a command that does this:

       chdev -1 hdiskN -a queue_depth=30 
    

    If this disk is an active hdisk of a vpath that belongs to an SDD volume group, and you run the varyonvg command to activate this SDD volume group, LVM might pick up the hdisk device rather than the vpath device. The result is that an SDD volume group partially uses SDD vpath devices, and partially uses ESS hdisk devices. The result is the volume group loses path fallover capability for that physical volume. The dpovgfix script tool fixes this problem. The command syntax is:

    dpovgfix vg-name
    

    lsvpcfg

    You can use the lsvpcfg script tool to display the configuration status of SDD devices. This displays the configuration status for all SDD devices. The lsvpcfg command can be issued in two ways.

    1. The command can be issued without parameters. The command syntax is:
      lsvpcfg
      

      See Verifying the SDD configuration for an example of the output and what it means.

    2. The command can also be issued using the vpath device name as a parameter. The command syntax is:
      lsvpcfg vpathN0 vpathN1 vpathN2
      

      You will see output similar to this:

      +--------------------------------------------------------------------------------+
      |vpath10 (Avail pv ) 13916392 = hdisk95 (Avail ) hdisk179 (Avail )               |
      |vpath20 (Avail ) 02816392 = hdisk23 (Avail ) hdisk106 (Avail )                  |
      |vpath30 (Avail ) 10516392 = hdisk33 (Avail ) hdisk116 (Avail )                  |
      +--------------------------------------------------------------------------------+

      See Verifying the SDD configuration for an explanation of the output.

    mkvg4vp

    You can use the mkvg4vp command to create an SDD volume group. For more information about this command, go to Configuring a volume group for fallover protection.

    extendvg4vp

    You can use the extendvg4vp command to extend an existing SDD volume group. For more information about this command, go to Extending an existing Subsystem Device Driver volume group.

    Using ESS devices directly

    After you configure the SDD, it creates SDD devices (vpath devices) for ESS LUNs. ESS LUNs are accessible through the connection between the AIX host server SCSI or FCP adapter and the ESS ports. The AIX disk driver creates the original or ESS devices (hdisks). Therefore, with SDD, an application now has two ways in which to access ESS devices.

    To use the SDD load-balancing and fallover features and access ESS devices, your application must use the SDD vpath devices rather than the ESS hdisk devices.

    Two types of applications use ESS disk storage. One type of application accesses ESS devices through the SDD vpath device (raw device). The other type of application accesses ESS devices through the AIX logical volume manager (LVM). For this type of application, you must create a volume group with the SDD vpath devices.

    If your application used ESS hdisk device special files directly before installing SDD, convert it to using the SDD vpath device special files. After installing SDD, perform the following steps:

    1. Type smitty from your desktop window. The System Management Interface Tool is displayed.

      Tip:

    2. Select Devices and press Enter. The Devices panel is displayed.
    3. Select Data Path Devices and press Enter. The Data Path Devices panel is displayed.
    4. Select Display Data Path Device Configuration. The system displays all SDD vpaths with their attached multiple paths (hdisks).
    5. Search the list of hdisks to locate the hdisks your application is using.
    6. Replace each hdisk with its corresponding SDD vpath device.
      Note:
      Depending upon your application, the manner in which you replace these files is different. If this is a new application, use the SDD vpath rather than hdisk to use the SDD load-balancing and fallover features.
    Note:
    Alternately, you can type lsvpcfg from the command-line interface rather than using SMIT. This displays all configured SDD vpath devices and their underlying paths (hdisks).

    Using ESS devices through AIX LVM

    Attention:

    If your application accesses ESS devices through LVM, determine the volume group that it uses before you convert volume groups. Then, perform the following steps to convert the volume group from the original ESS device hdisks to the SDD vpaths:

    1. Determine the file systems or logical volumes that your application accesses.
    2. Type smitty from your desktop window. The System Management Interface Tool is displayed.
    3. Select System Storage Management (Physical & Logical Storage) and press Enter. The System Storage Management (Physical & Logical Storage) panel is displayed.
    4. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
    5. Select Logical Volume and press Enter. The Logical Volume panel is displayed.
    6. Select List All Logical Volumes by Volume Group to determine the logical volumes that belong to this volume group and their logical volume mount points.
    7. Press Enter. The logical volumes are listed by volume group.

      To determine the file systems, perform the following steps:

      1. Type smitty from your desktop window. The System Management Interface Tool is displayed.
      2. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
      3. Select File Systems and press Enter. The File Systems panel is displayed.
      4. Select List All File Systems to locate all file systems that have the same mount points as the logical volumes.
      5. Press Enter. The file systems are listed.
      6. Note the file system name of that volume group and the file system mount point, if it is mounted.
      7. Unmount these file systems.
    8. Enter the following to convert the volume group from the original ESS hdisks to SDD vpaths:
        hd2vp vgname
      
    9. When the conversion is complete, mount all file systems that you previously unmounted.

    When the conversion is complete, your application now accesses ESS physical LUNs through SDD vpath devices. This provides load balancing and fallover protection for your application.

    Migrating a non-SDD volume group to an ESS SDD multipath volume group in concurrent mode

    Before you migrate your non-SDD volume group to an SDD volume group, make sure that you have completed the following tasks:

    You should complete the following steps to migrate a non-SDD volume group to a multipath SDD volume group in concurrent mode:

    1. Add new SDD vpath devices to an existing non-SDD volume group:
      1. Type smitty and press Enter from your desktop window. The System Management Interface Tool panel is displayed.

        Tip:

        • The SMIT facility runs in two interfaces, nongraphical and graphical. This procedure uses the nongraphical interface. You can type smit to invoke the graphical user interface.
        • The list items on the SMIT panel might be worded differently from one AIX version to another.
      2. Select System Storage Management (Physical & Logical) and press Enter. The System Storage Management (Physical & Logical) panel is displayed.
      3. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
      4. Select Volume Group and press Enter. The Volume Group panel is displayed.
      5. Select Add a Data Path Volume to a Volume Group and press Enter.
      6. Type the volume group name and physical volume name and press Enter. Alternately, you can use the F4 key to list all the available SDD vpath devices and use the F7 key to select the physical volumes you want to add.
    2. Mirror logical volumes from the original volume to a Subsystem Device Driver ESS volume. Issue the command:
      smitty mklvcopy  
      

      Use the new Subsystem Device Driver vpath devices for copying all logical volumes. Do not forget to include JFS log volumes.

      Note:
      The command smitty mklvcopy copies one logical volume at a time. A fast-path command to mirror all the logical volumes on a volume group is mirrorvg.
    3. Synchronize logical volumes (LVs) or force synchronization. Issue the following command to synchronize all the volumes:
      smitty syncvg 
      

      There are two options on the smitty panel:

      The fast way to synchronize logical volumes is to select the Synchronize by Physical Volume option.

    4. Remove the mirror and delete the original LVs. Type the following command to remove the original copy of the logical volumes from all original non-Subsystem Device Driver physical volumes:
      smitty rmlvcopy 
      
    5. Remove the original non-Subsystem Device Driver devices from the volume group. Type the following command:
      smitty reducevg
      

      The Remove a Physical Volume panel is displayed. Remove all non-SDD devices.

    Note:
    A non-SDD volume group can consist of non-ESS or ESS hdisk devices. There is no fallover protection unless multiple paths are configured for each LUN.

    Migrating an existing non-SDD volume group to SDD vpath devices in concurrent mode

    This procedure shows how to migrate an existing AIX volume group to use SDD vpath (pseudo) devices that have multipath capability. You do not take the volume group out of service. The example shown starts with a volume group, vg1, made up of one ESS device, hdisk13.

    To perform the migration, you must have vpath devices available that are greater than or equal to the size of each of the hdisks making up the volume group. In this example, we have a pseudo device, vpath12, with two paths, hdisk14 and hdisk30, that we will migrate the volume group to.

    1. Add the vpath device to the volume group as an Available volume:
      1. Type smitty and press Enter from your desktop window. The System Management Interface Tool panel is displayed.

        Tip:

        • The SMIT facility runs in two interfaces, nongraphical and graphical. This procedure uses the nongraphical interface. You can type smit to invoke the graphical user interface.
        • The list items on the SMIT panel might be worded differently from one AIX version to another.
      2. Select System Storage Management (Physical & Logical) and press Enter. The System Storage Management (Physical & Logical) panel is displayed.
      3. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
      4. Select Volume Group and press Enter. The Volume Group panel is displayed.
      5. Select Add a Data Path Volume to a Volume Group and press Enter.
      6. Type vg1 in the Volume Group Name field. Type vpath12 in the Physical Volume Name field. Press Enter.

        You can also type the command:

        extendvg4vp -f vg1 vpath12
        
    2. Mirror logical volumes from the original volume to the new SDD vpath volume:
      1. Type smitty and press Enter from your desktop window. The System Management Interface Tool panel is displayed.
      2. Select System Storage Management (Physical & Logical) and press Enter. The System Storage Management (Physical & Logical) panel is displayed.
      3. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
      4. Select Volume Group and press Enter. The Volume Group panel is displayed.
      5. Select Mirror a Volume Group and press Enter. The Mirror a Volume Group panel is displayed.
      6. Type a volume group name. Type a physical volume name. Press Enter.

        You can also type the command:

        mirrorvg vg1 vpath12
        
    3. Synchronize the logical volumes in the volume group:
      1. Type smitty and press Enter from your desktop window. The System Management Interface Tool panel is displayed.
      2. Select System Storage Management (Physical & Logical) and press Enter. The System Storage Management (Physical & Logical) panel is displayed.
      3. Select Logical Volume Manager and press Enter. The Logical Volume Manager panel is displayed.
      4. Select Volume Group and press Enter. The Volume Group panel is displayed.
      5. Select Synchronize LVM Mirrors and press Enter. The Synchronize LVM Mirrors panel is displayed.
      6. Select Synchronize by Physical Volume.
      You can also type the command:
      syncvg -p hdisk13 vpath12
      
    4. Delete copies of all logical volumes from the original physical volume:
      1. Type smitty and press Enter from your desktop window. The System Management Interface Tool panel is displayed.
      2. Select Logical Volumes and press Enter. The Logical Volumes panel is displayed.
      3. Select Set Characteristic of a Logical Volume and press Enter. The Set Characteristic of a Logical Volume panel is displayed.
      4. Select Remove Copy from a Logical Volume and press Enter. The Remove Copy from a Logical Volume panel is displayed.
      You can also type the command:
      rmlvcopy loglv01
      1 hdisk13 and rmlvcopy lv01 1 hdisk13
      
    5. Remove the old physical volume from the volume group:
      1. Type smitty and press Enter from your desktop window. The System Management Interface Tool panel is displayed.
      2. Select Logical Volume manager and press Enter. The Logical Volume manager panel is displayed.
      3. Select Volume Groups and press Enter. The Volume Groups panel is displayed.
      4. Select Set Characteristics of a Volume Group and press Enter. The Set Characteristics of a Volume Group panel is displayed.
      5. Select Remove a Physical Volume from a Volume Group and press Enter. The Remove a Physical Volume from a Volume Group panel is displayed.
      You can also type the command:
      reducevg vg1 hdisk13
      

    Using the trace function

    SDD supports AIX trace functions. The SDD trace ID is 2F8. Trace ID 2F8 traces routine entry, exit, and error paths of the algorithm. To use it, manually turn on the trace function before the program starts to run, then turn off the trace function either after the program stops, or any time you need to read the trace report. To start the trace function, type:

       trace -a -j 2F8
    

    To stop the trace function, type:

       trcstop
    

    To read the report, type:

       trcrpt | pg
    

    To save the trace data to a file, type:

       trcrpt > filename
    
    
    Note:
    To perform the AIX trace function, you must have the bos.sysmgt.trace installation package installed on your system.

    Error log messages

    SDD logs error status into the AIX errlog system. To check if SDD has generated an error log message, type the following command:

       errpt -a | grep VPATH
    

    The following list shows the SDD error log messages and explains each one:

    VPATH_XBUF_NOMEM
    An attempt was made to open an SDD vpath file and to allocate kernel-pinned memory. The system returned a null pointer to the calling program and kernel-pinned memory was not available. The attempt to open the file failed.

    VPATH_PATH_OPEN
    SDD device file failed to open one of its paths (hdisks). An attempt to open a vpath device is successful if at least one attached path opens. The attempt to open a vpath device fails only when all the vpath device paths fail to open.

    VPATH_DEVICE_OFFLINE
    Several attempts to retry an I/O request for a vpath device on a path have failed. The path status is set to Dead and the path is taken Offline. Issue the datapath command to set the Offline path to Online. For more information, see Chapter 7, Using the datapath commands.

    VPATH_DEVICE_ONLINE
    SDD supports Dead path auto_failback and Dead path reclamation. A Dead path is selected to send an I/O, after it has been bypassed by 2000 I/O requests on an operational path. If the I/O is successful, the Dead path is put Online, and its status is changed back to Open; a Dead path is put Online, and its status changes to Open after it has been bypassed by 50 000 I/O requests on an operational path.

    New and modified error log messages by SDD for HACMP

    The following list shows the new and modified error log messages generated by SDD installed from the ibmSdd_433.rte or ibmSdd_510nchacmp.rte installation package. This SDD release is for HACMP environments only. See SDD installation package attributes for more information on this release.

    VPATH_DEVICE_OPEN
    The SDD device file failed to open one of its paths (hdisks). An attempt to open a vpath device is successful if at least one attached path opens. The attempt to open a vpath device fails only when all the vpath device paths fail to open. In addition, this error log message is posted when the vpath device fails to register its underlying paths or fails to read the persistent reserve key for the device.

    VPATH_OUT_SERVICE
    There is no path available to retry an I/O request that failed for a vpath device. The I/O request is returned to the calling program and this error log is posted.

    VPATH_FAIL_RELPRESERVE
    An attempt was made to close a vpath device that was not opened with the RETAIN_RESERVE option on the persistent reserve. The attempt to close the vpath device was successful; however, the persistent reserve was not released. The user is notified that the persistent reserve is still in effect, and this error log is posted.

    VPATH_RESV_CFLICT
    An attempt was made to open a vpath device, but the reservation key of the vpath device is different from the reservation key currently in effect. The attempt to open the device fails and this error log is posted. The device could not be opened because it is currently reserved by someone else.

    Chapter 3. SDD for a Windows NT host system

    This chapter provides procedures for you to install, configure, remove, and use the SDD on a Windows NT host system that is attached to an ESS. For updated and additional information not included in this chapter, see the README file on the compact disc or visit the SDD Web site at:www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates


    Verifying the hardware and software requirements

    Hardware

    You must have the following hardware and software components in order to successfully install SDD.

    Software

    Non-support environments

    SDD does not support the following environments:

    ESS requirements

    To successfully install SDD, ensure that your host system is configured to the ESS as an Intel-based PC (personal computer) server with Windows NT 4.0 or higher.

    Host system requirements

    To successfully install SDD, your Windows NT host system must be an Intel-based system with Windows NT Version 4.0 Service Pack 3 or higher installed. The host system can be a uniprocessor or a multiprocessor system.

    SCSI requirements

    To use the SDD SCSI support, ensure your host system meets the following requirements:

    Fibre requirements

    To use the SDD fibre support, ensure that your host system meets the following requirements:

    For information about the fibre-channel adapters that can attach to your Windows NT host system, go to the following Web site:
    www.storage.ibm.com/hardsoft/products/ess/supserver.htm


    Preparing for SDD installation

    Before you install SDD, you must configure the ESS to your host system and required fibre-channel adapters are attached.

    Configuring the ESS

    Before you install SDD, configure your ESS for single-port or multiport access for each LUN. SDD requires a minimum of two independent paths that share the same LUN to use the load-balancing and failover features.

    For information about configuring your ESS, see IBM Enterprise Storage Server Introduction and Planning Guide.

    Configuring fibre-channel adapters

    You must configure the fibre-channel adapters that are attached to your Windows NT host system before you install SDD. Follow the adapter-specific configuration instructions to configure the adapters attached to your Windows NT host systems.

    SDD supports Emulex LP8000 with the full port driver. When you configure Emulex LP8000 for multipath functions, select Allow Multiple paths to SCSI Targets in the Emulex Configuration Tool panel.

    Make sure that your Windows NT host system has Service Pack 3 or higher. See IBM TotalStorage Enterprise Storage Server Host System Attachment Guide for more information about installing and configuring fibre-channel adapters for your Windows NT host systems.

    Configuring SCSI adapters

    Attention: Failure to disable the BIOS of attached nonstart devices may cause your system to attempt to start from an unexpected nonstart device.

    Before you install and use SDD, you must configure your SCSI adapters. For SCSI adapters that attach start devices, ensure that the BIOS for the adapter is enabled. For all other adapters that attach nonstart devices, ensure that the BIOS for the adapter is disabled.

    Note:
    When the adapter shares the SCSI bus with other adapters, the BIOS must be disabled.

    Installing the Subsystem Device Driver

    The following section describes how to install SDD. Make sure that all hardware and software requirements are met before you install the Subsystem Device Driver. See Verifying the hardware and software requirements for more information.

    Perform the following steps to install the SDD filter and application programs on your system:

    1. Log on as the administrator user.
    2. Insert the SDD installation compact disc into the CD-ROM drive.
    3. Start the Windows NT Explorer program.
    4. Select the CD-ROM drive. A list of all the installed directories on the compact disc is displayed.
    5. Select the \winNt\IBMsdd directory.
    6. Run the setup.exe program. The Installshield starts.
    7. Click Next. The Software License agreement is displayed.
    8. Click Yes. The User Information window opens.
    9. Type your name and your company name.
    10. Click Next. The Choose Destination Location window opens.
    11. Click Next. The Setup windows opens.
    12. Select the type of setup you prefer from the following setup choices. IBM recommends that you select Typical.
      Typical
      Selects all options.
      Compact
      You select the minimum required options only (the installation driver and the README file).
      Custom
      Select the options that you need.
    13. Click Next. The Setup Complete window opens.
    14. Click Finish. The SDD program prompts you to start your computer again.
    15. Click Yes to start your computer again. When you log on again, you see a Subsystem Device Driver Management entry in your Program menu containing the following files:
      1. SDD management
      2. Subsystem Device Driver manual
      3. README file
    Note:
    You can use the datapath query device command to verify the SDD installation. SDD is successfully installed if the command runs successfully.

    Configuring the Subsystem Device Driver

    To activate SDD, you need to restart your Windows NT system after it is installed. In fact, a restart is required to activate multipath support whenever a new file system or partition is added.

    Adding paths to SDD devices

    Attention: Ensure that SDD is installed before you add a new path to a device. Otherwise, the Windows NT server could lose the ability to access existing data on that device.

    This section contains the procedures for adding paths to SDD devices in multipath environments.

    Reviewing the existing SDD configuration information

    Before adding any additional hardware, review the configuration information for the adapters and devices currently on your Windows NT server.

    Verify that the number of adapters and the number of paths to each ESS volume match the known configuration. Perform the following steps to display information about the adapters and devices:

    1. You must log on as an administrator user to have access to the Windows NT disk administrator.
    2. Click Start --> Program --> Subsystem Device Driver --> Subsystem Device Driver Management. An MS-DOS window opens.
    3. Type datapath query adapter and press Enter. The output includes information about all the installed adapters. In the example shown in the following output, one SCSI adapter has 10 active paths:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |Active Adapters :1                                                              |
      |                                                                                |
      |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
      |    0  Scsi Port6 Bus0  NORMAL   ACTIVE        542          0     10      10    |
      |                                                                                |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    4. Type datapath query device and press Enter. In the example shown in the following output, 10 devices are attached to the SCSI path:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |Total Devices : 10                                                              |
      |                                                                                |
      |DEV#:   0  DEVICE NAME: Disk2 Part0  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part0     OPEN   NORMAL         14          0    |
      |                                                                                |
      |DEV#:   1  DEVICE NAME: Disk2 Part1  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part1     OPEN   NORMAL         94          0    |
      |                                                                                |
      |DEV#:   2  DEVICE NAME: Disk3 Part0  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part0     OPEN   NORMAL         16          0    |
      |                                                                                |
      |DEV#:   3  DEVICE NAME: Disk3 Part1  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part1     OPEN   NORMAL         94          0    |
      |                                                                                |
      |DEV#:   4  DEVICE NAME: Disk4 Part0  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part0     OPEN   NORMAL         14          0    |
      |                                                                                |
      |DEV#:   5  DEVICE NAME: Disk4 Part1  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part1     OPEN   NORMAL         94          0    |
      |                                                                                |
      |DEV#:   6  DEVICE NAME: Disk5 Part0  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part0     OPEN   NORMAL         14          0    |
      |                                                                                |
      |DEV#:   7  DEVICE NAME: Disk5 Part1  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part1     OPEN   NORMAL         94          0    |
      |                                                                                |
      |DEV#:   8  DEVICE NAME: Disk6 Part0  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part0     OPEN   NORMAL         14          0    |
      |                                                                                |
      |DEV#:   9  DEVICE NAME: Disk6 Part1  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part1     OPEN   NORMAL         94          0    |
      |                                                                                |
      +--------------------------------------------------------------------------------+

    Installing and configuring additional paths

    Perform the following steps to install and configure additional paths to a vpath device:

    1. Install any additional hardware on the Windows NT server.
    2. Install any additional hardware on the ESS.
    3. Configure the new paths to the server.
    4. Restart the Windows NT server. Restarting will ensure correct multipath access to both existing and new storage and to your Windows NT server.
    5. Verify that the path is added correctly. See Verifying additional paths are installed correctly.

    Verifying additional paths are installed correctly

    After installing additional paths to SDD devices, verify the following conditions:

    Perform the following steps to verify that the additional paths have been installed correctly:

    1. Click Start --> Program --> Subsystem Device Driver --> Subsystem Device Driver Management. An MS-DOS window opens.
    2. Type datapath query adapter and press Enter. The output includes information about any additional adapters that were installed. In the example shown in the following output, an additional path is installed to the previous configuration:
      +--------------------------------------------------------------------------------+
      |Active Adapters :2                                                              |
      |                                                                                |
      |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
      |    0  Scsi Port6 Bus0  NORMAL   ACTIVE        188          0     10      10    |
      |    1  Scsi Port7 Bus0  NORMAL   ACTIVE        204          0     10      10    |
      |                                                                                |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    3. Type datapath query device and press Enter. The output includes information about any additional devices that were installed. In the example shown in the following output, the output includes information about the new SCSI adapter that was assigned:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |                                                                                |
      |Total Devices : 10                                                              |
      |                                                                                |
      |DEV#:   0  DEVICE NAME: Disk2 Part0  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part0     OPEN   NORMAL          5          0    |
      |    1    Scsi Port7 Bus0/Disk7 Part0     OPEN   NORMAL          9          0    |
      |                                                                                |
      |DEV#:   1  DEVICE NAME: Disk2 Part1  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part1     OPEN   NORMAL         32          0    |
      |    1    Scsi Port7 Bus0/Disk7 Part1     OPEN   NORMAL         32          0    |
      |                                                                                |
      |DEV#:   2  DEVICE NAME: Disk3 Part0  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part0     OPEN   NORMAL          7          0    |
      |    1    Scsi Port7 Bus0/Disk8 Part0     OPEN   NORMAL          9          0    |
      |                                                                                |
      |DEV#:   3  DEVICE NAME: Disk3 Part1  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part1     OPEN   NORMAL         28          0    |
      |    1    Scsi Port7 Bus0/Disk8 Part1     OPEN   NORMAL         36          0    |
      |                                                                                |
      |DEV#:   4  DEVICE NAME: Disk4 Part0  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part0     OPEN   NORMAL          8          0    |
      |    1    Scsi Port7 Bus0/Disk9 Part0     OPEN   NORMAL          6          0    |
      |                                                                                |
      |DEV#:   5  DEVICE NAME: Disk4 Part1  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part1     OPEN   NORMAL         35          0    |
      |    1    Scsi Port7 Bus0/Disk9 Part1     OPEN   NORMAL         29          0    |
      |                                                                                |
      |DEV#:   6  DEVICE NAME: Disk5 Part0  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part0     OPEN   NORMAL          6          0    |
      |    1   Scsi Port7 Bus0/Disk10 Part0     OPEN   NORMAL          8          0    |
      |                                                                                |
      |DEV#:   7  DEVICE NAME: Disk5 Part1  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part1     OPEN   NORMAL         24          0    |
      |    1   Scsi Port7 Bus0/Disk10 Part1     OPEN   NORMAL         40          0    |
      |                                                                                |
      |DEV#:   8  DEVICE NAME: Disk6 Part0  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part0     OPEN   NORMAL          8          0    |
      |    1   Scsi Port7 Bus0/Disk11 Part0     OPEN   NORMAL          6          0    |
      |                                                                                |
      |DEV#:   9  DEVICE NAME: Disk6 Part1  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part1     OPEN   NORMAL         35          0    |
      |    1   Scsi Port7 Bus0/Disk11 Part1     OPEN   NORMAL         29          0    |
      |                                                                                |
      +--------------------------------------------------------------------------------+

      The definitive way to identify unique volumes on the ESS is by the serial number displayed. The volume appears at the SCSI level as multiple disks (more properly, Adapter/Bus/ID/LUN), but it's the same volume on the ESS. The previous example shows two paths to each partition (path 0: Scsi Port6 Bus0/Disk2, and path 1: Scsi Port7 Bus0/Disk7).

      The example shows partition 0 (Part0) for each of the devices. This partition stores information about the Windows partition on the drive. The operating system masks this partition from the user, but it still exists. In general, you will see one more partition from the output of the datapath query device command than what is being displayed from the Disk Administrator application.

    Upgrading the Subsystem Device Driver

    If you attempt to install over an existing version of SDD or Data Path Optimizer (DPO), the installation fails. You must uninstall any previous version of the SDD or DPO before installing a new version of SDD.

    Perform the following steps to upgrade to a newer SDD version:

    1. Uninstall the previous version of SDD. (See Removing the Subsystem Device Driver for instructions.)

      Attention: After uninstalling the previous version, you must immediately install the new version of SDD to avoid any potential data loss. If you perform a system restart before installing the new version, you may loose access to your assigned volumes.

    2. Install the new version of SDD. (See Installing the Subsystem Device Driver for instructions.)

    Adding or modifying a multipath storage configuration to the ESS

    This section contains the procedures for adding new storage to an existing configuration in multipath environments.

    Reviewing the existing SDD configuration information

    Before adding any additional hardware, review the configuration information for the adapters and devices currently on your Windows NT server.

    Verify that the number of adapters and the number of paths to each ESS volume match the known configuration. Perform the following steps to display information about the adapters and devices:

    1. Click Start --> Program --> Subsystem Device Driver --> Subsystem Device Driver Management. An MS-DOS window opens.
    2. Type datapath query adapter and press Enter. The output includes information about all the installed adapters. In the example shown in the following output, two SCSI adapters are installed on the Windows NT host server:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |Active Adapters :2                                                              |
      |                                                                                |
      |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
      |    0  Scsi Port6 Bus0  NORMAL   ACTIVE        188          0     10      10    |
      |    1  Scsi Port7 Bus0  NORMAL   ACTIVE        204          0     10      10    |
      |                                                                                |
      |Previous configuration with one additional path                                 |
      |                                                                                |
      |                                                                                |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    3. Type datapath query device and press Enter. In the example shown in the following output, 10 devices are attached to the SCSI path:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |Total Devices : 10                                                              |
      |                                                                                |
      |DEV#:   0  DEVICE NAME: Disk2 Part0  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part0     OPEN   NORMAL          5          0    |
      |    1    Scsi Port7 Bus0/Disk7 Part0     OPEN   NORMAL          9          0    |
      |                                                                                |
      |DEV#:   1  DEVICE NAME: Disk2 Part1  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part1     OPEN   NORMAL         32          0    |
      |    1    Scsi Port7 Bus0/Disk7 Part1     OPEN   NORMAL         32          0    |
      |                                                                                |
      |DEV#:   2  DEVICE NAME: Disk3 Part0  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part0     OPEN   NORMAL          7          0    |
      |    1    Scsi Port7 Bus0/Disk8 Part0     OPEN   NORMAL          9          0    |
      |                                                                                |
      |DEV#:   3  DEVICE NAME: Disk3 Part1  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part1     OPEN   NORMAL         28          0    |
      |    1    Scsi Port7 Bus0/Disk8 Part1     OPEN   NORMAL         36          0    |
      |                                                                                |
      |DEV#:   4  DEVICE NAME: Disk4 Part0  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part0     OPEN   NORMAL          8          0    |
      |    1    Scsi Port7 Bus0/Disk9 Part0     OPEN   NORMAL          6          0    |
      |                                                                                |
      |DEV#:   5  DEVICE NAME: Disk4 Part1  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part1     OPEN   NORMAL         35          0    |
      |    1    Scsi Port7 Bus0/Disk9 Part1     OPEN   NORMAL         29          0    |
      |                                                                                |
      |DEV#:   6  DEVICE NAME: Disk5 Part0  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part0     OPEN   NORMAL          6          0    |
      |    1   Scsi Port7 Bus0/Disk10 Part0     OPEN   NORMAL          8          0    |
      |                                                                                |
      |DEV#:   7  DEVICE NAME: Disk5 Part1  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part1     OPEN   NORMAL         24          0    |
      |    1   Scsi Port7 Bus0/Disk10 Part1     OPEN   NORMAL         40          0    |
      |                                                                                |
      |DEV#:   8  DEVICE NAME: Disk6 Part0  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part0     OPEN   NORMAL          8          0    |
      |    1   Scsi Port7 Bus0/Disk11 Part0     OPEN   NORMAL          6          0    |
      |                                                                                |
      |DEV#:   9  DEVICE NAME: Disk6 Part1  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part1     OPEN   NORMAL         35          0    |
      |    1   Scsi Port7 Bus0/Disk11 Part1     OPEN   NORMAL         29          0    |
      |                                                                                |
      +--------------------------------------------------------------------------------+

    Adding new storage to existing configuration

    Perform the following steps to install additional storage:

    1. Install any additional hardware to the ESS.
    2. Configure the new storage to the server.
    3. Restart the Windows NT server. Restarting will ensure correct multipath access to both existing and new storage and to your Windows NT server.
    4. Verify that the new storage is added correctly. See Verifying new storage is installed correctly.

    Verifying new storage is installed correctly

    After adding new storage to existing configuration, you should verify the following conditions:

    Perform the following steps to verify that the additional storage has been installed correctly:

    1. Click Start --> Program --> Subsystem Device Driver --> Subsystem Device Driver Management. An MS-DOS window opens.
    2. Type datapath query adapter and press Enter. The output includes information about all the installed adapters. In the example shown in the following output, two SCSI adapters are installed on the Windows NT host server:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |Active Adapters :2                                                              |
      |                                                                                |
      |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
      |    0  Scsi Port6 Bus0  NORMAL   ACTIVE        295          0     16      16    |
      |    1  Scsi Port7 Bus0  NORMAL   ACTIVE        329          0     16      16    |
      |                                                                                |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    3. Type datapath query device and press Enter. The output includes information about any additional devices that were installed. In the example shown in the following output, the output includes information about the new devices that were assigned:
      +--------------------------------------------------------------------------------+
      |                                                                                |
      |Total Devices : 16                                                              |
      |                                                                                |
      |DEV#:   0  DEVICE NAME: Disk2 Part0  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part0     OPEN   NORMAL          9          0    |
      |    1   Scsi Port7 Bus0/Disk10 Part0     OPEN   NORMAL          5          0    |
      |                                                                                |
      |DEV#:   1  DEVICE NAME: Disk2 Part1  TYPE: 2105E20   SERIAL: 00A12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk2 Part1     OPEN   NORMAL         26          0    |
      |    1   Scsi Port7 Bus0/Disk10 Part1     OPEN   NORMAL         38          0    |
      |                                                                                |
      |DEV#:   2  DEVICE NAME: Disk3 Part0  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part0     OPEN   NORMAL          9          0    |
      |    1   Scsi Port7 Bus0/Disk11 Part0     OPEN   NORMAL          7          0    |
      |                                                                                |
      |DEV#:   3  DEVICE NAME: Disk3 Part1  TYPE: 2105E20   SERIAL: 00B12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk3 Part1     OPEN   NORMAL         34          0    |
      |    1   Scsi Port7 Bus0/Disk11 Part1     OPEN   NORMAL         30          0    |
      |                                                                                |
      |DEV#:   4  DEVICE NAME: Disk4 Part0  TYPE: 2105E20   SERIAL: 31512028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part0     OPEN   NORMAL          8          0    |
      |    1   Scsi Port7 Bus0/Disk12 Part0     OPEN   NORMAL          6          0    |
      |                                                                                |
      |DEV#:   5  DEVICE NAME: Disk4 Part1  TYPE: 2105E20   SERIAL: 31512028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk4 Part1     OPEN   NORMAL         35          0    |
      |    1   Scsi Port7 Bus0/Disk12 Part1     OPEN   NORMAL         28          0    |
      |                                                                                |
      |DEV#:   6  DEVICE NAME: Disk5 Part0  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part0     OPEN   NORMAL          5          0    |
      |    1   Scsi Port7 Bus0/Disk13 Part0     OPEN   NORMAL          9          0    |
      |                                                                                |
      |DEV#:   7  DEVICE NAME: Disk5 Part1  TYPE: 2105E20   SERIAL: 00D12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk5 Part1     OPEN   NORMAL         28          0    |
      |    1   Scsi Port7 Bus0/Disk13 Part1     OPEN   NORMAL         36          0    |
      |                                                                                |
      |DEV#:   8  DEVICE NAME: Disk6 Part0  TYPE: 2105E20   SERIAL: 40812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part0     OPEN   NORMAL          5          0    |
      |    1   Scsi Port7 Bus0/Disk14 Part0     OPEN   NORMAL          9          0    |
      |                                                                                |
      |DEV#:   9  DEVICE NAME: Disk6 Part1  TYPE: 2105E20   SERIAL: 40812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk6 Part1     OPEN   NORMAL         25          0    |
      |    1   Scsi Port7 Bus0/Disk14 Part1     OPEN   NORMAL         38          0    |
      |                                                                                |
      |DEV#:  10  DEVICE NAME: Disk7 Part0  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk7 Part0     OPEN   NORMAL          7          0    |
      |    1   Scsi Port7 Bus0/Disk15 Part0     OPEN   NORMAL          7          0    |
      |                                                                                |
      |DEV#:  11  DEVICE NAME: Disk7 Part1  TYPE: 2105E20   SERIAL: 50812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk7 Part1     OPEN   NORMAL         34          0    |
      |    1   Scsi Port7 Bus0/Disk15 Part1     OPEN   NORMAL         30          0    |
      |                                                                                |
      |DEV#:  12  DEVICE NAME: Disk8 Part0  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk8 Part0     OPEN   NORMAL          7          0    |
      |    1   Scsi Port7 Bus0/Disk16 Part0     OPEN   NORMAL          7          0    |
      |                                                                                |
      |DEV#:  13  DEVICE NAME: Disk8 Part1  TYPE: 2105E20   SERIAL: 60012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk8 Part1     OPEN   NORMAL         29          0    |
      |    1   Scsi Port7 Bus0/Disk16 Part1     OPEN   NORMAL         35          0    |
      |                                                                                |
      |DEV#:  14  DEVICE NAME: Disk9 Part0  TYPE: 2105E20   SERIAL: 00812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk9 Part0     OPEN   NORMAL          6          0    |
      |    1   Scsi Port7 Bus0/Disk17 Part0     OPEN   NORMAL          8          0    |
      |                                                                                |
      |DEV#:  15  DEVICE NAME: Disk9 Part1  TYPE: 2105E20   SERIAL: 00812028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port6 Bus0/Disk9 Part1     OPEN   NORMAL         28          0    |
      |    1   Scsi Port7 Bus0/Disk17 Part1     OPEN   NORMAL         36          0    |
      |                                                                                |
      |                                                                                |
      +--------------------------------------------------------------------------------+

      The definitive way to identify unique volumes on the ESS is by the serial number displayed. The volume appears at the SCSI level as multiple disks (more properly, Adapter/Bus/ID/LUN), but it's the same volume on the ESS. The previous example shows two paths to each partition (path 0: Scsi Port6 Bus0/Disk2, and path 1: Scsi Port7 Bus0/Disk10).

      The example shows partition 0 (Part0) for each of the device. This partition stores information about the Windows partition on the drive. The operating system masks this partition from the user, but it still exists. In general, you will see one more partition from the output of the datapath query device command than what is being displayed from the Disk Administrator application.


    Removing the Subsystem Device Driver

    Perform the following steps to uninstall SDD on a Windows NT host system:

    1. Log on as the administrator user.
    2. Click Start --> Settings --> Control Panel. The Control Panel window opens.
    3. Double click Add/Remove Programs. The Add/Remove Programs window opens.
    4. In the Add/Remove Programs window, select Subsystem Device Driver from the Currently installed programs selection list.
    5. Click on the Add/Remove button.

    Attention:


    Displaying the current version of the Subsystem Device Driver

    You can display the current SDD version on a Windows NT host system by viewing the sddpath.sys file properties. Perform the following steps to view the properties of the sddpath.sys file:

    1. Click Start --> Run --> Programs --> Accessories --> Windows Explorer. Windows will open Windows Explorer.
    2. In Windows Explorer, go to your_installation_directory_letter:\Winnt\system32\drivers directory where

      (your_installation_directory_letter\ is the directory letter where you have installed the sddpath.sys file).

    3. Click the sddpath.sys file in your_installation_directory_letter:\Winnt\system32\drivers directory

      where your_installation_directory_letter refers to the letter of the directory in which you have installed the sddpath.sys file.

    4. Right-click on the sddpath.sys file and then click Properties. The sddpath.sys properties window will open.
    5. In the sddpath.sys properties window, click Version. The file version and copyright information about the sddpath.sys file is displayed.

    Support for Windows NT clustering

    The following items are required to support the Windows NT operating system is a clustering environment:

    Note:
    SDD 1.2.1 or higher does not support I/O load-balancing in a Windows NT clustering environment.

    Special considerations in the Windows NT clustering environment

    There are subtle differences in the way that SDD handles path reclamation in a Windows NT clustering environment compared to a nonclustering environment. When the Windows NT server loses a path in a nonclustering environment, the path condition changes from open to dead and the adapter condition changes from active to degraded. The adapter and path condition will not change until the path is made operational again. When the Windows NT server loses a path in a clustering environment, the path condition changes from open to dead and the adapter condition changes from active to degraded. However, after a period of time, the path condition changes back to open and the adapter condition changes back to normal, even if the path has not been made operational again.

    The datapath set adapter # offline command operates differently in a clustering environment as compared to a nonclustering environment. In a clustering environment, the datapath set adapter offline command does not change the condition of the path if the path is active or being reserved. If you issue the command, the following message is displayed: to preserve access some paths left online.

    Configuring a Windows NT cluster with SDD

    The following variables are used in this procedure:

    server_1 represents the first server with two host bus adapters (HBAs).

    server_2 represents the second server with two HBAs.

    hba_a represents the first HBA for server_1.

    hba_b represents the second HBA for server_1.

    hba_c represents the first HBA for server_2.

    hba_d represents the second HBA for server_2.

    Perform the following steps to configure a Windows NT cluster with SDD:

    1. Configure LUNs on the ESS as shared for all HBAs on both server_1 and server_2.
    2. Connect hba_a to the ESS, and restart server_1.
    3. Click Start --> Programs --> Administrative Tools --> Disk Administrator. The Disk Administrator is displayed. Use the Disk Administrator to verify the number of LUNs that are connected to server_1.

      The operating system recognizes each additional path to the same LUN as a device.

    4. Disconnect hba_a and connect hba_b to the ESS. Restart server_1.
    5. Click Start --> Programs --> Administrative Tools --> Disk Administrator. The Disk Administrator is displayed. Use the Disk Administrator to verify the number of LUNs that are connected to server_1.

      If the number of LUNs that are connected to server_1 is correct, proceed to 6.

      If the number of LUNs that are connected to server_1 is incorrect, perform the following steps:

      1. Verify that the cable for hba_b is connected to the ESS.
      2. Verify your LUN configuration on the ESS.
      3. Repeat steps 2 - 5.
    6. Install SDD on server_1, and restart server_1.

      For installation instructions , go to Installing the Subsystem Device Driver section.

    7. Connect hba_c to the ESS, and restart server_2.
    8. Click Start --> Programs --> Administrative Tools --> Disk Administrator. The Disk Administrator is displayed. Use the Disk Administrator to verify the number of LUNs that are connected to server_2.

      The operating system recognizes each additional path to the same LUN as a device.

    9. Disconnect hba_c and connect hba_d to the ESS. Restart server_2.
    10. Click Start --> Programs --> Administrative Tools --> Disk Administrator. The Disk Administrator is displayed. Use the Disk Administrator to verify that the correct number of LUNs are connected to server_2.

      If the number of LUNs that are connected to server_2 is correct, proceed to 11.

      If the number of LUNs that are connected to server_2 is incorrect, perform the following steps:

      1. Verify that the cable for hba_d is connected to the ESS.
      2. Verify your LUN configuration on the ESS.
      3. Repeat steps 7 - 10.
    11. Install SDD on server_2, and restart server_2.

      For installation instructions , go to Installing the Subsystem Device Driver section.

    12. Connect both hba_c and hba_d on server_2 to the ESS, and restart server_2.
    13. Use the datapath query adapter and datapath query device commands to verify the number of LUNs and paths on server_2.
    14. Click Start --> Programs --> Administrative Tools --> Disk Administrator. The Disk Administrator is displayed. Use the Disk Administrator to verify the number of LUNs as online devices. You also need to verify that all additional paths are shown as offline devices.
    15. Format the raw devices with NTFS.

      Make sure to keep track of the assigned drive letters on server_2.

    16. Connect both hba_a and hba_b on server_1 to the ESS, and restart server_1.
    17. Use the datapath query adapter and datapath query device commands to verify the correct number of LUNs and paths on server_1.

      Verify that the assigned drive letters on server_1 match the assigned drive letters on server_2.

    18. Restart server_2.
    19. Use the datapath query adapter and datapath query device commands to verify the correct number of LUNs and paths on server_1 and server_2. (This step is optional.)
    Note:
    You can use the datapath query adapter and datapath query device commands to show all the physical volumes and logical volumes for the host server. The secondary server only shows the physical volumes and the logical volumes that it owns.

    Chapter 4. SDD for a Windows 2000 host system

    This chapter provides procedures for you to install, configure, remove, and use the SDD on a Windows 2000 host system that is attached to an ESS. For updated and additional information not included in this chapter, see the README file on the compact disc or visit the SDD Web site at:
    www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates


    Verifying the hardware and software requirements

    You must have the following hardware and software components in order to install SDD:

    Hardware

    Software

    Non-supported environments

    SDD does not support the following environments:

    ESS requirements

    To successfully install SDD, make sure that you configure the ESS devices as IBM 2105xxx (where xxx is the ESS model number) on your Windows 2000 host system.

    Host system requirements

    To successfully install SDD, your host system should have Windows 2000 Service Pack 2 installed. The host system can be a uniprocessor or a multiprocessor system.

    To install all components, you must have 1 MB (MB equals approximately 1 000 000 bytes) of disk space available.

    SCSI requirements

    To use the SDD SCSI support, ensure your host system meets the following requirements:

    Fibre requirements

    To use the SDD fibre support, ensure that your host system meets the following requirements:


    Preparing for SDD 1.3.2.1 installation

    Before installing SDD 1.3.2.1, you must:

    1. Ensure that all hardware and software requirements are met before you install the Subsystem Device Driver. See Verifying the hardware and software requirements for more information.
    2. Configure the ESS to your host system. See Configuring the ESS for more information.
    3. Configure the fibre-channel adapters that are attached to your Windows 2000 host system. See Configuring fibre-channel adapters for more information.
    4. Configure the SCSI adapters that are attached to your Windows 2000 host system. See Configuring SCSI adapters for more information.
    5. Uninstall any previously installed version of SDD on your host system.

      For SDD uninstallation and installation instructions, go to:

      1. Uninstalling the Subsystem Device Driver
      2. Installing SDD 1.3.2.1

    Configuring the ESS

    Before you install SDD, configure your ESS for single-port or multiport access for each LUN. SDD requires a minimum of two independent paths that share the same logical unit to use the load-balancing and failover features.

    For information about configuring your ESS, see the IBM Enterprise Storage Server Introduction and Planning Guide.

    Note:
    During heavy usage, the Windows 2000 operating system might slow down while trying to recover from error conditions.

    Configuring fibre-channel adapters

    You must configure the fibre-channel adapters that are attached to your Windows 2000 host system before you install SDD. Follow the adapter-specific configuration instructions to configure the adapters attached to your Windows 2000 host systems.

    SDD supports Emulex LP8000 with the full port driver. When you configure Emulex LP8000 for multipath functions, select Allow Multiple Paths to SCSI Targets in the Emulex Configuration Tool panel.

    Configuring SCSI adapters

    Attention: Failure to disable the BIOS of attached nonstart devices may cause your system to attempt to restart from an unexpected nonstart device.

    Before you install and use SDD, you must configure your SCSI adapters. For SCSI adapters that are attached to start devices, ensure that the BIOS for the adapter is enabled. For all other adapters that are attached to nonstart devices, ensure that the BIOS for the adapter is disabled.

    Note:
    When the adapter shares the SCSI bus with other adapters, the BIOS must be disabled.

    Installing SDD 1.3.2.1

    The following section describes how to install SDD 1.3.2.1.

    Note:
    • Ensure that SDD is installed before adding additional paths to a device. Otherwise, the Windows 2000 server could lose the ability to access existing data on that device.
    • If you have not already done so, disconnect all cables that connect the host bus adapters and the ESS devices (while your Windows 2000 host system is turned off) before proceeding with the installation procedures.

    Perform the following steps to install SDD 1.3.2.1 on your system:

    1. Log on as the administrator user.
    2. Insert the SDD installation CD-ROM into the selected drive.
    3. Start the Windows 2000 Explorer program.
    4. Select the CD-ROM drive. A list of all the installed directories on the compact disc is displayed.
    5. Select the \win2k\IBMsdd directory (or your installation subdirectory)
    6. Run the setup.exe program. The setup program starts.

      Tip:

      • If you have previously installed an earlier version of SDD, you will see an "Upgrade?" question while the setup program is running. You must answer n to this question. After exiting from the setup program, go to Preparing for SDD 1.3.2.1 installation section for more information.
    7. Press Enter to exit the setup program.
    8. Shut down your Windows 2000 host system.
    9. Reconnect all cables that connect the host bus adapters and the ESS devices.
    10. Turn on your Windows 2000 host system.

    After completing the installation procedures and when you log on again, you will see a Subsystem Device Driver entry in your Program menu containing the following selections:

    1. Subsystem Device Driver management
    2. SDD Technical Support Web site
    3. README file
    Note:

    Displaying the current version of the Subsystem Device Driver

    You can display the current version of SDD on a Windows 2000 host system by viewing the sddbus.sys file properties. Perform the following steps to view the properties of sddbus.sys file:

    1. Click Start --> Run --> Programs --> Accessories --> Windows Explorer to open Windows Explorer.
    2. In Windows Explorer, go to your_installation_directory_drive_letter:\Winnt\system32\drivers directory

      where your_installation_directory_drive_letter is the letter of the directory in which you have installed the sddbus.sys file.

    3. Click the sddbus.sys file in your_installation_directory_drive_leltter:\Winnt\system32\drivers directory.
    4. Right-click on the sddbus.sys file, and then click Properties. The sddbus.sys properties window opens.
    5. In the sddbus.sys properties window, click Version. The file version and copyright information about the sddbus.sys file is displayed.

    Uninstalling the Subsystem Device Driver

    Perform the following steps to uninstall SDD on a Windows 2000 host system:

    1. Shut down your Windows 2000 host system.
    2. Disconnect all cables that connect the host bus adapters and the ESS devices.
    3. Turn on your Windows 2000 host system.
    4. Log on as the administrator user.
    5. Click Start --> Settings --> Control Panel. The Control Panel opens.
    6. Double click Add/Remove Programs. The Add/Remove Programs window opens.
    7. In the Add/Remove Programs window, select the Subsystem Device Driver from the currently installed programs selection list.
    8. Click on the Add/Remove button.
    9. Restart your system.

      Attention:

      • You must install SDD 1.3.2.1 immediately after performing a system restart to avoid any potential data loss. Go to Installing SDD 1.3.2.1 for instructions.
      • If you are not planning to reinstall the Subsystem Device Driver after the uninstallation, ensure that there is single path connection from the system to the storage device before performing a restart to avoid any potential data loss.

    Adding paths to SDD devices

    To activate SDD, you need to restart your Windows 2000 system after it is installed.

    Attention: Ensure that SDD is installed before you add additional paths to a device. Otherwise, the Windows 2000 server could lose the ability to access existing data on that device.

    Before adding any additional hardware, review the configuration information for the adapters and devices currently on your Windows 2000 server. Perform the following steps to display information about the adapters and devices:

    1. You must log on as an administrator user to have access to the Windows 2000 Computer Management.
    2. Click Start --> Program --> Subsystem Device Driver --> Subsystem Device Driver Management. An MS-DOS window opens.
    3. Type datapath query adapter and press Enter. The output includes information about all the installed adapters. In the example shown in the following output, one SCSI adapter is installed:
      +--------------------------------------------------------------------------------+
      |Active Adapters :1                                                              |
      |                                                                                |
      |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
      |    0  Scsi Port1 Bus0  NORMAL   ACTIVE       4057          0      8       8    |
      +--------------------------------------------------------------------------------+
    4. Type datapath query device and press Enter. In the example shown in the following output, 8 devices are attached to the SCSI path:
      +--------------------------------------------------------------------------------+
      |Total Devices : 8                                                               |
      |                                                                                |
      |DEV#:   0  DEVICE NAME: Disk7 Part7  TYPE: 2105E20   SERIAL: 01312028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk7 Part0     OPEN   NORMAL       1045          0    |
      |                                                                                |
      |DEV#:   1  DEVICE NAME: Disk6 Part6  TYPE: 2105E20   SERIAL: 01212028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk6 Part0     OPEN   NORMAL        391          0    |
      |                                                                                |
      |DEV#:   2  DEVICE NAME: Disk5 Part5  TYPE: 2105E20   SERIAL: 01112028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk5 Part0     OPEN   NORMAL       1121          0    |
      |                                                                                |
      |DEV#:   3  DEVICE NAME: Disk4 Part4  TYPE: 2105E20   SERIAL: 01012028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk4 Part0     OPEN   NORMAL        332          0    |
      |                                                                                |
      |DEV#:   4  DEVICE NAME: Disk3 Part3  TYPE: 2105E20   SERIAL: 00F12028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk3 Part0     OPEN   NORMAL        375          0    |
      |                                                                                |
      |DEV#:   5  DEVICE NAME: Disk2 Part2  TYPE: 2105E20   SERIAL: 31412028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk2 Part0     OPEN   NORMAL        258          0    |
      |                                                                                |
      |DEV#:   6  DEVICE NAME: Disk1 Part1  TYPE: 2105E20   SERIAL: 31312028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk1 Part0     OPEN   NORMAL        267          0    |
      |                                                                                |
      |DEV#:   7  DEVICE NAME: Disk0 Part0  TYPE: 2105E20   SERIAL: 31212028           |
      |=====================================================================           |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk0 Part0     OPEN   NORMAL        268          0    |
      |                                                                                |
      +--------------------------------------------------------------------------------+

    Activating additional paths

    Perform the following steps to activate additional paths to a vpath device:

    1. Install any additional hardware on the Windows 2000 server or the ESS.
    2. Restart the Windows 2000 server.
    3. Verify that the path is added correctly. See Verifying additional paths are installed correctly.

    Verifying additional paths are installed correctly

    After installing additional paths to SDD devices, verify that the additional paths have been installed correctly.

    Perform the following steps to verify that the additional paths have been installed correctly:

    1. Click Start --> Program --> Subsystem Device Driver --> Subsystem Device Driver Management. An MS-DOS window opens.
    2. Type datapath query adapter and press Enter. The output includes information about any additional adapters that were installed. In the example shown in the following output, an additional SCSI adapter has been installed:
      +--------------------------------------------------------------------------------+
      |Active Adapters :2                                                              |
      |                                                                                |
      |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
      |    0  Scsi Port1 Bus0  NORMAL   ACTIVE       1325          0      8       8    |
      |    1  Scsi Port2 Bus0  NORMAL   ACTIVE       1312          0      8       8    |
      +--------------------------------------------------------------------------------+
    3. Type datapath query device and press Enter. The output should include information about any additional devices that were installed. In this example, the output includes information about the new SCSI adapter and the new device numbers that were assigned. The following output is displayed:
      +--------------------------------------------------------------------------------+
      |Total Devices : 8                                                               |
      |                                                                                |
      |DEV#:   0  DEVICE NAME: Disk7 Part7  TYPE: 2105E20   SERIAL: 01312028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk7 Part0     OPEN   NORMAL        190          0    |
      |    1   Scsi Port2 Bus0/Disk15 Part0     OPEN   NORMAL        179          0    |
      |                                                                                |
      |DEV#:   1  DEVICE NAME: Disk6 Part6  TYPE: 2105E20   SERIAL: 01212028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk6 Part0     OPEN   NORMAL        179          0    |
      |    1   Scsi Port2 Bus0/Disk14 Part0     OPEN   NORMAL        184          0    |
      |                                                                                |
      |DEV#:   2  DEVICE NAME: Disk5 Part5  TYPE: 2105E20   SERIAL: 01112028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk5 Part0     OPEN   NORMAL        194          0    |
      |    1   Scsi Port2 Bus0/Disk13 Part0     OPEN   NORMAL        179          0    |
      |                                                                                |
      |DEV#:   3  DEVICE NAME: Disk4 Part4  TYPE: 2105E20   SERIAL: 01012028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk4 Part0     OPEN   NORMAL        187          0    |
      |    1   Scsi Port2 Bus0/Disk12 Part0     OPEN   NORMAL        173          0    |
      |                                                                                |
      |DEV#:   4  DEVICE NAME: Disk3 Part3  TYPE: 2105E20   SERIAL: 00F12028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk3 Part0     OPEN   NORMAL        215          0    |
      |    1   Scsi Port2 Bus0/Disk11 Part0     OPEN   NORMAL        216          0    |
      |                                                                                |
      |DEV#:   5  DEVICE NAME: Disk2 Part2  TYPE: 2105E20   SERIAL: 31412028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk2 Part0     OPEN   NORMAL        115          0    |
      |    1   Scsi Port2 Bus0/Disk10 Part0     OPEN   NORMAL        130          0    |
      |                                                                                |
      |DEV#:   6  DEVICE NAME: Disk1 Part1  TYPE: 2105E20   SERIAL: 31312028           |
      |=======================================================================         |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk1 Part0     OPEN   NORMAL        122          0    |
      |    1    Scsi Port2 Bus0/Disk9 Part0     OPEN   NORMAL        123          0    |
      |                                                                                |
      |DEV#:   7  DEVICE NAME: Disk0 Part0  TYPE: 2105E20   SERIAL: 31212028           |
      |=========================================================================       |
      |Path#              Adapter/Hard Disk    State     Mode     Select     Errors    |
      |    0    Scsi Port1 Bus0/Disk0 Part0     OPEN   NORMAL        123          0    |
      |    1    Scsi Port2 Bus0/Disk8 Part0     OPEN   NORMAL        128          0    |
      +--------------------------------------------------------------------------------+

    Support for Windows 2000 clustering

    SDD 1.3.0.0 or higher is required to support Windows 2000 clustering. SDD 1.3.0.0 or higher does not support I/O load-balancing in a Windows 2000 clustering environment.

    When running Windows 2000 clustering, failover/failback may not occur when the last path is being removed from the shared resources. See Microsoft article Q294173 for additional information. Windows 2000 does not support dynamic disks in the MSCS environment.

    Special considerations in the Windows 2000 clustering environment

    There are subtle differences in the way that SDD handles path reclamation in a Windows 2000 clustering environment compared to a nonclustering environment. When the Windows 2000 server loses a path in a nonclustering environment, the path condition changes from open to dead and the adapter condition changes from active to degraded. The adapter and path condition will not change until the path is made operational again. When the Windows 2000 server loses a path in a clustering environment, the path condition changes from open to dead and the adapter condition changes from active to degraded. However, after a period of time, the path condition changes back to open and the adapter condition changes back to normal, even if the path has not been made operational again.

    The datapath set adapter # offline command operates differently in a clustering environment as compared to a nonclustering environment. In a clustering environment, the datapath set adapter offline command does not change the condition of the path if the path is active or being reserved. If you issue the command, the following message is displayed: to preserve access some paths left online.

    Configuring a Windows 2000 cluster with SDD

    The following variables are used in this procedure:

    server_1
    represents the first server with two host bus adapters (HBAs).

    server_2
    represents the second server with two HBAs.

    hba_a
    represents the first HBA for server_1

    hba_b
    represents the second HBA for server_1

    hba_c
    represents the first HBA for server_2

    hba_d
    represents the second HBA for server_2

    Perform the following steps to configure a Windows 2000 cluster with SDD:

    1. Configure LUNs on the ESS as shared for all HBAs on both server_1 and server_2.
    2. Connect hba_a to the ESS, and restart server_1.
    3. Click Start --> Programs --> Administrative Tools--> Computer Management. The Computer Management window opens. From the Computer Management window, select Storage and then Disk Management to work with the storage devices attached to the host system.

      The operating system will recognize each additional path to the same LUN as a device.

    4. Disconnect hba_a and connect hba_b to the ESS. Restart server_1.
    5. Click Start --> Programs --> Administrative Tools--> Computer Management. The Computer Management window opens. From the Computer Management window, select Storage and then Disk Management to verify the correct number of LUNs that are connected to server_1.

      If the number of LUNs that are connected to server_1 is correct, proceed to 6.

      If the number of LUNs that are connected to server_1 is incorrect, perform the following steps:

      1. Verify that the cable for hba_b is connected to the ESS.
      2. Verify your LUN configuration on the ESS.
      3. Repeat steps 2 - 5.
    6. Install SDD on server_1, and restart server_1.

      For installation instructions , go to Installing SDD 1.3.2.1 section.

    7. Connect hba_c to the ESS, and restart server_2.
    8. Click Start --> Programs --> Administrative Tools--> Computer Management. The Computer Management window opens. From the Computer Management window, select Storage and then Disk Management to verify the correct number of LUNs that are connected to server_2.

      The operating system will see each additional path to the same LUN as a device.

    9. Disconnect hba_c and connect hba_d to the ESS. Restart server_2.
    10. Click Start --> Programs --> Administrative Tools--> Computer Management. The Computer Management window is displayed. From the Computer Management window, select Storage and then Disk Management to verify the correct number of LUNs that are connected to server_2.

      If the number of LUNs that are connected to server_2 is correct, proceed to 11.

      If the number of LUNs that are connected to server_2 is incorrect, perform the following steps:

      1. Verify that the cable for hba_d is connected to the ESS.
      2. Verify your LUN configuration on the ESS.
      3. Repeat steps 7 - 10.
    11. Install SDD on server_2, and restart server_2.

      For installation instructions , go to Installing SDD 1.3.2.1 section.

    12. Connect both hba_c and hba_d on server_2 to the ESS, and restart server_2.
    13. Use the datapath query adapter and datapath query device commands to verify the correct number of LUNs and paths on server_2.
    14. Click Start --> Programs --> Administrative Tools--> Computer Management. The Computer Management window opens. From the Computer Management window, select Storage and then Disk Management to verify that the actual number of LUNs as online devices is correct.
    15. Format the raw devices with NTFS.

      Make sure to keep track of the assigned drive letters on server_2.

    16. Connect both hba_a and hba_b on server_1 to the ESS, and restart server_1.
    17. Use the datapath query adapter and datapath query device commands to verify the correct number of LUNs and paths on server_1.

      Verify that the assigned drive letters on server_1 match the assigned drive letters on server_2.

    18. Restart server_2.
    19. Use the datapath query adapter and datapath query device commands to verify the correct number of LUNs and paths on server_1 and server_2. (This step is optional.)
      Note:
      You can use the datapath query adapter and datapath query device commands to show all the physical and logical volumes for the host server. The secondary server only shows the physical volumes and the logical volumes that it owns.

    Chapter 5. SDD for a Hewlett-Packard host system

    This chapter provides step-by-step procedures for you to install, configure, remove, and use the SDD on a Hewlett-Packard (HP) host system that is attached to an ESS. For updated and additional information not included in this manual, please see the README file on the compact disc or go to the SDD Web site at:

    www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates


    Verifying the hardware and software requirements

    You must install the following hardware and software components to ensure that SDD installs and operates successfully.

    Hardware

    Software

    Non-supported environments

    SDD does not support the following environments:

    SDD 1.3.0.5 support on HP-UX 11.0 and HP-UX 11i

    SDD supports the following ESS attachments:

    Attention: The patches in Table 19 (as appropriate for a 32-bit or 64-bit application) must be installed on your HP-UX 11.0 host system to ensure that SDD operates successfully. Do not install these patches on your HP-UX 11i host system.

    Understanding how SDD works on an HP host system

    SDD resides above the HP SCSI disk driver (sdisk) in the protocol stack. For more information about how SDD works, see Subsystem Device Driver.


    Preparing for SDD installation

    Before you install SDD, you must configure the ESS to your host system and attach the required SCSI or fibre-channel adapters.

    Configuring the ESS

    Before you install SDD, configure your ESS for single-port or multiport access for each LUN. SDD requires a minimum of two independent paths that share the same logical unit to use the load-balancing and path-failover protection features.

    For information about configuring your ESS, see IBM Enterprise Storage Server Introduction and Planning Guide.

    Planning for installation

    Before you install SDD on your HP host system, you need to understand what kind of software runs on your host. The way you install SDD depends on the kind of software you have running. There are two types of special device files that are supported:

    There are three possible scenarios for installing SDD. The scenario you choose depends on the kind of software you have installed:

    Scenario 1
    Your system has no software applications (other than UNIX) or DBMS that communicates directly to the HP-UX disk device layer.

    Scenario 2
    Your system already has a software application or DBMS, such as Oracle, that communicates directly with the HP-UX disk device layer.

    Scenario 3
    Your system already has SDD and you want to upgrade the software.

    Table 18 further describes the various installation scenarios and how you should proceed.

    Table 18. SDD installation scenarios
    Installation scenario Description How to proceed
    Scenario 1
    • SDD is not installed.
    • No software application or DBMS communicates directly to sdisk interface.

    Go to:
    1. Installing the Subsystem Device Driver
    2. Standard UNIX applications
    Scenario 2
    • SDD is not installed.
    • An existing application package or DBMS communicates directly to the sdisk interface.

    Go to:
    1. Installing the Subsystem Device Driver
    2. Using applications with SDD
    Scenario 3 SDD is installed. Go to Upgrading the Subsystem Device Driver

    The patches in Table 19 (as appropriate for a 32-bit or 64-bit application) must be installed on your HP-UX 11.0 host system to ensure that SDD operates successfully. Do not install the patches listed on your HP-UX 11i host system.

    Table 19. HP 11.0 patches necessary for proper operation of SDD
    Application mode HP patch Patch description
    32-bit PHKL_20674 Fix VxFS unmount hang & NMF, sync panics
    32-bit PHKL_20915 Trap-related panics/hangs
    32-bit PHKL_21834 Fibre channel Mass Storage Driver Patch
    32-bit PHKL_22759 SCSI IO Subsystem Cumulative patch
    32-bit PHKL_23001 Signal, threads, spinlock, scheduler, IDS, q3p
    32-bit PHKL_23406 Probe, sysproc, shmem, thread cumulative patch
    32-bit or 64-bit PHKL_21392 VxFS performance, hang, icache, DPFs
    32-bit or 64-bit PHKL_21624 start, JFS, PA8600, 3Gdata, NFS, IDS, PM, VM, async
    32-bit or 64-bit PHKL_21989 SCSI IO Subsystem Cumulative patch
    64-bit PHKL_21381 Fibre Channel Mass Storage driver


    Installing the Subsystem Device Driver

    Before you install SDD, make sure that you have root access to your HP host system and that all the required hardware and software is ready.

    Perform the following steps to install SDD on your HP host system:

    1. Make sure that the SDD compact disc (CD) is available.
    2. Insert the CD into your CD-ROM drive.
    3. Mount the CD-ROM drive using the mount command. Here are two examples of the mount command:
      mount /dev/dsk/c0t2d0 /cdrom
      

      or

      mount /dev/dsk/c0t2d0 /your_installation_directory
      

      where /cdrom or /your_installation_directory is the name of the directory to which you want to mount the CD-ROM drive.

    4. Run the sam program.
      > sam
      
    5. Select Software Management.
    6. Select Install Software to Local Host.
    7. At this point, the SD Install - Software Selection panel is displayed. Almost immediately afterwards, a Specify Source menu is displayed:
      1. For Source Depot Type, select the local CD-ROM.
      2. For Source Depot Path, choose the directory and the IBMdpo.depot file.

        For 32-bit mode applications, use:

        /cdrom/hp32bit/IBMdpo.depot

        or

        /your_installation_directory/hp32bit/IBMdpo.depot

        For 64-bit mode applications, use:

        /cdrom/hp64bit/IBMdpo.depot

        or

        /your_installation_directory/hp32bit/IBMdpo.depot

      3. Click OK.

        You will see output similar to that in either Figure 3 or Figure 4.

        Figure 3. IBMdpo Driver 32-bit

        +--------------------------------------------------------------------------------+
        |Name           Revision          Information                        Size(Kb)    |
        |IBMdpo_tag ->  B.11.00.01        IBMdpo Driver 32-bit               nnnn        |
        +--------------------------------------------------------------------------------+

        Figure 4. IBMdpo Driver 64-bit

        +--------------------------------------------------------------------------------+
        |Name           Revision          Information                        Size(Kb)    |
        |IBMdpo_tag ->  B.11.00.01        IBMdpo Driver 64-bit               nnnn        |
        +--------------------------------------------------------------------------------+
    8. Click the IBMdpo_tag product.
    9. Click Actions from the Bar menu, and then click Mark for Install.
    10. Click Actions from the Bar menu, and then click Install (analysis). An Install Analysis panel, is displayed showing the status of Ready.
    11. Click OK to proceed. A Confirmation window opens which states that the installation will begin.
    12. Type Yes and press Enter. The analysis phase starts.
    13. After the analysis phase has finished, another Confirmation window opens informing you that the system will be restarted after installation is complete. Type Yes and press Enter. The installation of IBMdpo will now proceed.
    14. An Install window opens informing you about the progress of the IBMdpo software installation. The window looks similar to the following:

      +--------------------------------------------------------------------------------+
      |Press 'Product Summary' and/or 'Logfile' for more target information.           |
      |Target          : XXXXX                                                         |
      |Status          : Building kernel                                               |
      |Percent Complete     : 17%                                                      |
      |Kbytes Installed   :  276 of 1393                                               |
      |Time Left (minutes) : 1                                                         |
      |Product Summary       Logfile                                                   |
      |Done                             Help                                           |
      +--------------------------------------------------------------------------------+
      The Done option is not available when the installation is in progress. It becomes available after the installation process completes.
    15. Click Done. A Note window opens informing you that the local system will restart with the newly installed software.
    16. Click OK to proceed. Before the machine console restarts, a message similar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |* A reboot of this system is being invoked. Please wait.                        |
      |                                                                                |
      |*** FINAL System shutdown message (XXXXX) ***                                   |
      |System going down IMMEDIATELY                                                   |
      +--------------------------------------------------------------------------------+
    Note:
    You can use the datapath query device command to verify the SDD installation. SDD is successfully installed if the command executes successfully.

    Post-installation

    After SDD is installed, the device driver resides above the HP SCSI disk driver (sdisk) in the protocol stack. In other words, SDD now communicates to the HP-UX device layer. The SDD software installation procedure installs a number of SDD components and updates some system files. Those components and files are listed in the following tables.

    Table 20. SDD components installed for HP host systems
    File Location Description
    libvpath.a /usr/conf/lib SDD device driver
    vpath /usr/conf/master.d SDD configuration file
    Executables /opt/IBMdpo/bin Configuration and status tools
    README.sd /opt/IBMdpo README file
    defvpath /sbin SDD configuration file used during startup

    Table 21. System files updated for HP host systems
    File Location Description
    system /stand/build Forces the loading of the SDD device driver
    lvmrc /etc Causes the defvpath command to run at start time

    Table 22. SDD commands and their descriptions for HP host systems
    Command Description
    cfgvpath Configures vpath devices
    defvpath Second part of the cfgvpath command configuration during startup time
    showvpath Lists the configuration mapping between SDD devices and underlying disks
    datapath SDD driver console command tool

    If you are not using a DBMS or an application package that communicates directly to the sdisk interface, the installation procedure is nearly complete. However, you still need to customize HP-UX so that standard UNIX applications can use SDD. Go to Standard UNIX applications for instructions. If you have a DBMS or an application package installed that communicates directly to the sdisk interface, such as Oracle, go to Using applications with SDD and read the information specific to the application you are using.

    During the installation process, the following files were copied from the IBMdpo_depot to the system:

    # Kernel-related files

    # SDD driver related files

    In addition, the /stand/vmunix kernel was created with the device driver. The /stand/system directory was modified in order to add the device driver entry into the file. After these files were created, the /opt/IBMdpo/bin/cfgvpath program was initiated in order to create vpaths in the /dev/dsk and /dev/rdsk directories for all IBM disks which are available on the system. This information is stored in the /opt/IBMdpo file for use after restarting the machine.

    Note:
    SDD devices are found in /dev/rdsk and /dev/dsk. The device is named according to the SDD number. A device with a number of 0 would be /dev/rdsk/vpath0.

    Upgrading the Subsystem Device Driver

    Upgrading the SDD consists of removing and reinstalling the IBMdpo package. If you are upgrading SDD, go to *** and then go to Installing the Subsystem Device Driver.

    Uninstalling the Subsystem Device Driver

    The following procedure explains how to remove the SDD. You must uninstall the current level of SDD before upgrading to a newer level.

    Complete the following procedure to uninstall SDD:

    1. Restart or unmount all SDD file systems.
    2. If you are using SDD with a database, such as Oracle, edit the appropriate database configuration files (database partition) to remove all the SDD devices.
    3. Run the sam program.
      > sam
      
    4. Click Software Management.
    5. Click Remove Software.
    6. Click Remove Local Host Software.
    7. Click the IBMdpo_tag selection.
      1. Click Actions from the Bar menu, and then select Mark for Remove.
      2. Click Actions from the Bar menu, and then select Remove (analysis). A Remove Analysis window opens and shows the status of Ready.
      3. Click OK to proceed. A Confirmation window opens and indicates that the uninstallation will begin.
      4. Type Yes. The analysis phase starts.
      5. After the analysis phase has finished, another Confirmation window opens indicating that the system will restarted after the uninstallation is complete. Type Yes and press Enter. The uninstallation of IBMdpo begins.
      6. An Uninstall window opens showing the progress of the IBMdpo software uninstallation. This is what the panel looks like:
        +--------------------------------------------------------------------------------+
        |Target        : XXXXX                                                           |
        |Status        : Executing unconfigure                                           |
        |Percent Complete    : 17%                                                       |
        |Kbytes Removed      : 340 of 2000                                               |
        |Time Left (minutes) : 5                                                         |
        |Removing Software   : IBMdpo_tag,...........                                    |
        +--------------------------------------------------------------------------------+
        The Done option is not available when the installation is in progress. It becomes available after the installation process completes.
    8. Click Done. A Note window opens informing you that the local system will restart with the newly installed software.
    9. Click OK to proceed. The following message is displayed on the machine console before it restarts:
      +--------------------------------------------------------------------------------+
      |* A reboot of this system is being invoked. Please wait.                        |
      |                                                                                |
      |*** FINAL System shutdown message (XXXXX) ***                                   |
      |System going down IMMEDIATELY                                                   |
      +--------------------------------------------------------------------------------+
    Note:
    When SDD has been successfully uninstalled, the first part of the procedure for upgrading the SDD is complete. To complete an upgrade, you need to reinstall SDD. See the installation procedure in Installing the Subsystem Device Driver.

    Changing an SDD hardware configuration

    When adding or removing multiport SCSI devices, you must reconfigure SDD to recognize the new devices. Perform the following steps to reconfigure SDD:

    1. Restart the system by typing:
      shutdown -r 0
      
    2. Issue the cfgvpath command to reconfigure the vpath by typing:
      /opt/IBMdpo/bin/cfgvpath -c
      
    3. Restart the system by typing:
      shutdown -r 0
      

    Using applications with SDD

    If your system already has a software application or a DBMS installed that communicates directly with the HP-UX disk device drivers, you need to insert the new SDD device layer between the software application and the HP-UX disk device layer. You also need to customize the software application to have it communicate with the SDD devices instead of the HP-UX devices.

    In addition, many software applications and DBMSs need to control certain device attributes such as ownership and permissions. Therefore, you must ensure that the new SDD devices that these software applications or DBMSs access in the future have the same attributes as the HP-UX sdisk devices that they replace. You need to customize the application or DBMS to accomplish this.

    This section contains the procedures for customizing the following software applications and DBMS for use with SDD:

    Standard UNIX applications

    If you have not already done so, install SDD using the procedure in Installing the Subsystem Device Driver. When this is done, SDD resides above the HP SCSI disk driver (sdisk) in the protocol stack. In other words, SDD now communicates to the HP-UX device layer. To use standard UNIX applications with SDD, you must make some changes to your logical volumes. You must convert your existing logical volumes or create new ones.

    Standard UNIX applications such as newfs, fsck, mkfs, and mount, which normally take a disk device or raw disk device as a parameter, also accept the SDD device as a parameter. Similarly, entries in files such as vfstab and dfstab (in the format of cntndnsn) can be replaced by entries for the corresponding SDD vpathNs devices. Make sure that the devices that are replaced are replaced with the corresponding SDD device. Issue the showvpath command to list all SDD devices and their underlying disks.

    To use the SDD driver for an existing logical volume, you must remove the existing logical volume and volume group and recreate it using the SDD device.

    Attention: Do not use the SDD for critical file systems needed at startup, such as /(root), /stand, /usr, /tmp or /var. Doing so may render your system unusable if SDD is ever uninstalled (for example, as part of an upgrade).

    Creating new logical volumes

    The task of creating a new logical volume to use SDD consists of the following subtasks.

    Note:
    You must have super-user privileges to perform the following subtasks.
    1. Determining the major number of the logical volume device
    2. Creating a device node for the logical volume device
    3. Creating a physical volume
    4. Creating a volume group
    5. Creating a logical volume
    6. Creating a file system on the volume group
    7. Mounting the logical volume.

    To create a new logical volume that uses SDD, you first need to determine the major number of the logical volume device.

    Determining the major number of the logical volume device

    Type the following command to determine the major number:

    # lsdev | grep lv
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |64          64         lv              lvm                                      |
    +--------------------------------------------------------------------------------+

    The first number in the message is the major number of the character device, which is the number you want to use.

    Creating a device node for the logical volume device

    Creating a device node actually consists of:

    1. Creating a subdirectory in the /dev directory for the volume group
    2. Changing to the /dev directory
    3. Creating a device node for the logical volume device

    If you do not have any other logical volume devices, you can use a minor number of 0x010000. In this example, assume that you have no other logical volume devices. A message similar to the following is displayed:

    # mknod group c 64 0x010000
    

    Create a physical volume, by performing the procedure in Creating a physical volume.

    Creating a directory in /dev for the volume group

    Type the following command to create a subdirectory in the /dev directory for the volume group:

    # mkdir /dev/vgibm
    

    In this example, vgibm is the name of the directory.

    Next, change to the directory that you just created.

    Changing to the /dev directory

    Type the following command to change to the /dev directory:

    # cd /dev/vgibm
    

    Next, recreate a device node for the logical volume device.

    Recreating the physical volume

    Type the following command to recreate the physical volume:

    # pvcreate /dev/rdsk/vpath0
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |Physical volume "/dev/rdsk/vpath0" has been successfully created.               |
    +--------------------------------------------------------------------------------+
    In this example, the SDD device associated with the underlying disk is vpath0. Verify the underlying disk by typing the following showvpath command:
    # /opt/IBMdpo/bin/showvpath
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |vpath0:                                                                         |
    |	/dev/dsk/c3t4d0                                                                |
    +--------------------------------------------------------------------------------+

    Next, create the physical volume.

    Creating a physical volume

    Type the following command to create a physical volume:

    # pvcreate /dev/rdsk/vpath0
    

    Now create the volume group.

    Creating a volume group

    Type the following command to create a volume group:

    # vgcreate /dev/vgibm /dev/dsk/vpath0
    

    Now create the logical volume.

    Creating a logical volume

    Type the following command to create logical volume lvol1:

    # lvcreate -L 100 -n lvol1 vgibm
    

    The -L 100 portion of the command makes a 100-MB volume group; you can make it larger if you want to. Now you are ready to create a file system on the volume group.

    Creating a file system on the volume group

    Type the following command to create a file system on the volume group:

    # newfs -F hfs /dev/vgibm/rlvol1
    

    Finally, mount the logical volume. This example assumes that you have a mount point called /mnt.

    Mounting the logical volume

    Type the following command to mount the logical volume lvol1:

    # mount /dev/vgibm/lvol1 /mnt
    

    Attention: In some cases it may be necessary to use standard HP recovery procedures to fix a volume group that has become damaged or corrupted. For information about using recovery procedures, such as, vgscan, vgextend, vpchange, or vgreduce, see the HP-UX Reference Volume 2 at the following Web site:

    docs.hp.com

    Removing logical volumes

    Perform the following procedures to remove logical volumes.

    Removing the existing logical volume

    Before the logical volume is removed, it must be unmounted. For example, type the following command to unmount logical volume lvol1:

    # umount /dev/vgibm/lvol1
    

    Next, remove the logical volume.

    For example, type the following command to remove logical volume lvol1:

    # lvremove /dev/vgibm/lvol1
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |The logical volume "/dev/vgibm/lvol1" is not empty;                             |
    |do you really want to delete the logical volume (y/n)                           |
    +--------------------------------------------------------------------------------+

    Type y and press Enter. A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |Logical volume "/dev/vgibm/lvol1" has been successfully removed.                |
    |Volume Group configuration for /dev/vgibm has been saved in                     |
    |/etc/lvmconf/vgibm.conf                                                         |
    +--------------------------------------------------------------------------------+

    When prompted to delete the logical volume, type y.

    Next, remove the volume group.

    Removing the existing volume group

    Type the following command to remove the volume group vgibm:

    # vgremove /dev/vgibm
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |Volume group "/dev/vgibm" has been successfully removed.                        |
    +--------------------------------------------------------------------------------+

    Now recreate the logical volume.

    Recreating the existing logical volumes

    The task of converting an existing logical volume to use SDD consists of the following subtasks:

    1. Determining the size of the logical volume
    2. Recreating the physical volume
    3. Recreating the volume group
    4. Recreating the logical volume
    5. Setting the correct timeout value for the logical volume manager
    Note:
    You must have super-user privileges to perform these subtasks.

    As an example, suppose you have a logical volume called lvol1 under a volume group vgibm, which is currently using the disk directly, (for example, through path /dev path /dev/dsk/c3t4d0). You would like to convert logical volume lvol1 to use SDD. To recreate the logical volume, you first need to determine the size of the logical volume.

    Determining the size of the logical volume

    Type the following command to determine the size of the logical volume:

    # lvdisplay | grep LV Size
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |LV Size (Mbytes) 100                                                            |
    +--------------------------------------------------------------------------------+

    In this case, the logical volume size is 100 MB. Next, recreate the physical volume.

    Recreating the physical volume

    Type the following command to recreate the physical volume:

    # pvcreate /dev/rdsk/vpath0
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |Physical volume "/dev/rdsk/vpath0" has been successfully created.               |
    +--------------------------------------------------------------------------------+

    In this example, the SDD device associated with the underlying disk is vpath0. Verify the underlying disk by typing the following command:

    # /opt/IBMdpo/bin/showvpath
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |vpath0:                                                                         |
    |	/dev/dsk/c3t4d0                                                                |
    +--------------------------------------------------------------------------------+

    Next, recreate the volume group.

    Recreating the volume group

    Type the following command to recreate the volume group:

    # vgcreate /dev/vgibm /dev/dsk/vpath0
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |Increased the number of physical extents per physical volume to 2187.           |
    |Volume group "/dev/vgibm" has been successfully created.                        |
    |Volume Group configuration for /dev/vgibm has been saved in                     |
    |/etc/lvmconf/vgibm.conf                                                         |
    +--------------------------------------------------------------------------------+

    Now recreate the logical volume.

    Recreating the logical volume

    Recreating the logical volume consists of a number of smaller steps:

    1. Recreating the physical volume
    2. Recreating the volume group
    3. Recreating the logical volume
    4. Setting the proper timeout value for the logical volume manager

    Attention: The recreated logical volume should be the same size as the original volume; otherwise, the recreated volume cannot store the data that was on the original.

    Type the following command to recreate the logical volume:

    # lvcreate -L 100 -n lvol1 vgibm
    

    A message similar to the following is displayed:

    +--------------------------------------------------------------------------------+
    |Logical volume "/dev/vgibm/lvol1" has been successfully created with            |
    |character device "/dev/vgibm/rlvol1".                                           |
    |Logical volume "/dev/vgibm/lvol1" has been successfully extended.               |
    |Volume Group configuration for /dev/vgibm has been saved in                     |
    |/etc/lvmconf/vgibm.conf                                                         |
    +--------------------------------------------------------------------------------+

    The -L 100 parameter comes from the size of the original logical volume, which is determined by using the lvdisplay command. In this example, the original logical volume was 100 MB in size.

    Setting the correct timeout value for the logical volume manager

    Attention: The timeout values for the logical volume manager must be set correctly for SDD to operate properly. This is particularly true if you are going to be using concurrent microcode download.

    If you are going to be using concurrent microcode download with single-path SCSI, perform the following steps to set the correct timeout value for the logical volume manager:

    1. Ensure that the timeout value for an SDD logical volume is set to the default. Type lvdisplay /dev/vgibm/lvol1 and press Enter. If the timeout value is not default, type lvchange -t 0 /dev/vgibm/lvol1 and press Enter to change it. (In this example, vgibm is the name of the logical volume group that was previously configured to use SDD; in your environment the name may be different.)
    2. Change the timeout value for an SDD physical volume to 240. Type pvchange -t 240 /dev/dsk/vpathn and press Enter. (n refers to the vpath number.) If you are not sure about the vpath number, type /opt/IBMdpo/bin/showvpath and press Enter to obtain this information.

    If you are going to be using concurrent microcode download with multipath SCSI, perform the following steps to set the proper timeout value for the logical volume manager:

    1. Ensure that the timeout value for an SDD logical volume is set to the default. Type lvdisplay /dev/vgibm/lvoly and press Enter. If the timeout value is not default, type lvchange -t 0 /dev/vgibm/lvoly and press Enter to change it. (In this example, vgibm is the name of logical volume group that was previously configured to use SDD; in your environment the name may be different.)
    2. Change the timeout value for an SDD physical volume to 240. Type pvchange -t 240 /dev/dsk/vpathn and press Enter. (n refers to the vpath number.) If you are not sure about the vpath number, type /opt/IBMdpo/bin/showvpath and press Enter to obtain this information.
    3. The recreated logical volume must be mounted before it can be accessed.

    Attention: In some cases it may be necessary to use standard HP recovery procedures to fix a volume group that has become damaged or corrupted. For information about using recovery procedures, such as, vgscan, vgextend, vpchange, or vgreduce, see the HP-UX Reference Volume 2 at the following Web site:

    docs.hp.com

    Installing SDD on a Network File System file server

    The procedures in this section show how to install SDD for use with an exported file system (Network File System file server).

    Setting up Network File System for the first time

    Perform the following steps if you are installing exported file systems on SDD devices for the first time:

    1. If you have not already done so, install SDD using the procedure in Installing the Subsystem Device Driver.
    2. Determine which SDD (vpathN) volumes you will use as file system devices.
    3. Create file systems on the selected SDD devices using the appropriate utilities for the type of file system that you will use. If you are using the standard HP-UX UFS file system, type the following command:
      # newfs /dev/rdsk/vpathN
      

      In this example, N is the SDD device instance of the selected volume. Create mount points for the new file systems.

    4. Install the file systems into the directory /etc/fstab. Click yes in the mount at boot field.
    5. Install the file system mount points into the /etc/exports directory for export.
    6. Restart the system.

    Installing SDD on a system that already has the Network File System file server

    Perform the following steps if you have the Network File System file server already configured to:

    1. List the mount points for all currently exported file systems by looking in the /etc/exports directory.
    2. Match the mount points found in step 1 with sdisk device link names (files named /dev/(r)dsk/cntndn) by looking in the /etc/fstab directory.
    3. Match the sdisk device link names found in step 2 with SDD device link names (files named /dev/(r)dsk/vpathN) by issueing the showvpath command.
    4. Make a backup copy of the current /etc/fstab file.
    5. Edit the /etc/fstab file, replacing each instance of an sdisk device link named /dev/(r)dsk/cntndn with the corresponding SDD device link.
    6. Restart the system.
    7. Verify that each exported file system:

      If there is a problem with any exported file system after completing step 7, restore the original /etc/fstab file and restart to restore Network File System service. Then review your steps and try again.

    Oracle

    You must have super-user privileges to perform the following procedures. You also need to have Oracle documentation on hand. These procedures were tested with Oracle 8.0.5 Enterprise server with the 8.0.5.1 patch set from Oracle.

    Installing an Oracle database for the first time

    You can set up your Oracle database in one of two ways. You can set it up to use a file system or raw partitions. The procedure for installing your database differs depending on the choice you make.

    Using a file system

    1. If you have not already done so, install SDD using the procedure in Installing the Subsystem Device Driver.
    2. Create and mount file systems on one or more SDD partitions. (Oracle recommends three mount points on different physical devices.)
    3. Follow the Oracle Installation Guide for instructions on installing to a file system. (During the Oracle installation, you will be asked to name three mount points. Supply the mount points for the file systems you created on the SDD partitions.)

    Using raw partitions

    Attention: When using raw partitions, make sure that the ownership and permissions of the SDD devices are the same as the ownership and permissions of the raw devices they are replacing. Make sure that all the databases are closed before making changes.

    In the following procedure you will be replacing the raw devices with the SDD devices.

    1. If you have not already done so, install SDD using the procedure in Installing the Subsystem Device Driver.
    2. Create the Oracle software owner user in the local server /etc/passwd file. You must also complete the following related activities:
      1. Complete the rest of the Oracle preinstallation tasks described in the Oracle8 Installation Guide. Plan the installation of Oracle8 on an file system residing on an SDD partition.
      2. Set up the Oracle user's ORACLE_BASE and ORACLE_ HOME environment variables to the directories of this file system.
      3. Create two more SDD-resident file systems on two other SDD volumes. Each of the resulting three mount points should have a subdirectory named oradata. The subdirectory is used as a control file and redo log location for the installer's default database (a sample database) as described in the Oracle8 Installation Guide. Oracle recommends using raw partitions for redo logs. To use SDD raw partitions as redo logs, create symbolic links from the three redo log locations to SDD raw device links (files named /dev/rdsk/vpathNs, where N is the SDD instance number, and s is the partition ID) that point to the slice.
    3. Determine which SDD (vpathN) volumes you will use as Oracle8 database devices.
    4. Partition the selected volumes using the HP-UX format utility. If SDD raw partitions are to be used by Oracle8 as database devices, be sure to leave disk cylinder 0 of the associated volume unused. This protects UNIX disk labels from corruption by Oracle8, as described in the Oracle8 Installation Guide.
    5. Ensure that the Oracle software owner has read and write privileges to the selected SDD raw partition device files under the /devices directory.
    6. Set up symbolic links from the oradata directory (under the first of the three mount points). Link the database files system <db>.dbf, tempdb.dbf, rbsdb.dbf, toolsd.bdbf, and usersdb.dbf to SDD raw device links (files named /dev/rdsk/vpathNs). Point to the partitions of the appropriate size, where db is the name of the database that you are creating. (The default is test.)
    7. Install the Oracle8 server following the instructions in the Oracle8 Installation Guide. Be sure to be logged in as the Oracle software owner when you run the orainst /m command. Select the Install New Product - Create Database Objects option. Select Raw Devices for the storage type. Specify the raw device links set up in steps 2 and 6 for the redo logs and database files of the default database.
    8. To set up other Oracle8 databases, you must set up control files, redo logs, and database files following the guidelines in the Oracle8 Administrator's Reference. Make sure any raw devices and file systems you set up reside on SDD volumes.
    9. Launch the sqlplus utility.
    10. Issue the create database SQL command, specifying the control, log, and system data files that you have set up.
    11. Issue the create tablespace SQL command to set up each of the temp, rbs, tools, and users database files that you created.
    12. Issue the create rollback segment SQL command to create the three redo log files that you set. For the syntax of these three create commands, see the Oracle8 Server SQL Language Reference Manual.

    Installing an SDD on a system that already has Oracle in place

    The installation procedure for a new SDD installation differs depending on whether you are using a file system or raw partitions for your Oracle database.

    If using a file system

    Perform the following procedure if you are installing SDD for the first time on a system with an Oracle database that uses a file system:

    1. Record the raw disk partitions being used (they are in the cntndnsn format) or the partitions where the Oracle file systems reside. You can get this information from the /etc/vfstab file if you know where the Oracle files are. Your database administrator can tell you where the Oracle files are, or you can check for directories with the name oradata.
    2. Complete the basic installation steps in Installing the Subsystem Device Driver.
    3. Change to the directory where you installed the SDD utilities. Issue the showvpath command.
    4. Check the display to see whether you find a cntndn directory that is the same as the one where the Oracle files are.
    5. Use the SDD partition identifiers instead of the original HP-UX identifiers when mounting the file systems.

      If you originally used the following HP-UX identifiers:

      mount /dev/dsk/c1t3d2 /oracle/mp1
      

      You now use the following SDD partition identifiers:

      mount /dev/dsk/vpath2 /oracle/mp1
      

      For example, assume that you found that vpath2 was the SDD identifier.

    Follow the instructions in the Oracle Installation Guide for setting ownership and permissions.

    If using raw partitions

    Perform the following procedure if you have Oracle8 already installed and want to reconfigure it to use SDD partitions instead of sdisk partitions (for example, partitions accessed through /dev/rdsk/cntndn files).

    All Oracle8 control, log, and data files are accessed either directly from mounted file systems or using links from the oradata subdirectory of each Oracle mount point set up on the server. Therefore, the process of converting an Oracle installation from sdisk to SDD has two parts:

    Converting an Oracle installation from sdisk to SDD

    Perform the following conversion steps:

    1. Back up your Oracle8 database files, control files, and redo logs.
    2. Obtain the sdisk device names for the Oracle8 mounted file systems by looking up the Oracle8 mount points in /etc/fstab and extracting the corresponding sdisk device link name (for example, /dev/rdsk/c1t4d0).
    3. Launch the sqlplus utility.
    4. Type the command:
      select * from sys.dba_data_files;
      

      Determine the underlying device where each data file resides, either by looking up mounted file systems in /etc/fstab or by extracting raw device link names directly from the select command output.

    5. Fill in the following table, which is for planning purposes:
      Oracle Device Link File Attributes SDD Device Link
      Owner Group Permissions
      /dev/rdsk/c1tld0 oracle dba 644 /dev/rdsk/vpath4
    6. Fill in column 2 by issuing the command ls -l on each device link listed in column 1 and extracting the link source device file name.
    7. Fill in the File Attributes columns by issuing the command ls -l on each Actual Device Node from column 2.
    8. Install SDD following the instructions in the Installing the Subsystem Device Driver.
    9. Fill in the Subsystem Device Driver Device Links column by matching each cntndnsn device link listed in the Oracle Device Link column with its associated vpathN device link name by typing the following command:
      /opt/IBMdpo/bin/showvpath
      
    10. Fill in the Subsystem Device Driver Device Nodes column by issuing the command ls -l on each SDD Device Link and tracing back to the link source file.
    11. Change the attributes of each node listed in the Subsystem Device Driver Device Nodes column to match the attributes listed to the left of it in the File Attributes column using the UNIX chown, chgrp, and chmod commands.
    12. Make a copy of the existing /etc/fstab file. Edit the /etc/fstab file, changing each Oracle device link to its corresponding SDD device link.
    13. For each link found in an oradata directory, recreate the link using the appropriate SDD device link as the source file instead of the associated sdisk device link listed in the Oracle Device Link column.
    14. Restart the server.
    15. Verify that all file system and database consistency checks complete successfully.

    Chapter 6. SDD for a Sun host system

    This chapter provides step-by-step procedures for you to install, configure, remove, and use SDD on a Sun host system that is attached to an ESS. For updated and additional information not included in this manual, see the README file on the compact disc or visit the SDD Web site:

    www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates


    Verifying the hardware and software requirements

    You must install the following hardware and software components to ensure that SDD installs and operates successfully.

    Hardware

    Software

    Supported environments

    SDD supports the following environments:

    Non-supported environments

    SDD does not support the following environments:

    Understanding how SDD works on a Sun host system

    SDD resides above the Sun SCSI disk driver (sd) in the protocol stack. For more information about how SDD works, see Subsystem Device Driver.


    Preparing for SDD installation

    Before you install SDD, you must configure the ESS to your host system and attach the required fibre-channel adapters.

    Configuring the ESS

    Before you install SDD, configure your ESS for single-port or multiport access for each LUN. SDD requires a minimum of two independent paths that share the same logical unit to use the load-balancing and path-failover protection features.

    For information about configuring your ESS, see IBM Enterprise Storage Server Introduction and Planning Guide.

    Planning for installation

    Before you install SDD on your Sun host system, you need to understand what kind of software is running on it. The way you install SDD depends on the kind of software you are running. There are three types of software that communicate directly to raw or block disk device interfaces such as sd and SDD:

    There are three possible scenarios for installing SDD. The scenario you choose depends on the kind of software you have installed:

    Scenario 1
    Your system has no volume manager, DBMS, or software applications (other than UNIX) that communicate directly to the Solaris disk device layer

    Scenario 2
    Your system already has a volume manager, software application, or DBMS, such as Oracle, that communicates directly with the Solaris disk device drivers

    Scenario 3
    Your system already has SDD and you want to upgrade the software

    Table 23 further describes the various installation scenarios and how you should proceed.

    Table 23. SDD installation scenarios
    Installation scenario Description How to proceed
    Scenario 1
    • SDD is not installed.
    • No volume managers are installed.
    • No software application or DBMS are installed that communicates directly to the sd interface.

    Go to:
    1. Installing the Subsystem Device Driver
    2. Standard UNIX applications
    Scenario 2
    • SDD is not installed.
    • An existing volume manager, software application, or DBMS is installed that communicates directly to the sd interface.

    Go to:
    1. Installing the Subsystem Device Driver
    2. Using applications with SDD
    Scenario 3 SDD is installed. Go to: Upgrading the Subsystem Device Driver

    Table 24 lists the installation package file names that come with SDD.

    Table 24. SDD package file names
    Package file names Description
    sun32bit/IBMdpo Solaris 2.6, Solaris 7, Solaris 8
    sun64bit/IBMdpo Solaris 7
    sun64bit/IBMdpo Solaris 8

    For SDD to operate properly, ensure that the Solaris patches in Table 25 are installed on your operating system.

    Table 25. Solaris patches necessary for proper operation of SDD
    Patch name Solaris 2.6 Solaris 7
    glm 105580-15 106925-04
    isp 105600-19 106924-06
    sd & ssd 105356-16 107458-10

    Attention: Analyze and study your operating and application environment to ensure there are no conflicts with these patches prior to their installation.

    Go to the following Web site for the latest information about Solaris patches:

    sunsolve.Sun.COM


    Installing the Subsystem Device Driver

    Before you install SDD, make sure that you have root access to your Sun host system and that all the required hardware and software is ready.

    Perform the following steps to install SDD on your Sun host system:

    1. Make sure that the SDD compact disc (CD) is available.
    2. Insert the CD into your CD-ROM drive.
    3. Change to the installation directory:
      # cd /cdrom/cdrom0/sun32bit or
      # cd /cdrom/cdrom0/sun64bit 
      
    4. Issue the pkgadd command, and point the -d option of the pkgadd command to the directory containing IBMdpo. For example,
      pkgadd -d /cdrom/cdrom0/sun32bit IBMdpo or
      pkgadd -d /cdrom/cdrom0/sun64bit IBMdpo
      
    5. A message similar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |Processing package instance <IBMdpo> from <var/spool/pkg>                       |
      |                                                                                |
      |                                                                                |
      |IBM DPO driver                                                                  |
      |(sparc) 1                                                                       |
      |## Processing package information.                                              |
      |## Processing system information.                                               |
      |## Verifying disk space requirements.                                           |
      |## Checking for conflicts with packages already installed.                      |
      |## Checking for setuid/setgid programs.                                         |
      |                                                                                |
      |This package contains scripts which will be executed with super-user            |
      |permission during the process of installing this package.                       |
      |                                                                                |
      |Do you want to continue with the installation of <IBMdpo> [y,n,?]               |
      +--------------------------------------------------------------------------------+
    6. Type y and press Enter to proceed. A message similar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |Installing IBM DPO driver as <IBMdpo>                                           |
      |                                                                                |
      |## Installing part 1 of 1.                                                      |
      |/etc/defvpath                                                                   |
      |/etc/rc2.d/S00vpath-config                                                      |
      |/etc/rcS.d/S20vpath-config                                                      |
      |/kernel/drv/vpathdd                                                             |
      |/kernel/drv/vpathdd.conf                                                        |
      |/opt/IBMdpo/cfgvpath                                                            |
      |/opt/IBMdpo/datapath                                                            |
      |/opt/IBMdpo/devlink.vpath.tab                                                   |
      |/opt/IBMdpo/etc.system                                                          |
      |/opt/IBMdpo/pathtest                                                            |
      |/opt/IBMdpo/showvpath                                                           |
      |/usr/sbin/vpathmkdev                                                            |
      |[ verifying class <none>  ]                                                     |
      |## Executing postinstall script.                                                |
      |                                                                                |
      |DPO: Configuring 24 devices (3 disks * 8 slices)                                |
      |                                                                                |
      |Installation of <IBMdpo> was successful.                                        |
      |                                                                                |
      |The following packages are available:                                           |
      |1 IBMcli ibm2105cli                                                             |
      |         (sparc) 1.1.0.0                                                        |
      |2 IBMdpo IBM DPO driver Version: May-10-2000 16:51                              |
      |         (sparc) 1                                                              |
      |Select package(s) you wish to process (or 'all' to process                      |
      |all packages). (default: all) [?,??,q]:                                         |
      +--------------------------------------------------------------------------------+
    7. Type q and press Enter to proceed. A message similar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |*** IMPORTANT NOTICE ***                                                        |
      |This machine must now be rebooted in order to ensure                            |
      |sane operation. Execute                                                         |
      |       shutdown -y -i6 -g0                                                      |
      |and wait for the "Console Login:" prompt.                                       |
      |                                                                                |
      |DPO is now installed. Proceed to Post-Installation.                             |
      +--------------------------------------------------------------------------------+

    Verifying the SDD installation

    To verify that SDD has been successfully installed, type datapath query device. If the command executes, SDD is installed.

    Uninstalling the Subsystem Device Driver

    The following procedure explains how to uninstall an SDD. You must uninstall the current level of SDD before upgrading to a newer level.

    Attention: Do not restart between the uninstallation and the reinstallation of SDD.

    Perform the following steps to uninstall SDD:

    1. Restart or unmount all SDD file systems.
    2. If you are using SDD with a database, such as Oracle, edit the appropriate database configuration files (database partition) to remove all the SDD devices.
    3. If you are using a database, restart the database.
    4. Type # pkgrm IBMdpo and press Enter.

      Attention: A number of different installed packages is displayed. Make sure you specify the correct package to uninstall.

      A message similar to the following is displayed:

      +--------------------------------------------------------------------------------+
      |The following packages are available:                                           |
      |1 IBMcli ibm2105cli                                                             |
      |         (sparc) 1.1.0.0                                                        |
      |2 IBMdpo IBM DPO driver Version: May-10-2000 16:51                              |
      |         (sparc) 1                                                              |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    5. Type y and press Enter. A message similar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |## Removing installed package instance <IBMdpo>                                 |
      |                                                                                |
      |This package contains scripts that will be executed with super-user             |
      |permission during the process of removing this package.                         |
      |                                                                                |
      |Do you want to continue with the removal of this package [y,n,?,q] y            |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    6. Type y and press Enter. A message similar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |## Verifying package dependencies.                                              |
      |## Processing package information.                                              |
      |## Executing preremove script.                                                  |
      |Device busy                                                                     |
      |Cannot unload module: vpathdd                                                   |
      |Will be unloaded upon reboot.                                                   |
      |## Removing pathnames in class <none>                                           |
      |/usr/sbin/vpathmkdev                                                            |
      |/opt/IBMdpo                                                                     |
      |/kernel/drv/vpathdd.conf                                                        |
      |/kernel/drv/vpathdd                                                             |
      |/etc/rcS.d/S20vpath-config                                                      |
      |/etc/rc2.d/S00vpath-config                                                      |
      |/etc/defvpath                                                                   |
      |## Updating system information.                                                 |
      |                                                                                |
      |Removal of <IBMdpo> was successful.                                             |
      |                                                                                |
      +--------------------------------------------------------------------------------+

      Attention: Do not restart at this time.

    When SDD has been successfully uninstalled, the first part of the procedure for upgrading the SDD is complete. To complete the upgrade, you now need to reinstall SDD. See Installing the Subsystem Device Driver for detailed procedures.

    Post-installation

    After the installation is complete, manually unmount the compact disc. Issue the umount /cdrom command from the root directory. Go to the CD-ROM drive and press the Eject button.

    After SDD is installed, your system must be restarted to ensure proper operation. Type the command:

    # shutdown -i6 -g0 -y
    

    SDD devices are found in the /dev/rdsk and /dev/dsk directories. The device is named according to the SDD instance number. A device with an instance number of 0 would be: /dev/rdsk/vpath0a where a denotes the slice. Therefore, /dev/rdsk/vpath0c would be instance zero and slice 2.

    After SDD is installed, the device driver resides above the Sun SCSI disk driver (sd) in the protocol stack. In other words, SDD now communicates to the Solaris device layer. The SDD software installation procedure installs a number of SDD components and updates some system files. Those components and files are listed in the following tables.

    Table 26. SDD components installed for Sun host systems
    File Location Description
    vpathdd /kernel/drv Device driver
    vpathdd.conf /kernel/drv SDD config file
    Executables /opt/IBMdpo/bin Configuration and status tools
    S20vpath-config /etc/rcS.d Boot initialization script (See Note.)
    Note:
    This script must come before other LVM initialization scripts

    Table 27. System files updated for Sun host systems
    File Location Description
    /etc/system /etc Forces the loading of SDD
    /etc/devlink.tab /etc Tells the system how to name SDD devices in /dev

    Table 28. SDD commands and their descriptions for Sun host systems
    Command Description
    cfgvpath Configures vpath devices
    showvpath Lists all SDD devices and their underlying disks
    vpathmkdev Create SDD devices for /dev/dsk entries
    datapath SDD driver console command tool

    If you are not using a volume manager, software application, or DBMS that communicates directly to the sd interface, then the installation procedure is nearly complete. If you have a volume manager, software application, or DBMS installed that communicates directly to the sd interface, such as Oracle, go to Using applications with SDD and read the information specific to the application you are using.

    Upgrading the Subsystem Device Driver

    Upgrading SDD consists of uninstalling and reinstalling the IBMdpo package. If you are upgrading SDD, go to Uninstalling the Subsystem Device Driver and then go to Installing the Subsystem Device Driver.

    Changing an SDD hardware configuration

    When adding or removing multiport SCSI devices from your system, you must reconfigure SDD to recognize the new devices. Perform the following steps to reconfigure SDD:

    1. Shut down the system. Type shutdown -i0 -g0 -y and press Enter.
    2. Perform a configuration restart. From the OK prompt, type boot -r and press Enter. This uses the current SDD entries during restart, not the new entries. The restart forces the new disks to be recognized.
    3. Run the SDD configuration utility to make the changes to the directory /opt/IBMdpo/bin. Type cfgvpath -c and press Enter.
    4. Shut down the system. Type shutdown -i6 -g0 -y and press Enter.
    5. After the restart, change to the /opt/IBMdpo/bin directory by typing:
      cd /opt/IBMdpo/bin
      
    6. Type drvconfig and press Enter to reconfigure all the drives.
    7. Type vpathmkdev and press Enter to create all the vpath devices.

    Using applications with SDD

    If your system already has a volume manager, software application, or DBMS installed that communicates directly with the Solaris disk device drivers, you need to insert the new SDD device layer between the program and the Solaris disk device layer. You also need to customize the volume manager, software application, or DBMS in order to have it communicate with the SDD devices instead of the Solaris devices.

    In addition, many software applications and DBMS need to control certain device attributes such as ownership and permissions. Therefore, you must ensure that the new SDD devices that these software applications or DBMSs access have the same attributes as the Solaris sd devices that they replace. You need to customize the software application or DBMS to accomplish this.

    This section describes how to use the following applications with SDD:

    Standard UNIX applications

    If you have not already done so, install SDD using the procedure in the Installing the Subsystem Device Driver section. When this is done, the device driver resides above the Solaris SCSI disk driver (sd) in the protocol stack. In other words, SDD now communicates to the Solaris device layer.

    Standard UNIX applications, such as newfs, fsck, mkfs, and mount, that normally take a disk device or raw disk device as a parameter, also accept the SDD device as a parameter. Similarly, entries in files such as vfstab and dfstab (in the format of cntndnsn) can be replaced by entries for the corresponding SDD vpathNs devices. Make sure that the devices that are replaced are replaced with the corresponding SDD device. Issue the showvpath command to list all SDD devices and their underlying disks.

    Installing SDD on a Network File System file server

    The procedures in this section show how to install SDD for use with an exported file system (Network File System file server).

    Setting up the Network File System for the first time

    Perform the following steps if you are installing exported file systems on SDD devices for the first time:

    1. If you have not already done so, install SDD using the procedure in the Installing the Subsystem Device Driver section.
    2. Determine which SDD (vpathN) volumes you will use as file system devices.
    3. Partition the selected volumes using the Solaris format utility.
    4. Create file systems on the selected SDD devices using the appropriate utilities for the type of file system that you will use. If you are using the standard Solaris UFS file system, type the following command:
      # newfs /dev/rdsk/vpathNs
      

      In this example, N is the SDD device instance of the selected volume. Create mount points for the new file systems.

    5. Install the file systems into the /etc/fstab directory. Click yes in the mount at boot field.
    6. Install the file system mount points into the directory /etc/exports for export.
    7. Restart the system.

    Installing SDD on a system that already has the Network File System file server

    Perform the following steps if you have the Network File System file server already configured to:

    1. List the mount points for all currently exported file systems by looking in the /etc/exports directory.
    2. Match the mount points found in step 1 with sdisk device link names (files named /dev/(r)dsk/cntndn) by looking in the /etc/fstab directory.
    3. Match the sd device link names found in step 2 with SDD device link names (files named /dev/(r)dsk/vpathN) by issuing the showvpath command.
    4. Make a backup copy of the current /etc/fstab file.
    5. Edit the /etc/fstab file, replacing each instance of an sd device link named /dev/(r)dsk/cntndn with the corresponding SDD device link.
    6. Restart the system
    7. Verify that each exported file system passes:

    Oracle

    You must have super-user privileges, to perform the following procedures. You also need to have Oracle documentation on hand. These procedures were tested with Oracle 8.0.5 Enterprise server with the 8.0.5.1 patch set from Oracle.

    Installing an Oracle database for the first time

    You can set up your Oracle database in one of two ways. You can set it up to use a file system or raw partitions. The procedure for installing your database differs depending on the choice you make.

    Using a file system

    1. If you have not already done so, install SDD using the procedure in the Installing the Subsystem Device Driver section.
    2. Create and mount file systems on one or more SDD partitions. (Oracle recommends three mount points on different physical devices.)
    3. Follow the Oracle Installation Guide for instructions on installing to a file system. (During the Oracle installation, you will be asked to name three mount points. Supply the mount points for the file systems you created on the SDD partitions.)

    Using raw partitions

    Attention: If using raw partitions make sure all the databases are closed before going further. Make sure that the ownership and permissions of the SDD devices are the same as the ownership and permissions of the raw devices they are replacing. Do not use disk cylinder 0 (sector 0), which is the disk label. Using it corrupts the disk. For example, slice 2 on Sun is the whole disk. If you use this device without repartitioning it to start at sector 1, the disk label is corrupted.

    In the following procedure you will be replacing the raw devices with the SDD devices.

    1. If you have not already done so, install SDD using the procedure outlined in the Installing the Subsystem Device Driver section.
    2. Create the Oracle software owner user in the local server /etc/passwd file. You must also complete the following related activities:
      1. Complete the rest of the Oracle preinstallation tasks described in the Oracle8 Installation Guide. Plan to install Oracle8 on a file system that resides on an SDD partition.
      2. Set up the Oracle user's ORACLE_BASE and ORACLE_ HOME environment variables to be directories of this file system.
      3. Create two more SDD-resident file systems on two other SDD volumes. Each of the resulting three mount points should have a subdirectory named oradata. The subdirectory is used as a control file and redo log location for the installer's default database (a sample database) as described in the Installation Guide. Oracle recommends using raw partitions for redo logs. To use SDD raw partitions as redo logs, create symbolic links from the three redo log locations to SDD raw device links (files named /dev/rdsk/vpathNs, where N is the SDD instance number, and s is the partition ID) that point to the slice.
    3. Determine which SDD (vpathN) volumes you will use as Oracle8 database devices.
    4. Partition the selected volumes using the Solaris format utility. If SDD raw partitions are to be used by Oracle8 as database devices, be sure to leave sector 0/disk cylinder 0 of the associated volume unused. This protects UNIX disk labels from corruption by Oracle8.
    5. Ensure that the Oracle software owner has read and write privileges to the selected SDD raw partition device files under the /devices/pseudo directory.
    6. Set up symbolic links in the oradata directory under the first of the three mount points. See step 2. Link the database files to SDD raw device links (files named /dev/rdsk/vpathNs) that point to partitions of the appropriate size.
    7. Install the Oracle8 server following the instructions in the Oracle Installation Guide. Be sure to be logged in as the Oracle software owner when you run the orainst /m command. Select the Install New Product - Create Database Objects option. Select Raw Devices for the storage type. Specify the raw device links set up in step 2 for the redo logs. Specify the raw device links set up in step 3 for the database files of the default database.
    8. To set up other Oracle8 databases, you must set up control files, redo logs, and database files following the guidelines in the Oracle8 Administrator's Reference. Make sure any raw devices and file systems you set up reside on SDD volumes.
    9. Launch the sqlplus utility.
    10. Issue the create database SQL command, specifying the control, log, and system data files that you have set up.
    11. Issue the create tablespace SQL command to set up each of the temp, rbs, tools, and users database files that you created.
    12. Issue the create rollback segment SQL command to create the three redo log files that you set. For the syntax of these three create commands, see the Oracle8 Server SQL Language Reference Manual.

    Installing an SDD on a system that already has Oracle in place

    The installation procedure for a new SDD installation differs depending on whether you are using a file system or raw partitions for your Oracle database.

    If using a file system

    Perform the following procedure if you are installing SDD for the first time on a system with an Oracle database that uses a file system:

    1. Record the raw disk partitions being used (they are in the cntndnsn format) or the partitions where the Oracle file systems reside. You can get this information from the /etc/vfstab file if you know where the Oracle files are. Your database administrator can tell you where the Oracle files are, or you can check for directories with the name oradata.
    2. Complete the basic installation steps in the Installing the Subsystem Device Driver section.
    3. Change to the directory where you installed the SDD utilities. Issue the showvpath command.
    4. Check the display to see whether you find a cntndn directory that is the same as the one where the Oracle files are. For example, if the Oracle files are on c1t8d0s4, look for c1t8d0s2. If you find it, you will know that /dev/dsk/vpath0c is the same as /dev/dsk/clt8d2s2. (SDD partition identifiers end in an alphabetical character from a -g rather than s0, s1, s2, etc.) Write this down. A message similiar to the following is displayed:
      +--------------------------------------------------------------------------------+
      |vpath0c                                                                         |
      |    c1t8d0s2   /devices/pci@1f,0/pci@1/scsi@2/sd@1,0:c,raw                      |
      |    c2t8d0s2   /devices/pci@1f,0/pci@1/scsi@2,1/sd@1,0:c,raw                    |
      |                                                                                |
      +--------------------------------------------------------------------------------+
    5. Use the SDD partition identifiers instead of the original Solaris identifiers when mounting the file systems.

      If you originally used the following Sun identifiers:
      mount /dev/dsk/c1t3d2s4 /oracle/mp1

      you now use the following SDD partition identifiers:
      mount /dev/dsk/vpath2e /oracle/mp1

      For example, assume you found that vpath2c was the SDD identifier.

    Follow the instructions in Oracle Installation Guide for setting ownership and permissions.

    If using raw partitions

    Perform the following procedure if you have Oracle8 already installed and want to reconfigure it to use SDD partitions instead of sd partitions (for example, partitions accessed through /dev/rdsk/cntndn files).

    All Oracle8 control, log, and data files are accessed either directly from mounted file systems, or through links from the oradata subdirectory of each Oracle mount point set up on the server. Therefore, the process of converting an Oracle installation from sd to SDD is changing the Oracle mount points' physical devices in /etc/fstab from sd device partition links to the SDD device partition links that access the same physical partitions, and then recreating any links to raw sd device links to point to raw SDD device links that access the same physical partitions.

    Converting an Oracle installation from sd to SDD partitions

    Perform the following steps to convert an Oracle installation from sd to SDD partitions:

    1. Back up your Oracle8 database files, control files, and redo logs.
    2. Obtain the sd device names for the Oracle8 mounted file systems by looking up the Oracle8 mount points in /etc/vfstab and extracting the corresponding sd device link name (for example, /dev/rdsk/c1t4d0s4).
    3. Launch the sqlplus utility.
    4. Type the command:
      select * from sys.dba_data_files;
      

      The output lists the locations of all data files in use by Oracle. Determine the underlying device that each data file resides on, either by looking up mounted file systems in the /etc/vfstab file or by extracting raw device link names directly from the select command output.

    5. Issue the ls -l command on each device link found in step 4 and extract the link source device file name. For example, if you type the command:
      # ls -l /dev/rdsk/c1t1d0s4
      

      A message similar to the following is displayed:

      +--------------------------------------------------------------------------------+
      |/dev/rdsk/c1t1d0s4 /devices/pci@1f,0/pci@1/scsi@2/sd@1,0:e                      |
      +--------------------------------------------------------------------------------+
    6. Write down the file ownership and permissions by issuing the ls -lL command on either the files in /dev/ or /devices (it yields the same result). For example, if you type the command:
      # ls -lL /dev/rdsk/c1t1d0s4
      

      A message similar to the following is displayed:

      +--------------------------------------------------------------------------------+
      |crw-r--r-- oracle dba 32,252 Nov 16 11:49 /dev/rdsk/c1t1d0s4                    |
      +--------------------------------------------------------------------------------+
    7. Complete the basic installation steps in the Installing the Subsystem Device Driver section.
    8. Match each cntndns device with its associated vpathNs device link name by issuing the showvpath command. Remember that vpathNs partition names use the letters a - h in the s position to indicate slices 0 - 7 in the corresponding cntndnsn slice names.
    9. Issue the ls -l command on each SDD device link.
    10. Write down the SDD device nodes for each SDD device link by tracing back to the link source file.
    11. Change the attributes of each SDD device to match the attributes of the corresponding disk device using the chgrp and chmod commands.
    12. Make a copy of the existing /etc/vfstab file for recovery purposes. Edit the /etc/vfstab file, changing each Oracle device link to its corresponding SDD device link.
    13. For each link found in an oradata directory, recreate the link using the appropriate SDD device link as the source file instead of the associated sd device link. As you perform this step, generate a reversing shell script that can restore all the original links in case of error.
    14. Restart the server.
    15. Verify that all file system and database consistency checks complete successfully.

    Solstice DiskSuite

    For these procedures, you need access to the Solaris answerbook facility. These procedures were tested using Solstice DiskSuite 4.2 with the patch 106627-04 (DiskSuite patch) installed. You should have a copy of the DiskSuite Administration Guide available to complete these procedures. You must have super-user privileges to perform these procedures.

    Installing Solstice DiskSuite for the first time

    Perform the following steps if you are installing Solstice DiskSuite on the multiport subsystem server for the first time. Perform the following steps to install the Solstice DiskSuite for the first time on an SDD system:

    1. Install SDD using the procedure in the Installing the Subsystem Device Driver section, if you have not already done so.
    2. Configure the Sparc server to recognize all devices over all paths using the boot -r command.
    3. Install the Solstice DiskSuite packages and the answerbook. Do not restart yet.
      Note:
      Do not install the DiskSuite Tool (metatool).
    4. Determine which vpath devices you will use to create Disk Suite metadevices. Partition these devices by selecting them in the Solaris format utility. The devices appear as vpathNs, where N is the vpath driver instance number. Use the partition submenu, just as you would for an sd device link of the form, cntndn. If you want to know which cntndn links correspond to a particular vpath device, type the showvpath command and press Enter. Reserve at least three partitions of three cylinders each for use as DiskSuite Replica database locations.
      Note:
      You do not need to partition any sd (cntndn) devices.
    5. Set up the replica databases on a partition of its own. This partition needs to be at least three partitions of three cylinders. Do not use a partition that includes Sector 0 for this database replica partition. Perform the following instructions for setting up replica databases onthe vpathNs partitions, where N is the vpath device instance number and s is the letter denoting the three-cylinder partition, or slice, of the device that you wish to use as a replica. Remember that partitions a - h of a vpath device correspond to slices 0 - 7 of the underlying multiport subsystem device.
    6. Follow the instructions in the DiskSuite Administration Guide to build the types of metadevices you need, using the metainit command and the /dev/(r)dsk/vpathNs device link names, wherever the instructions specify /dev/(r)dsk/cntndnsn device link names.
    7. Insert the setup of all vpathNs devices used by DiskSuite into the /etc/opt/SUNWmd/md.tab file.

    Installing SDD on a system that already has Solaris DiskSuite in place

    Perform the following steps if Solaris DiskSuite is already installed:

    1. Back up all data.
    2. Back up the current Solstice configuration by making a copy of the /etc/opt/SUNWmd/md.tab file, and recording the output of the metastat and metadb -i commands. Make sure all sd device links in use by DiskSuite are entered in md.tab, and that they all come up properly after a restart.
    3. Install SDD using the procedure in the Installing the Subsystem Device Driver section, if you have not already done so. After the installation completes, type shutdown -i6 -y -g0 and press Enter. This verifies the vpath installation.
      Note:
      Do not do a reconfiguration restart.
    4. Using a plain sheet of paper, make a two-column list matching up the /dev/(r)dsk/cntndnsn device links found in step 2 with the corresponding /dev/(r)dsk/vpathNs device links using the showvpath command.
    5. Delete each replica database currently configured with a /dev/(r)dsk/cntndnsn device by using the metadb -d -f <device> command. Replace the replica database with the corresponding /dev/(r)dsk/vpathNs device found in step 2 by using the metadb -a <device> command.
    6. Create a new md.tab file. Insert the corresponding vpathNs device link name in place of each cntndnsn device link name. Do not do this for start device partitions (vpath does not currently support these). When you are confident that the new file is correct, install it in the /etc/opt/SUNWmd directory.
    7. Restart the server, or proceed to the next step if you wish to avoid restarting your system.

      To back out vpath in case of any problems following step 7, reverse the procedures in step 6, reinstall the original md.tab into /etc/opt/SUNWmd, issue the pkgrm IBMdpo command, and restart.

    8. Stop all applications using DiskSuite, including file systems.
    9. Type the following commands for each existing metadevice:
      metaclear <device>
      metainit -a
      
    10. Restart your applications.

    Setting up UFS logging on a new system

    For these procedures, you need access to the Solaris answerbook facility. You must have super-user privileges to perform these procedures.

    Perform the following steps if you are installing a new UFS logging file system on vpath devices:

    1. Install SDD using the procedure in the Installing the Subsystem Device Driver section, if you have not already done so.
    2. Determine which vpath (vpathNs) volumes you will use as file system devices. Partition the selected vpath volumes using the Solaris format utility. Be sure to create partitions for UFS logging devices as well as for UFS master devices.
    3. Create file systems on the selected vpath UFS master device partitions using the newfs command.
    4. Install Solstice DiskSuite if you have not already done so.
    5. Create the metatrans device using metainit. For example, assume /dev/dsk/vpath0d is your UFS master device used in step 3, /dev/dsk/vpath0e is its corresponding log device, and d0 is the trans device you want to create for UFS logging. Type metainit d0 -t vpath0d vpath0e and press Enter.
    6. Create mount points for each UFS logging file system you have created using steps 3 and 5.
    7. Install the file systems into the /etc/vfstab directory, specifying /dev/md/(r)dsk/d <metadevice number> for the raw and block devices. Set the mount at boot field to yes.
    8. Restart your system.

    Installing vpath on a system that already has UFS logging in place

    Perform the following steps if you already have UFS logging file systems residing on a multiport subsystem and if you wish to use vpath partitions instead of sd partitions to access them.

    1. Make a list of the DiskSuite metatrans devices for all existing UFS logging file systems by looking in the /etc/vfstab directory. Make sure that all configured metatrans devices are correctly set up in the /etc/opt/SUNWmd/md.tab file. If the devices are not set up now, set them up before continuing. Save a copy of the md tab file.
    2. Match the device names found in step 1 with sd device link names (files named /dev/(r)dsk/cntndnsn) using the metastat command.
    3. Install SDD using the procedure in the Installing the Subsystem Device Driver section, if you have not already done so.
    4. Match the sd device link names found in step 2 with vpath device link names (files named /dev/(r)dsk/vpathNs) by executing the /opt/IBMdpo/bin/showvpath command.
    5. Unmount all current UFS logging file systems known to reside on the multiport subsystem through the umount command.
    6. Type metaclear -a and press Enter.
    7. Create new metatrans devices from the vpathNs partitions found in step 4 corresponding to the sd device links found in step 2. Remember that vpath partitions a - h correspond to sd slices 0 - 7. Use the metainit d <metadevice number> -t <"vpathNs" - master device> <"vpathNs" - logging device> command. Be sure to use the same metadevice numbering as was originally used with the sd partitions. Edit the /etc/opt/SUNWmd/md.tab file to change each metatrans device entry to use vpathNs devices.
    8. Restart the system.
    Note:
    If there is a problem with a metatrans device after steps 7 and 8, restore the original /etc/opt/SUNWmd/md.tab file and restart the system. Review your steps and try again.

    Chapter 7. Using the datapath commands

    SDD provides commands that you can use to display the status of adapters that are used to access managed devices, or to display the status of devices that the device driver manages. You can also set individual path conditions either to online or offline, or you can set all paths that are connected to an adapter or bus either to online or offline. This chapter includes descriptions of these commands. Table 29 provides an alphabetical list of these commands, a brief description, and where to go in this chapter for more information.

    Table 29. Commands
    Command Description Page
    datapath query adapter Displays information about adapters datapath query adapter command
    datapath query adaptstats Displays performance information for all SCSI and FCS adapters that are attached to SDD devices datapath query adaptstats command
    datapath query device Displays information about devices datapath query device command
    datapath query devstats Displays performance information for a single SDD device or all SDD devices datapath query devstats command
    datapath set adapter Sets all device paths that are attached to an adapter to online or offline datapath set adapter command
    datapath set device Sets the path of a device to online or offline datapath set device command

    datapath query adapter command

    The datapath query adapter command displays information about a single adapter or all adapters.

    Syntax

    >>-datapath query adapter-adapter number-----------------------><
     
     
    

    Parameters

    adapter number
    The adapter number for which you want information displayed. If you do not enter an adapter number, information about all adapters is displayed.

    Examples

    If you type the datapath query adapter command, the following output is displayed:

    +--------------------------------------------------------------------------------+
    |Active Adapters :4                                                              |
    |                                                                                |
    |Adpt#     Adapter Name   State     Mode     Select     Errors  Paths  Active    |
    |    0            scsi3  NORMAL   ACTIVE  129062051          0     64       0    |
    |    1            scsi2  NORMAL   ACTIVE   88765386        303     64       0    |
    |    2           fscsi2  NORMAL   ACTIVE  407075697       5427   1024       0    |
    |    3           fscsi0  NORMAL   ACTIVE  341204788      63835    256       0    |
    +--------------------------------------------------------------------------------+

    The terms used in the output are defined as follows:

    Adpt #
    The number of the adapter.

    Adapter Name
    The name of the adapter.

    State
    The condition of the named adapter. It can be either:
    Normal
    Adapter is in use.
    Degraded
    One or more paths are not functioning.
    Failed
    The adapter is no longer being used by SDD.

    Mode
    The mode of the named adapter, which is either Active or Offline.

    Select
    The number of times this adapter was selected for input or output.

    Errors
    The number of errors on all paths that are attached to this adapter.

    Paths
    The number of paths that are attached to this adapter.
    Note:
    In the Windows NT host system, this is the number of physical and logical devices that are attached to this adapter.

    Active
    The number of functional paths that are attached to this adapter. The number of functional paths is equal to the number of paths minus any that are identified as failed or offline.

    datapath query adaptstats command

    The datapath query adaptstats command displays performance information for all SCSI and FCS adapters that are attached to SDD devices. If you do not enter a device number, information about all devices is displayed.

    Syntax

    >>-datapath query adaptstats-adapter number--------------------><
     
     
    

    Parameters

    adapter number
    The adapter number for which you want information displayed. If you do not enter an adapter number, information about all adapters is displayed.

    Examples

    If you type the datapath query adaptstats 0 command, the following output is displayed:

    Adapter #:  0
    =============
                    Total Read  Total Write  Active Read  Active Write   Maximum
    I/O:                  1442     41295166            0             2        75
    SECTOR:             156209    750217654            0            32      2098
     
    /*-------------------------------------------------------------------------*/                                                       
     
    

    The terms used in the output are defined as follows:

    Total Read

    Total Write

    Active Read

    Active Write

    Maximum

    datapath query device command

    The datapath query device command displays information about a single device or all devices. If you do not enter a device number, information about all devices is displayed.

    Syntax

    >>-datapath query device-device number-------------------------><
     
     
    

    Parameters

    device number
    The device number refers to the device index number, rather than the SDD device number.

    Examples

    If you type the datapath query device 35 command, the following output is displayed:

    DEV#:  35  DEVICE NAME: vpath0  TYPE: 2105E20   SERIAL: 60012028
    ================================================================
    Path#      Adapter/Hard Disk   State      Mode    Select        Errors
     0             scsi6/hdisk58    OPEN    NORMAL   7861147             0
     1             scsi5/hdisk36    OPEN    NORMAL   7762671             0                                                       
     
    
    Note:
    Usually, the device number and the device index number are the same. However, if the devices are configured out of order, the two numbers are not always consistent. To find the corresponding index number for a specific device, you should always run the datapath query device command first.

    The terms used in the output are defined as follows:

    Dev#
    The number of this device.

    Name
    The name of this device.

    Type
    The device product ID from inquiry data.

    Serial
    The logical unit number (LUN) for this device.

    Path
    The path number.

    Adapter
    The name of the adapter that the path is attached to.

    Hard Disk
    The name of the logical device that the path is bound to.

    State
    The condition of the named device:
    Open
    Path is in use.
    Close
    Path is not being used.
    Dead
    Path is no longer being used. It was either removed by SDD due to errors or manually removed using the datapath set device M path N offline or datapath set adapter N offline command.
    Invalid
    Path verification failed. The path was not opened.

    Mode
    The mode of the named device. It is either Normal or Offline.

    Select
    The number of times this path was selected for input or output.

    Errors
    The number of errors on a path that is attached to this device.

    datapath query devstats command

    The datapath query devstats command displays performance information for a single SDD device or all SDD devices. If you do not enter a device number, information about all devices is displayed.

    Syntax

    >>-datapath query devstats-device number-----------------------><
     
     
    

    Parameters

    device number
    The device number refers to the device index number, rather than the SDD device number.

    Examples

    If you type the datapath query devstats 0 command, the following output is displayed:

    Device #:   0
    =============
                    Total Read  Total Write  Active Read  Active Write   Maximum
    I/O:                   387     24502563            0             0        62
    SECTOR:               9738    448308668            0             0      2098
     
    Transfer Size:      <= 512        <= 4k       <= 16K        <= 64K     > 64K
                       4355850      1024164     19121140          1665       130
     
    /*-------------------------------------------------------------------------*/                                                       
     
    

    The terms used in the output are defined as follows:

    Total Read

    Total Write

    Active Read

    Active Write

    Maximum

    Transfer size

    datapath set adapter command

    The datapath set adapter command sets all device paths attached to an adapter either to online or offline. If all paths are attached to a single fibre-channel adapter, that connects to multiple ESS ports through a switch, the set adapter 0 offline command fails; all the paths are not set offline. The datapath set adapter offline command fails if there is any device having the last path attached to this adapter.

    Attention: The datapath set adapter offline command will not remove the last path to a device. This command can be issued even when the device are closed.

    Syntax

    >>-datapath set adapter-adapter number-+- online--+------------><
                                           '- offline-'
     
     
    

    Parameters

    adapter number
    The adapter number that you want to change.

    online
    Sets the adapter online.

    offline
    Sets the adapter offline.

    Examples

    If you type the datapath set adapter 0 offline command, adapter 0 changes to Offline mode and its state changes to failed; while all paths attached to adapter 0 change to Offline mode and their states change to Dead, if they were in the Open state.

    datapath set device command

    The datapath set device command sets the path of a device either to online or offline. You cannot remove the last path to a device from service. This prevents a data access failure from occurring. The datapath set device command can be issued even when the device is closed.

    Syntax

    >>-datapath set device-device number path number-+- online--+--><
                                                     '- offline-'
     
     
    

    Parameters

    device number
    The device index number that you want to change.

    path number
    The path number that you want to change.

    online
    Sets the path online.

    offline
    Removes the path from service.

    Examples

    If you type the datapath set device 0 path 0 offline command, path 0 for device 0 changes to Offline mode.


    Notices

    This information was developed for products and services offered in the U.S.A.

    IBM may not offer the products, services, or features discussed in this document in other countries. Consult your local IBM representative for information on the products and services currently available in your area. Any reference to an IBM product, program, or service is not intended to state or imply that only that IBM product, program, or service may be used. Any functionally equivalent product, program, or service that does not infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to evaluate and verify the operation of any non-IBM product, program, or service.

    IBM may have patents or pending patent applications covering subject matter described in this document. The furnishing of this document does not give you any license to these patents. You can send license inquiries, in writing, to:
    IBM Director of Licensing
    IBM Corporation
    North Castle Drive
    Armonk, NY 10504-1785
    U.S.A

    The following paragraph does not apply to the United Kingdom or any other country where such provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS PUBLICATIONS "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of express or implied warranties in certain transactions, therefore, this statement may not apply to you.

    This information could include technical inaccuracies or typographical errors. Changes are periodically made to the information herein; these changes will be incorporated in new editions of the publications. IBM may make improvements and/or changes in the product(s) and/or program(s) described in this publication at any time without notice.

    IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring any obligation to you.

    Information concerning non-IBM products was obtained from the suppliers of those products, their published announcements or other publicly available sources. IBM has not tested those products and cannot confirm the accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products.


    Trademarks

    The following terms are trademarks of the International Business Machines Corporation in the United States, other countries, or both:

    AIX
    AS/400
    DFSMS/MVS
    ES/9000
    ESCON
    FICON
    FlashCopy
    HACMP/6000
    IBM
    Enterprise Storage Server
    IBM TotalStorage
    eServer
    MVS/ESA
    Netfinity
    NetVista
    NUMA-Q
    Operating System/400
    OS/390
    OS/400
    RS/6000
    S/390
    Seascape
    SNAPSHOT
    SP
    StorWatch
    System/360
    System/370
    System/390
    TotalStorage
    Versatile Storage Server
    VM/ESA
    VSE/ESA

    Microsoft and Windows NT are trademarks of Microsoft Corporation in the United States, other countries, or both.

    Java and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other countries, or both.

    Linux is a trademark of Linus Torvalds and others.

    UNIX is a registered trademark of The Open Group in the United States and other countries.

    Other company, product, and service names may be trademarks or service marks of others.


    Electronic emission notices

    This section contains the electronic emission notices or statements for the United States and other countries.

    Federal Communications Commission (FCC) statement

    This equipment has been tested and complies with the limits for a Class A digital device, pursuant to Part 15 of the FCC Rules. These limits are designed to provide reasonable protection against harmful interference when the equipment is operated in a commercial environment. This equipment generates, uses, and can radiate radio frequency energy and, if not installed and used in accordance with the instruction manual, might cause harmful interference to radio communications. Operation of this equipment in a residential area is likely to cause harmful interference, in which case the user will be required to correct the interference at his own expense.

    Properly shielded and grounded cables and connectors must be used to meet FCC emission limits. IBM is not responsible for any radio or television interference caused by using other than recommended cables and connectors, or by unauthorized changes or modifications to this equipment. Unauthorized changes or modifications could void the users authority to operate the equipment.

    This device complies with Part 15 of the FCC Rules. Operation is subject to the following two conditions: (1) this device might not cause harmful interference, and (2) this device must accept any interference received, including interference that might cause undesired operation.

    Industry Canada compliance statement

    This Class A digital apparatus complies with Canadian ICES-003.

    Cet appareil numérique de la classe A est conform à la norme NMB-003 du Canada.

    European community compliance statement

    This product is in conformity with the protection requirements of EC Council Directive 89/336/EEC on the approximation of the laws of the Member States relating to electromagnetic compatibility. IBM cannot accept responsibility for any failure to satisfy the protection requirements resulting from a nonrecommended modification of the product, including the fitting of non-IBM option cards.

    Germany only

    Zulassungsbescheinigung laut Gesetz ueber die elektromagnetische Vertraeglichkeit von Geraeten (EMVG) vom 30. August 1995.

    Dieses Geraet ist berechtigt, in Uebereinstimmung mit dem deutschen EMVG das EG-Konformitaetszeichen - CE - zu fuehren.

    Der Aussteller der Konformitaetserklaeung ist die IBM Deutschland.

    Informationen in Hinsicht EMVG Paragraph 3 Abs. (2) 2:

      Das Geraet erfuellt die Schutzanforderungen nach EN 50082-1 und
      EN 55022 Klasse A.
    

    EN 55022 Klasse A Geraete beduerfen folgender Hinweise:

    Nach dem EMVG:

    "Geraete duerfen an Orten, fuer die sie nicht ausreichend entstoert
    sind, nur mit besonderer Genehmigung des Bundesministeriums
    fuer Post und Telekommunikation oder des Bundesamtes fuer Post und
    Telekommunikation
    betrieben werden. Die Genehmigung wird erteilt, wenn keine
    elektromagnetischen Stoerungen zu erwarten sind." (Auszug aus dem
    EMVG, Paragraph 3, Abs.4)
     
    Dieses Genehmigungsverfahren ist nach Paragraph 9 EMVG in Verbindung
    mit der entsprechenden Kostenverordnung (Amtsblatt 14/93)
    kostenpflichtig.
    

    Nach der EN 55022:

    "Dies ist eine Einrichtung der Klasse A. Diese Einrichtung kann im
    Wohnbereich Funkstoerungen verursachen; in diesem Fall kann vom
    Betreiber verlangt werden, angemessene Massnahmen durchzufuehren
    und dafuer aufzukommen."
    

    Anmerkung:

    Um die Einhaltung des EMVG sicherzustellen, sind die Geraete wie in den
    Handbuechern angegeben zu installieren und zu betreiben.
    

     

     

    Japanese Voluntary Control Council for Interference (VCCI) class A statement



    cmvc filename: vs2kan

    Korean government Ministry of Communication (MOC) statement

    Please note that this device has been approved for business purpose with regard to electromagnetic interference. If you find this is not suitable for your use, you may exchange it for a nonbusiness purpose one.

    Taiwan class A compliance statement



    cmvc filename: vs07171l


    IBM agreement for licensed internal code

    Read Before Using

    IMPORTANT

    YOU ACCEPT THE TERMS OF THIS IBM LICENSE AGREEMENT FOR MACHINE CODE BY YOUR USE OF THE HARDWARE PRODUCT OR MACHINE CODE. PLEASE READ THE AGREEMENT CONTAINED IN THIS BOOK BEFORE USING THE HARDWARE PRODUCT. SEE IBM agreement for licensed internal code.

    You accept the terms of this Agreement 1 by your initial use of a machine that contains IBM Licensed Internal Code (called "Code"). These terms apply to Code used by certain machines IBM or your reseller specifies (called "Specific Machines"). International Business Machines Corporation or one of its subsidiaries ("IBM") owns copyrights in Code or has the right to license Code. IBM or a third party owns all copies of Code, including all copies made from them.

    If you are the rightful possessor of a Specific Machine, IBM grants you a license to use the Code (or any replacement IBM provides) on, or in conjunction with, only the Specific Machine for which the Code is provided. IBM licenses the Code to only one rightful possessor at a time.

    Under each license, IBM authorizes you to do only the following:

    1. execute the Code to enable the Specific Machine to function according to its Official Published Specifications (called "Specifications");
    2. make a backup or archival copy of the Code (unless IBM makes one available for your use), provided you reproduce the copyright notice and any other legend of ownership on the copy. You may use the copy only to replace the original, when necessary; and
    3. execute and display the Code as necessary to maintain the Specific Machine.

    You agree to acquire any replacement for, or additional copy of, Code directly from IBM in accordance with IBM's standard policies and practices. You also agree to use that Code under these terms.

    You may transfer possession of the Code to another party only with the transfer of the Specific Machine. If you do so, you must 1) destroy all your copies of the Code that were not provided by IBM, 2) either give the other party all your IBM-provided copies of the Code or destroy them, and 3) notify the other party of these terms. IBM licenses the other party when it accepts these terms. These terms apply to all Code you acquire from any source.

    Your license terminates when you no longer rightfully possess the Specific Machine.

    Actions you must not take

    You agree to use the Code only as authorized above. You must not do, for example, any of the following:

    1. Otherwise copy, display, transfer, adapt, modify, or distribute the Code (electronically or otherwise), except as IBM may authorize in the Specific Machine's Specifications or in writing to you;
    2. Reverse assemble, reverse compile, or otherwise translate the Code unless expressly permitted by applicable law without the possibility of contractual waiver;
    3. Sublicense or assign the license for the Code; or
    4. Lease the Code or any copy of it.

    Glossary


    This glossary includes terms for the IBM TotalStorage Enterprise Storage Server (ESS) and other Seascape solution products.

    This glossary includes selected terms and definitions from:

    This glossary uses the following cross-reference form:

    See
    This refers the reader to one of three kinds of related information:

    A

    access
    (1) To obtain the use of a computer resource.
    (2) In computer security, a specific type of interaction between a subject and an object that results in flow of information from one to the other.

    access-any mode
    One of the two access modes that can be set for the ESS during initial configuration. It enables all fibre-channel-attached host systems with no defined access profile to access all logical volumes on the ESS. With a profile defined in ESS Specialist for a particular host, that host has access only to volumes that are assigned to the WWPN for that host. See pseudo-host and worldwide port name.

    active Copy Services server
    The Copy Services server that manages the Copy Services domain. Either the primary or the backup Copy Services server can be the active Copy Services server. The backup Copy Services server is available to become the active Copy Services server if the primary Copy Services server fails. See backup Copy Services server and primary Copy Services server.

    alert
    A message or log that a storage facility generates as the result of error event collection and analysis. An alert indicates that a service action is required.

    allegiance
    In Enterprise Systems Architecture/390, a relationship that is created between a device and one or more channel paths during the processing of certain conditions. See implicit allegiance, contingent allegiance, and reserved allegiance.

    allocated storage
    On an ESS, the space allocated to volumes, but not yet assigned. See assigned storage.

    American National Standards Institute (ANSI)
    An organization of producers, consumers, and general interest groups that establishes the procedures by which accredited organizations create and maintain voluntary industry standards in the United States. (A)

    Anonymous
    The label in ESS Specialist on an icon representing all connections using fibre-channel adapters between the ESS and hosts that are not completely defined to the ESS. See anonymous host, pseudo-host, and access-any mode.

    anonymous host
    Synonym for "pseudo-host" (in contrast to the Anonymous label that appears on some pseudo-host icons. See Anonymous and pseudo-host.

    ANSI
    See American National Standards Institute.

    APAR
    See authorized program analysis report.

    arbitrated loop
    For fibre-channel connections, a topology that enables the interconnection of a set of nodes. See point-to-point connection and switched fabric.

    array
    An ordered collection, or group, of physical devices (disk drive modules) that are used to define logical volumes or devices. More specifically, regarding the ESS, an array is a group of disks designated by the user to be managed by the RAID-5 technique. See redundant array of inexpensive disks.

    ASCII
    American Standard Code for Information Interchange. An ANSI standard (X3.4-1977) for assignment of 7-bit numeric codes (plus 1 bit for parity) to represent alphabetic and numeric characters and common symbols. Some organizations, including IBM, have used the parity bit to expand the basic code set.

    assigned storage
    On an ESS, the space allocated to a volume and assigned to a port.

    authorized program analysis report (APAR)
    A report of a problem caused by a suspected defect in a current, unaltered release of a program.

    availability
    The degree to which a system or resource is capable of performing its normal function. See data availability.

    B

    backup Copy Services server
    One of two Copy Services servers in a Copy Services domain. The other Copy Services server is the primary Copy Services server. The backup Copy Services server is available to become the active Copy Services server if the primary Copy Services server fails. A Copy Services server is software that runs in one of the two clusters of an ESS, and manages data-copy operations for that Copy Services server group. See primary Copy Services server and active Copy Services server.

    bay
    Physical space on an ESS used for installing SCSI, ESCON, and fibre-channel host adapter cards. The ESS has four bays, two in each cluster. See service boundary.

    bit
    (1) A binary digit.
    (2) The storage medium required to store a single binary digit.
    (3) Either of the digits 0 or 1 when used in the binary numeration system. (T) See byte.

    block
    A group of consecutive bytes used as the basic storage unit in fixed-block architecture (FBA). All blocks on the storage device are the same size (fixed size). See fixed-block architecture and data record.

    byte
    (1) A group of eight adjacent binary digits that represent one EBCDIC character.
    (2) The storage medium required to store eight bits. See bit.

    C

    cache
    A buffer storage that contains frequently accessed instructions and data, thereby reducing access time.

    cache fast write
    A form of the fast-write operation in which the subsystem writes the data directly to cache where it is available for later destaging.

    cache hit
    An event that occurs when a read operation is sent to the cluster, and the requested data is found in cache. The opposite of cache miss.

    cache memory
    Memory, typically volatile memory, that a subsystem uses to improve access times to instructions or data. The cache memory is typically smaller and faster than the primary memory or storage medium. In addition to residing in cache memory, the same data also resides on the storage devices in the storage facility.

    cache miss
    An event that occurs when a read operation is sent to the cluster, but the data is not found in cache. The opposite of cache hit.

    call home
    A communication link established between the ESS and a service provider. The ESS can use this link to place a call to IBM or to another service provider when it requires service. With access to the machine, service personnel can perform service tasks, such as viewing error logs and problem logs or initiating trace and dump retrievals. See heartbeat and remote technical assistance information network.

    cascading
    (1) Connecting network controllers to each other in a succession of levels, to concentrate many more lines than a single level permits.
    (2) In high-availability cluster multiprocessing (HACMP), cascading pertains to a cluster configuration in which the cluster node with the highest priority for a particular resource acquires the resource if the primary node fails. The cluster node relinquishes the resource to the primary node upon reintegration of the primary node into the cluster.

    catcher
    A server that service personnel use to collect and retain status data that an ESS sends to it.

    CCR
    See channel-command retry.

    CCW
    See channel command word.

    CD-ROM
    See compact disc, read-only memory.

    CEC
    See computer-electronic complex.

    channel
    In Enterprise Systems Architecture/390, the part of a channel subsystem that manages a single I/O interface between a channel subsystem and a set of control units.

    channel command retry (CCR)
    In Enterprise Systems Architecture/390, the protocol used between a channel and a control unit that enables the control unit to request that the channel reissue the current command.

    channel command word (CCW)
    In Enterprise Systems Architecture/390, a data structure that specifies an I/O operation to the channel subsystem.

    channel path
    In Enterprise Systems Architecture/390, the interconnection between a channel and its associated control units.

    channel subsystem
    In Enterprise Systems Architecture/390, the part of a host computer that manages I/O communication between the program and any attached control units.

    channel-subsystem image
    In Enterprise Systems Architecture/390, the logical functions that a system requires to perform the function of a channel subsystem. With ESCON multiple image facility (EMIF), one channel subsystem image exists in the channel subsystem for each logical partition (LPAR). Each image appears to be an independent channel subsystem program, but all images share a common set of hardware facilities.

    CKD
    See count key data.

    CLI
    See command-line interface.

    cluster
    (1) A partition in the ESS capable of performing all ESS functions. With two clusters in the ESS, any operational cluster can take over the processing of a failing cluster.
    (2) On an AIX platform, a group of nodes within a complex.

    cluster processor complex (CPC)
    The unit within a cluster that provides the management function for the storage server. It consists of cluster processors, cluster memory, and related logic.

    command-line interface (CLI)
    (1) An interface provided by an operating system that defines a set of commands and enables a user (or a script-like language) to issue these commands by typing text in response to the command prompt (for example, DOS commands, UNIX shell commands).
    (2) An optional ESS software that enables a user to issue commands to and retrieve information from the Copy Services server.

    compact disc, read-only memory (CD-ROM)
    High-capacity read-only memory in the form of an optically read compact disc.

    compression
    (1) The process of eliminating gaps, empty fields, redundancies, and unnecessary data to shorten the length of records or blocks.
    (2) Any encoding that reduces the number of bits used to represent a given message or record.

    computer-electronic complex (CEC)
    The set of hardware facilities associated with a host computer.

    Concurrent Copy
    A facility on a storage server that enables a program to make a backup of a data set while the logical volume remains available for subsequent processing. The data in the backup copy is frozen at the point in time that the server responds to the request.

    concurrent installation of licensed internal code
    Process of installing licensed internal code on an ESS while applications continue to run.

    concurrent maintenance
    Service that is performed on a unit while it is operational.

    concurrent media maintenance
    Service performed on a disk drive module (DDM) without losing access to the data.

    configure
    To define the logical and physical configuration of the input/output (I/O) subsystem through the user interface provided for this function on the storage facility.

    consistent copy
    A copy of a data entity (a logical volume, for example) that contains the contents of the entire data entity at a single instant in time.

    console
    A user interface to a server, such as can be provided by a personal computer. See IBM TotalStorage ESS Master Console.

    contingent allegiance
    In Enterprise Systems Architecture/390, a relationship that is created in a control unit between a device and a channel when unit-check status is accepted by the channel. The allegiance causes the control unit to guarantee access; the control unit does not present the busy status to the device. This enables the channel to retrieve sense data that is associated with the unit-check status on the channel path associated with the allegiance.

    control unit (CU)
    (1) A device that coordinates and controls the operation of one or more input/output devices, and synchronizes the operation of such devices with the operation of the system as a whole.
    (2) In Enterprise Systems Architecture/390, a storage server with ESCON, FICON, or OEMI interfaces. The control unit adapts a native device interface to an I/O interface supported by an ESA/390 host system. On an ESS, the control unit would be the parts of the storage server that support the attachment of emulated CKD devices over ESCON, FICON, or OEMI interfaces. See cluster.

    control-unit image
    In Enterprise Systems Architecture/390, a logical subsystem that is accessed through an ESCON or FICON I/O interface. One or more control-unit images exist in each control unit. Each image appears as an independent control unit, but all control-unit image share a common set of hardware facilities. The ESS can emulate 3990-3, TPF, 3990-6, or 2105 control units.

    control-unit initiated reconfiguration (CUIR)
    A software mechanism used by the ESS to request that an operating system verify that one or more subsystem resources can be taken off-line for service. The ESS can use this process to automatically vary channel paths offline and online to facilitate bay service or concurrent code installation. Depending on the operating system, support for this process may be model-dependent, may depend on the IBM Subsystem Device Driver, or may not exist.

    Coordinated Universal Time (UTC)
    The international standard of time that is kept by atomic clocks around the world.

    Copy Services client
    Software that runs on each ESS cluster in the Copy Services server group and that performs the following functions:

    Copy Services server group
    A collection of user-designated ESS clusters participating in Copy Services functions managed by a designated active Copy Services server. A Copy Services server group is also called a Copy Services domain.

    count field
    The first field of a count key data (CKD) record. This eight-byte field contains a four-byte track address (CCHH). It defines the cylinder and head that are associated with the track, and a one-byte record number (R) that identifies the record on the track. It defines a one-byte key length that specifies the length of the record's key field (0 means no key field). It defines a two-byte data length that specifies the length of the record's data field (0 means no data field). Only the end-of-file record has a data length of zero.

    count key data (CKD)
    In Enterprise Systems Architecture/390, a data-record format employing self-defining record formats in which each record is represented by up to three fields--a count area identifying the record and specifying its format, an optional key area that can be used to identify the data area contents; and an optional data area that typically would contain the user data for the record. For CKD records on the ESS, the logical volume size is defined in terms of the device emulation mode (3390 or 3380 track format). The count field is always 8 bytes long and contains the lengths of the key and data fields, the key field has a length of 0 to 255 bytes, and the data field has a length of 0 to 65 535 or the maximum that will fit on the track. Typically, customer data appears in the data field. The use of the key field is dependent on the software managing the storage. See data record.

    CPC
    See cluster processor complex.

    CRC
    See cyclic redundancy check.

    CU
    See control unit.

    CUIR
    See control-unit initiated reconfiguration.

    customer console
    See consoleand IBM TotalStorage ESS Master Console.

    CUT
    See Coordinated Universal Time.

    cyclic redundancy check (CRC)
    A redundancy check in which the check key is generated by a cyclic algorithm. (T)

    cylinder
    A unit of storage on a CKD device. A cylinder has a fixed number of tracks.

    D

    DA
    See device adapter and SSA adapter.

    daisy chain
    See serial connection.

    DASD
    See direct access storage device.

    DASD fast write (DFW)
    Caching of active write data by a storage server by journaling the data in nonvolatile storage, avoiding exposure to data loss.

    data availability
    The degree to which data is available when needed, typically measured as a percentage of time that the system would be capable of responding to any data request (for example, 99.999% available).

    data compression
    A technique or algorithm used to encode data such that the encoded result can be stored in less space than the original data. The original data can be recovered from the encoded result through a reverse technique or reverse algorithm. See compression.

    Data Facility Storage Management Subsystem
    An operating environment that helps automate and centralize the management of storage. To manage storage, DFSMS provides the storage administrator with control over data class, storage class, management class, storage group, and automatic class selection routine definitions.

    data field
    The optional third field of a count key data (CKD) record. The count field specifies the length of the data field. The data field contains data that the program writes.

    data record
    The basic unit of S/390 and zSeries storage on an ESS, also known as a count-key-data (CKD) record. Data records are stored on a track. The records are sequentially numbered starting with 0. The first record, R0, is typically called the track descriptor record and contains data normally used by the operating system to manage the track. See count-key-data and fixed-block architecture.

    data sharing
    The ability of homogeneous or divergent host systems to concurrently utilize data that they store on one or more storage devices. The storage facility enables configured storage to be accessible to any, or all, attached host systems. To use this capability, the host program must be designed to support data that it is sharing.

    DDM
    See disk drive module.

    DDM group
    See disk drive module group.

    dedicated storage
    Storage within a storage facility that is configured such that a single host system has exclusive access to the storage.

    demote
    To remove a logical data unit from cache memory. A subsystem demotes a data unit in order to make room for other logical data units in the cache. It might also demote a data unit because the logical data unit is not valid. A subsystem must destage logical data units with active write units before they can be demoted.

    destaging
    (1) Movement of data from an online or higher priority to an offline or lower
    (2) priority device. The ESS stages incoming data into cache and then destages it to disk.

    device
    In Enterprise Systems Architecture/390, a disk drive.

    device adapter (DA)
    A physical component of the ESS that provides communication between the clusters and the storage devices. The ESS has eight device adapters that it deploys in pairs, one from each cluster. DA pairing enables the ESS to access any disk drive from either of two paths, providing fault tolerance and enhanced availability.

    device address
    In Enterprise Systems Architecture/390, the field of an ESCON or FICON device-level frame that selects a specific device on a control-unit image.

    device interface card
    A physical subunit of a storage cluster that provides the communication with the attached DDMs.

    device number
    In Enterprise Systems Architecture/390, a four-hexadecimal-character identifier, for example 13A0, that the systems administrator associates with a device to facilitate communication between the program and the host operator. The device number is associated with a subchannel.

    device sparing
    A subsystem function that automatically copies data from a failing DDM to a spare DDM. The subsystem maintains data access during the process.

    direct access storage device (DASD)
    (1) A mass storage medium on which a computer stores data.
    (2) A disk device.

    disk drive
    Standard term for a disk-based nonvolatile storage medium. The ESS uses hard disk drives as the primary nonvolatile storage media to store host data.

    disk drive module (DDM)
    A field replaceable unit that consists of a single disk drive and its associated packaging.

    disk drive module group
    In the ESS, a group of eight disk drive modules (DDMs) contained in an 8-pack and installed as a unit.

    disk group
    In the ESS, a collection of seven or eight disk drives in the same SSA loop and set up by the ESS to be available to be assigned as a RAID-5 rank. You can format a disk group as CKD or FB, and as RAID or non-RAID, or leave it unassigned.

    DNS
    See domain name system.

    domain
    (1) That part of a computer network in which the data processing resources are under common control.
    (2) In TCP/IP, the naming system used in hierarchical networks.
    (3) A Copy Services server group, in other words, the set of clusters designated by the user to be managed by a particular Copy Services server.

    domain name system (DNS)
    In TCP/IP, the server program that supplies name-to-address translation by mapping domain names to internet addresses. The address of a DNS server is the internet address of the server that hosts the DNS software for the network.

    drawer
    A unit that contains multiple DDMs and provides power, cooling, and related interconnection logic to make the DDMs accessible to attached host systems.

    drive
    (1) A peripheral device, especially one that has addressed storage media. See disk drive module.
    (2) The mechanism used to seek, read, and write information on a storage medium.

    duplex
    (1) Regarding ESS Copy Services, the state of a volume pair after PPRC has completed the copy operation and the volume pair is synchronized.
    (2) In general, pertaining to a communication mode in which data can be sent and received at the same time.

    dynamic sparing
    The ability of a storage server to move data from a failing disk drive module (DDM) to a spare DDM while maintaining storage functions.

    E

    E10
    The forerunner of the F10 model of the ESS. See F10.

    E20
    The forerunner of the F20 model of the ESS. See F20

    EBCDIC
    See extended binary-coded decimal interchange code.

    EC
    See engineering change.

    ECKD
    See extended count key data.

    electrostatic discharge (ESD)
    An undesirable discharge of static electricity that can damage equipment and degrade electrical circuitry.

    emergency power off (EPO)
    A means of turning off power during an emergency, usually a switch.

    EMIF
    See ESCON multiple image facility.

    enclosure
    A unit that houses the components of a storage subsystem, such as a control unit, disk drives, and power source.

    end of file
    A coded character recorded on a data medium to indicate the end of the medium. On a CKD direct access storage device, the subsystem indicates the end of a file by including a record with a data length of zero.

    engineering change (EC)
    An update to a machine, part, or program.

    Enterprise Storage Server
    See IBM TotalStorage Enterprise Storage Server.

    Enterprise Systems Architecture/390(R) (ESA/390) and z/Architecture
    IBM architectures for mainframe computers and peripherals. Processor systems that follow the ESA/390 architecture include the ES/9000(R) family, while the IBM e(logo)server zSeries server uses the z/Architecture.

    Enterprise Systems Connection (ESCON)
    (1) An ESA/390 and zSeries computer peripheral interface. The I/O interface uses ESA/390 logical protocols over a serial interface that configures attached units to a communication fabric.
    (2) A set of IBM products and services that provide a dynamically connected environment within an enterprise.

    EPO
    See emergency power off.

    ERP
    See error recovery procedure.

    error recovery procedure (ERP)
    Procedures designed to help isolate and, where possible, to recover from errors in equipment. The procedures are often used in conjunction with programs that record information on machine malfunctions.

    ESA/390
    See Enterprise Systems Architecture/390.

    ESCD
    See ESCON director.

    ESCON
    See Enterprise System Connection.

    ESCON channel
    An S/390 or zSeries channel that supports ESCON protocols.

    ESCON director (ESCD)
    An I/O interface switch that provides for the interconnection of multiple ESCON interfaces in a distributed-star topology.

    ESCON host systems
    S/390 or zSeries hosts that attach to the ESS with an ESCON adapter. Such host systems run on MVS, VM, VSE, or TPF operating systems.

    ESCON multiple image facility (EMIF)
    In Enterprise Systems Architecture/390, a function that enables LPARs to share an ESCON channel path by providing each LPAR with its own channel-subsystem image.

    EsconNet
    In ESS Specialist, the label on a pseudo-host icon representing a host connection that uses the ESCON protocol and that is not completely defined on the ESS. See pseudo-host and access-any mode.

    ESD
    See electrostatic discharge.

    eserver
    See IBM e(logo)server.

    ESS
    See IBM TotalStorage Enterprise Storage Server.

    ESS Expert
    See IBM StorWatch Enterprise Storage Server Expert.

    ESS Specialist
    See IBM TotalStorage Enterprise Storage Server Specialist.

    ESS Copy Services
    See IBM TotalStorage Enterprise Storage Server Copy Services.

    ESS Master Console
    See IBM TotalStorage ESS Master Console.

    ESSNet
    See IBM TotalStorage Enterprise Storage Server Network.

    Expert
    See IBM StorWatch Enterprise Storage Server Expert.

    extended binary-coded decimal interchange code (EBCDIC)
    A coding scheme developed by IBM used to represent various alphabetic, numeric, and special symbols with a coded character set of 256 eight-bit codes.

    extended count key data (ECKD)
    An extension of the CKD architecture.

    Extended Remote Copy (XRC)
    A function of a storage server that assists a control program to maintain a consistent copy of a logical volume on another storage facility. All modifications of the primary logical volume by any attached host are presented in order to a single host. The host then makes these modifications on the secondary logical volume.

    extent
    A continuous space on a disk that is occupied by or reserved for a particular data set, data space, or file. The unit of increment is a track. See multiple allegiance and parallel access volumes.

    F

    F10
    A model of the ESS featuring a single-phase power supply. It has fewer expansion capabilities than the Model F20.

    F20
    A model of the ESS featuring a three-phase power supply. It has more expansion capabilities than the Model F10, including the ability to support a separate expansion enclosure.

    fabric
    In fibre-channel technology, a routing structure, such as a switch, receives addressed information and routes to the appropriate destination. A fabric can consist of more than one switch. When multiple fibre-channel switches are interconnected, they are said to be cascaded.

    failback
    Cluster recovery from failover following repair. See failover.

    failover
    On the ESS, the process of transferring all control of a storage facility to a single cluster when the other cluster in the storage facility fails.

    fast write
    A write operation at cache speed that does not require immediate transfer of data to a disk drive. The subsystem writes the data directly to cache, to nonvolatile storage, or to both. The data is then available for destaging. A fast-write operation reduces the time an application must wait for the I/O operation to complete.

    FBA
    See fixed-block architecture.

    FC-AL
    See Fibre Channel-Arbitrated Loop.

    FCP
    See fibre-channel protocol.

    FCS
    See fibre-channel standard.

    feature code
    A code that identifies a particular orderable option and that is used by service personnel to process hardware and software orders. Individual optional features are each identified by a unique feature code.

    fibre channel (FC)
    A data-transmission architecture based on the ANSI fibre-channel standard, which supports full-duplex communication. The ESS supports data transmission over fiber-optic cable through its fibre-channel adapters. See fibre-channel protocoland fibre-channel standard.

    Fibre Channel-Arbitrated Loop (FC-AL)
    An implementation of the fibre-channel standard that uses a ring topology for the communication fabric. Refer to American National Standards Institute (ANSI) X3T11/93-275. In this topology, two or more fibre-channel end points are interconnected through a looped interface. The ESS supports this topology.

    fibre-channel connection (FICON)
    A fibre-channel communications protocol designed for IBM mainframe computers and peripherals.

    fibre-channel protocol (FCP)
    For fibre-channel communication, the protocol has five layers. The layers define how fibre-channel ports interact through their physical links to communicate with other ports.

    fibre-channel standard (FCS)
    An ANSI standard for a computer peripheral interface. The I/O interface defines a protocol for communication over a serial interface that configures attached units to a communication fabric. The protocol has two layers. The IP layer defines basic interconnection protocols. The upper layer supports one or more logical protocols. Refer to American National Standards Institute (ANSI) X3.230-199x.

    FICON
    See fibre-channel connection.

    FiconNet
    In ESS Specialist, the label on a pseudo-host icon representing a host connection that uses the FICON protocol and that is not completely defined on the ESS. See pseudo-host and access-any mode.

    field replaceable unit (FRU)
    An assembly that is replaced in its entirety when any one of its components fails. In some cases, a field replaceable unit may contain other field replaceable units.

    FIFO
    See first-in-first-out.

    firewall
    A protection against unauthorized connection to a computer or a data storage system. The protection is usually in the form of software on a gateway server that grants access to users who meet authorization criteria.

    first-in-first-out (FIFO)
    A queuing technique in which the next item to be retrieved is the item that has been in the queue for the longest time. (A)

    fixed-block architecture (FBA)
    An architecture for logical devices that specifies the format of and access mechanisms for the logical data units on the device. The logical data unit is a block. All blocks on the device are the same size (fixed size). The subsystem can access them independently.

    fixed-block device
    An architecture for logical devices that specifies the format of the logical data units on the device. The logical data unit is a block. All blocks on the device are the same size (fixed size); the subsystem can access them independently. This is the required format of the logical data units for host systems that attach with a Small Computer System Interface (SCSI) or fibre-channel interface. See fibre-channel, Small Computer System Interface and SCSI-FCP.

    FlashCopy
    An optional feature for the ESS that can make an instant copy of data, that is, a point-in-time copy of a volume.

    FRU
    See field replaceable unit.

    full duplex
    See duplex.

    G

    GB
    See gigabyte.

    gigabyte (GB)
    A gigabyte of storage is 109 bytes. A gigabyte of memory is 230 bytes.

    GDPS
    Geographically Dispersed Parallel Sysplex, an S/390 multi-site application availability solution.

    group
    See disk drive module group or Copy Services server group.

    H

    HA
    See host adapter.

    HACMP
    Software that provides host clustering, so that a failure of one host is recovered by moving jobs to other hosts within the cluster; named for high-availability cluster multiprocessing

    hard disk drive (HDD)
    (1) A storage medium within a storage server used to maintain information that the storage server requires.
    (2) A mass storage medium for computers that is typically available as a fixed disk (such as the disks used in system units of personal computers or in drives that are external to a personal computer) or a removable cartridge.

    Hardware Service Manager (HSM)
    An option selected from System Service Tools or Dedicated Service Tools on the AS/400 or iSeries host that enables the user to display and work with system hardware resources, and to debug input-output processors (IOP), input-output adapters (IOA), and devices.

    HDA
    See head and disk assembly.

    HDD
    See hard disk drive.

    hdisk
    An AIX term for storage space.

    head and disk assembly (HDA)
    The portion of an HDD associated with the medium and the read/write head.

    heartbeat
    A status report sent at regular intervals from the ESS. The service provider uses this report to monitor the health of the call home process. See call home, heartbeat call home record, and remote technical assistance information network.

    heartbeat call home record
    Machine operating and service information sent to a service machine. These records might include such information as feature code information and product logical configuration information.

    High Speed Link (HSL)
    Bus technology for input-output tower attachment on iSeries host.

    home address
    A nine-byte field at the beginning of a track that contains information that identifies the physical track and its association with a cylinder.

    hop
    Interswitch connection. A hop count is the number of connections that a particular block of data traverses between source and destination. For example, data traveling from one hub over a wire to another hub traverses one hop.

    host adapter (HA)
    A physical subunit of a storage server that provides the ability to attach to one or more host I/O interfaces. The Enterprise Storage Server has four HA bays, two in each cluster. Each bay supports up to four host adapters.

    host processor
    A processor that controls all or part of a user application network. In a network, the processing unit in which the data communication access method resides. See host system.

    host system
    (1) A computer system that is connected to the ESS. The ESS supports both mainframe (S/390 or zSeries) hosts as well as open-systems hosts. S/390 or zSeries hosts are connected to the ESS through ESCON or FICON interfaces. Open-systems hosts are connected to the ESS by SCSI or fibre-channel interfaces.
    (2) The data processing system to which a network is connected and with which the system can communicate.
    (3) The controlling or highest level system in a data communication configuration.

    hot plug
    Pertaining to the ability to add or remove a hardware facility or resource to a unit while power is on.

    HSL
    See High Speed Link.

    I

    IBM e(logo)server
    The brand name for a series of server products that are optimized for e-commerce. The products include the iSeries, pSeries, xSeries, and zSeries.

    IBM product engineering (PE)
    The third-level of IBM service support. Product engineering is composed of IBM engineers who have experience in supporting a product or who are knowledgeable about the product.

    IBM StorWatch Enterprise Storage Server Expert (ESS Expert)
    The software that gathers performance data from the ESS and presents it through a Web browser.

    IBM TotalStorage Enterprise Storage Server (ESS)
    A member of the Seascape(R) product family of storage servers and attached storage devices (disk drive modules). The ESS provides for high-performance, fault-tolerant storage and management of enterprise data, providing access through multiple concurrent operating systems and communication protocols. High performance is provided by four symmetric multiprocessors, integrated caching, RAID support for the disk drive modules, and disk access through a high-speed serial storage architecture (SSA) interface.

    IBM TotalStorage Enterprise Storage Server Specialist (ESS Specialist)
    Software with a Web-browser interface for configuring the ESS.

    IBM TotalStorage Enterprise Storage Server Copy Services (ESS Copy Services)
    Software with a Web-browser interface for configuring, managing, and monitoring the data-copy functions of FlashCopy and PPRC.

    IBM TotalStorage Enterprise Storage Server Network (ESSNet)
    A private network providing Web browser access to the ESS. IBM installs the ESSNet software on an IBM workstation called the IBM TotalStorage ESS Master Console, supplied with the first ESS delivery.

    IBM TotalStorage ESS Master Console (ESS Master Console)
    An IBM workstation (formerly named the ESSNet console and hereafter referred to simply as the ESS Master Console) that IBM installs to provide the ESSNet facility when they install your ESS. It includes a Web browser that provides links to the ESS user interface, including ESS Specialist and ESS Copy Services.

    ID
    See identifier.

    identifier (ID)
    A unique name or address that identifies things such as programs, devices, or systems.

    IML
    See initial microprogram load.

    implicit allegiance
    In Enterprise Systems Architecture/390, a relationship that a control unit creates between a device and a channel path when the device accepts a read or write operation. The control unit guarantees access to the channel program over the set of channel paths that it associates with the allegiance.

    initial microprogram load (IML)
    To load and initiate microcode or firmware that controls a hardware entity such as a processor or a storage server.

    initial program load (IPL)
    To load and initiate the software, typically an operating system that controls a host computer.

    initiator
    A SCSI device that communicates with and controls one or more targets. An initiator is typically an I/O adapter on a host computer. A SCSI initiator is analogous to an S/390 channel. A SCSI logical unit is analogous to an S/390 device. See target.

    i-node
    The internal structure in an AIX operating system that describes the individual files in the operating system. It contains the code, type, location, and owner of a file.

    input/output (I/O)
    Pertaining to (a) input, output, or both or (b) a device, process, or channel involved in data input, data output, or both.

    Internet Protocol (IP)
    In the Internet suite of protocols, a protocol without connections that routes data through a network or interconnecting networks and acts as an intermediary between the higher protocol layers and the physical network. The upper layer supports one or more logical protocols (for example, a SCSI-command protocol and an ESA/390 command protocol). Refer to ANSI X3.230-199x. The IP acronym is the IP in TCP/IP. See Transmission Control Protocol/Internet Protocol.

    invalidate
    To remove a logical data unit from cache memory, because it cannot support continued access to the logical data unit on the device. This removal may be the result of a failure within the storage server or a storage device that is associated with the device.

    I/O
    See input/output.

    I/O adapter (IOA)
    Input-output adapter on the PCI bus.

    I/O device
    An addressable read and write unit, such as a disk drive device, magnetic tape device, or printer.

    I/O interface
    An interface that enables a host to perform read and write operations with its associated peripheral devices.

    I/O Priority Queueing
    Facility provided by the Workload Manager of OS/390 and supported by the ESS that enables the systems administrator to set priorities for queueing I/Os from different system images. See multiple allegiance and parallel access volume.

    I/O processor (IOP)
    Controls input-output adapters and other devices.

    IP
    See Internet Protocol.

    IPL
    See initial program load.

    iSeries
    An IBM e(logo)server product that emphasizes integration.

    J

    Java virtual machine (JVM)
    A software implementation of a central processing unit (CPU) that runs compiled Java code (applets and applications).

    JVM
    See Java virtual machine.

    K

    KB
    See kilobyte.

    key field
    The second (optional) field of a CKD record. The key length is specified in the count field. The key length determines the field length. The program writes the data in the key field and use the key field to identify or locate a given record. The subsystem does not use the key field.

    kilobyte (KB)
    (1) For processor storage, real, and virtual storage, and channel volume, 210 or 1024 bytes.
    (2) For disk storage capacity and communications volume, 1000 bytes.

    Korn shell
    Interactive command interpreter and a command programming language.

    KPOH
    See thousands of power-on hours.

    L

    LAN
    See local area network.

    last-in first-out (LIFO)
    A queuing technique in which the next item to be retrieved is the item most recently placed in the queue. (A)

    LBA
    See logical block address.

    LCU
    See logical control unit.

    least recently used (LRU)
    (1) The algorithm used to identify and make available the cache space that contains the least-recently used data.
    (2) A policy for a caching algorithm that chooses to remove from cache the item that has the longest elapsed time since its last access.

    LED
    See light-emitting diode.

    LIC
    See licensed internal code.

    licensed internal code (LIC)
    Microcode that IBM does not sell as part of a machine, but licenses to the customer. LIC is implemented in a part of storage that is not addressable by user programs. Some IBM products use it to implement functions as an alternate to hard-wired circuitry.

    LIFO
    See last-in first-out.

    light-emitting diode (LED)
    A semiconductor chip that gives off visible or infrared light when activated.

    link address
    On an ESCON or FICON interface, the portion of a source or destination address in a frame that ESCON or FICON uses to route a frame through an ESCON or FICON director. ESCON or FICON associates the link address with a specific switch port that is on the ESCON or FICON director. Equivalently, it associates the link address with the channel-subsystem or control unit link-level functions that are attached to the switch port.

    link-level facility
    The ESCON or FICON hardware and logical functions of a control unit or channel subsystem that allow communication over an ESCON or FICON write interface and an ESCON or FICON read interface.

    local area network (LAN)
    A computer network located on a user's premises within a limited geographic area.

    local e-mail
    An e-mail configuration option for storage servers that are connected to a host-system network that does not have a domain name system (DNS) server.

    logical address
    On an ESCON or FICON interface, the portion of a source or destination address in a frame used to select a specific channel-subsystem or control-unit image.

    logical block address (LBA)
    The address assigned by the ESS to a sector of a disk.

    logical control unit (LCU)
    See control-unit image.

    logical data unit
    A unit of storage that is accessible on a given device.

    logical device
    The facilities of a storage server (such as the ESS) associated with the processing of I/O operations directed to a single host-accessible emulated I/O device. The associated storage is referred to as a logical volume. The logical device is mapped to one or more host-addressable units, such as a device on an S/390 I/O interface or a logical unit on a SCSI I/O interface, such that the host initiating I/O operations to the I/O-addressable unit interacts with the storage on the associated logical device.

    logical partition (LPAR)
    A set of functions that create the programming environment that is defined by the ESA/390 architecture. ESA/390 architecture uses this term when more than one LPAR is established on a processor. An LPAR is conceptually similar to a virtual machine environment except that the LPAR is a function of the processor. Also the LPAR does not depend on an operating system to create the virtual machine environment.

    logical path
    For Copy Services, a relationship between a source logical subsystem and target logical subsystem that is created over a physical path through the interconnection fabric used for Copy Services functions.

    logical subsystem (LSS)
    Pertaining to the ESS, a construct that consists of a group of up to 256 logical devices. An ESS can have up to 16 CKD-formatted logical subsystems (4096 CKD logical devices) and also up to 16 fixed-block (FB) logical subsystems (4096 FB logical devices). The logical subsystem facilitates configuration of the ESS and may have other implications relative to the operation of certain functions. There is a one-to-one mapping between a CKD logical subsystem and an S/390 control-unit image.

    For S/390 or zSeries hosts, a logical subsystem represents a logical control unit (LCU). Each control-unit image is associated with only one logical subsystem. See control-unit image.

    logical unit
    The open-systems term for a logical disk drive.

    logical unit number (LUN)
    A SCSI term for a unique number used on a SCSI bus to enable it to differentiate between up to eight separate devices, each of which is a logical unit.

    logical volume
    The storage medium associated with a logical disk drive. A logical volume typically resides on one or more storage devices. The ESS administrator defines this unit of storage. The logical volume, when residing on a RAID-5 array, is spread over 6 +P or 7 +P drives, where P is parity. A logical volume can also reside on a non-RAID storage device. See count key data and fixed block address.

    logical volume manager (LVM)
    A set of system commands, library routines, and other tools that allow the user to establish and control logical volume storage. The LVM maps data between the logical view of storage space and the physical disk drive module (DDM).

    longitudinal redundancy check (LRC)
    A method of error-checking during data transfer that involves checking parity on a row of binary digits that are members of a set that forms a matrix. Longitudinal redundancy check is also called a longitudinal parity check.

    longwave laser adapter
    A connector used between host and the ESS to support longwave fibre-channel communication.

    loop
    The physical connection between a pair of device adapters in the ESS. See device adapter.

    LPAR
    See logical partition.

    LRC
    See longitudinal redundancy check.

    LRU
    See least recently used.

    LSS
    See logical subsystem.

    LUN
    See logical unit number.

    LVM
    See logical volume manager.

    M

    machine level control (MLC)
    A database that contains the EC level and configuration of products in the field.

    machine reported product data (MRPD)
    Product data gathered by a machine and sent to a destination such as an IBM support server or RETAIN. These records might include such information as feature code information and product logical configuration information.

    mainframe
    A computer, usually in a computer center, with extensive capabilities and resources to which other computers may be connected so that they can share facilities. (T)

    maintenance analysis procedure (MAP)
    A hardware maintenance document that gives an IBM service representative a step-by-step procedure for tracing a symptom to the cause of a failure.

    management information base (MIB)
    (1) A schema for defining a tree structure that identifies and defines certain objects that can be passed between units using an SNMP protocol. The objects passed typically contain certain information about the product such as the physical or logical characteristics of the product.
    (2) Shorthand for referring to the MIB-based record of a network device. Information about a managed device is defined and stored in the management information base (MIB) of the device. Each ESS has a MIB. SNMP-based network management software uses the record to identify the device. See simple network management protocol.

    MAP
    See maintenance analysis procedure.

    Master Console
    See IBM TotalStorage ESS Master Console.

    MB
    See megabyte.

    MCA
    See Micro Channel architecture.

    mean time between failures (MTBF)
    (1) A projection of the time that an individual unit remains functional. The time is based on averaging the performance, or projected performance, of a population of statistically independent units. The units operate under a set of conditions or assumptions.
    (2) For a stated period in the life of a functional unit, the mean value of the lengths of time between consecutive failures under stated conditions. (I) (A)

    medium
    For a storage facility, the disk surface on which data is stored.

    megabyte (MB)
    (1) For processor storage, real and virtual storage, and channel volume, 220 or 1 048 576 bytes.
    (2) For disk storage capacity and communications volume, 1 000 000 bytes.

    MES
    See miscellaneous equipment specification.

    MIB
    See management information base.

    Micro Channel architecture (MCA)
    The rules that define how subsystems and adapters use the Micro Channel bus in a computer. The architecture defines the services that each subsystem can or must provide.

    Microsoft Internet Explorer (MSIE)
    Web browser software manufactured by Microsoft.

    MIH
    See missing-interrupt handler.

    mirrored pair
    Two units that contain the same data. The system refers to them as one entity.

    mirroring
    In host systems, the process of writing the same data to two disk units within the same auxiliary storage pool at the same time.

    miscellaneous equipment specification (MES)
    IBM field-installed change to a machine.

    missing-interrupt handler (MIH)
    An MVS and MVS/XA facility that tracks I/O interrupts. MIH informs the operator and creates a record whenever an expected interrupt fails to occur before a specified elapsed time is exceeded.

    MLC
    See machine level control.

    mobile service terminal (MoST)
    The mobile terminal used by service personnel.

    Model 100
    A 2105 Model 100, often simply referred to as a Mod 100, is an expansion enclosure for the ESS. See 2105 and

    MoST
    See mobile service terminal.

    MRPD
    See machine reported product data.

    MSIE
    See Microsoft Internet Explorer.

    MTBF
    See mean time between failures.

    multiple allegiance
    An ESS hardware function that is independent of software support. This function enables multiple system images to concurrently access the same logical volume on the ESS as long as the system images are accessing different extents. See extent and parallel access volumes.

    multiple virtual storage (MVS)
    Implies MVS/390, MVS/XA, MVS/ESA, and the MVS element of the OS/390 operating system.

    MVS
    See multiple virtual storage.

    N

    Netfinity
    Obsolete brand name of an IBM Intel-processor-based server.

    Netscape Navigator
    Web browser software manufactured by Netscape.

    node
    The unit that is connected in a fibre-channel network. An ESS is a node in a fibre-channel network.

    non-RAID
    A disk drive set up independently of other disk drives and not set up as part of a disk drive module group to store data using the redundant array of disks (RAID) data-striping methodology.

    nonremovable medium
    A recording medium that cannot be added to or removed from a storage device.

    nonretentive data
    Data that the control program can easily recreate in the event it is lost. The control program may cache nonretentive write data in volatile memory.

    nonvolatile storage (NVS)
    (1) Typically refers to nonvolatile memory on a processor rather than to a nonvolatile disk storage device. On a storage facility, nonvolatile storage is used to store active write data to avoid data loss in the event of a power loss.
    (2) A storage device whose contents are not lost when power is cut off.

    NVS
    See nonvolatile storage.

    O

    octet
    In Internet Protocol (IP) addressing, one of the four parts of a 32-bit integer presented in dotted decimal notation. dotted decimal notation consists of four 8-bit numbers written in base 10. For example, 9.113.76.250 is an IP address containing the octets 9, 113, 76, and 250.

    OEMI
    See original equipment manufacturer's information.

    open system
    A system whose characteristics comply with standards made available throughout the industry and that therefore can be connected to other systems complying with the same standards. Applied to the ESS, such systems are those hosts that connect to the ESS through SCSI or SCSI-FCP adapters. See Small Computer System Interface and SCSI-FCP.

    organizationally unique identifier (OUI)
    An IEEE-standards number that identifies an organization with a 24-bit globally unique assigned number referenced by various standards. OUI is used in the family of 802 LAN standards, such as Ethernet and Token Ring.

    original equipment manufacturer's information (OEMI)
    A reference to an IBM guideline for a computer peripheral interface. The interface uses ESA/390 logical protocols over an I/O interface that configures attached units in a multidrop bus topology.

    OUI
    See organizationally unique identifier.

    P

    panel
    The formatted display of information that appears on a display screen.

    parallel access volume (PAV)
    An advanced function of the ESS that enables OS/390 and z/OS systems to issue concurrent I/O requests against a CKD logical volume by associating multiple devices of a single control-unit image with a single logical device. Up to 8 device addresses can be assigned to a parallel access volume. PAV enables two or more concurrent writes to the same logical volume, as long as the writes are not to the same extents. See extent, I/O Priority Queueing, and multiple allegiance.

    parity
    A data checking scheme used in a computer system to ensure the integrity of the data. The RAID implementation uses parity to recreate data if a disk drive fails.

    path group
    The ESA/390 term for a set of channel paths that are defined to a control unit as being associated with a single logical partition (LPAR). The channel paths are in a group state and are online to the host. See logical partition.

    path group identifier
    The ESA/390 term for the identifier that uniquely identifies a given logical partition (LPAR). The path group identifier is used in communication between the LPAR program and a device. The identifier associates the path group with one or more channel paths, thereby defining these paths to the control unit as being associated with the same LPAR.

    PAV
    See parallel access volume.

    PCI
    See peripheral component interconnect.

    PE
    See IBM product engineering.

    Peer-to-Peer Remote Copy (PPRC)
    A function of a storage server that maintains a consistent copy of a logical volume on the same storage server or on another storage server. All modifications that any attached host performs on the primary logical volume are also performed on the secondary logical volume.

    peripheral component interconnect (PCI)
    An architecture for a system bus and associated protocols that supports attachments of adapter cards to a system backplane.

    physical path
    A single path through the I/O interconnection fabric that attaches two units. For Copy Services, this is the path from a host adapter on one ESS (through cabling and switches) to a host adapter on another ESS.

    point-to-point connection
    For fibre-channel connections, a topology that enables the direct interconnection of ports. See arbitrated loop and switched fabric.

    POST
    See power-on self test.

    power-on self test (POST)
    A diagnostic test run by servers or computers when they are turned on.

    PPRC
    See Peer-to-Peer Remote Copy.

    predictable write
    A write operation that can cache without knowledge of the existing format on the medium. All writes on FBA DASD devices are predictable. On CKD DASD devices, a write is predictable if it does a format write for the first data record on the track.

    primary Copy Services server
    One of two Copy Services servers in a Copy Services domain. The primary Copy Services server is the active Copy Services server until it fails; it is then replaced by the backup Copy Services server. A Copy Services server is software that runs in one of the two clusters of an ESS and performs data-copy operations within that group. See active Copy Services server and backup Copy Services server.

    product engineering
    See IBM product engineering.

    program
    On a computer, a generic term for software that controls the operation of the computer. Typically, the program is a logical assemblage of software modules that perform multiple related tasks.

    program-controlled interruption
    An interruption that occurs when an I/O channel fetches a channel command word with the program-controlled interruption flag on.

    program temporary fix (PTF)
    A temporary solution or bypass of a problem diagnosed by IBM in a current unaltered release of a program

    promote
    To add a logical data unit to cache memory.

    protected volume
    An AS/400 term for a disk storage device that is protected from data loss by RAID techniques. An AS/400 host does not mirror a volume configured as a protected volume, while it does mirror all volumes configured as unprotected volumes. The ESS, however, can be configured to indicate that an AS/400 volume is protected or unprotected and give it RAID protection in either case.

    pSeries
    An IBM e(logo)server product that emphasizes performance.

    pseudo-host
    A host connection that is not explicitly defined to the ESS and that has access to at least one volume that is configured on the ESS. The FiconNet pseudo-host icon represents the FICON protocol. The EsconNet pseudo-host icon represents the ESCON protocol. The pseudo-host icon labelled "Anonymous" represents hosts connected through the SCSI-FCP protocol. Anonymous host is a commonly used synonym for pseudo-host. The ESS adds a pseudo-host icon only when the ESS is set to access-any mode. See access-any mode.

    PTF
    See program temporary fix.

    PV Links
    Short for Physical Volume Links, an alternate pathing solution from Hewlett-Packard providing for multiple paths to a volume, as well as static load balancing.

    R

    rack
    See enclosure.

    RAID
    See redundant array of inexpensive disks and array. RAID also is expanded to redundant array of independent disks.

    RAID 5
    A type of RAID that optimizes cost-effective performance through data striping. RAID 5 provides fault tolerance for up to two failed disk drives by distributing parity across all of the drives in the array plus one parity disk drive. The ESS automatically reserves spare disk drives when it assigns arrays to a device adapter pair (DA pair). See device adapter.

    random access
    A mode of accessing data on a medium in a manner that requires the storage device to access nonconsecutive storage locations on the medium.

    redundant array of inexpensive disks (RAID)
    A methodology of grouping disk drives for managing disk storage to insulate data from a failing disk drive.

    remote technical assistance information network (RETAIN)
    The initial service tracking system for IBM service support, which captures heartbeat and call-home records. See support catcher and support catcher telephone number.

    REQ/ACK
    See request for acknowledgement and acknowledgement.

    request for acknowledgement and acknowledgement (REQ/ACK)
    A cycle of communication between two data transport devices for the purpose of verifying the connection, which starts with a request for acknowledgement from one of the devices and ends with an acknowledgement from the second device.

    reserved allegiance
    In Enterprise Systems Architecture/390, a relationship that is created in a control unit between a device and a channel path when a Sense Reserve command is completed by the device. The allegiance causes the control unit to guarantee access (busy status is not presented) to the device. Access is over the set of channel paths that are associated with the allegiance; access is for one or more channel programs, until the allegiance ends.

    RETAIN
    See remote technical assistance information network.

    R0
    See track-descriptor record.

    S

    S/390 and zSeries
    IBM enterprise servers based on Enterprise Systems Architecture/390 (ESA/390) and z/Architecture, respectively. "S/390" is a shortened form of the original name "System/390".

    S/390 and zSeries storage
    Storage arrays and logical volumes that are defined in the ESS as connected to S/390 and zSeries servers. This term is synonymous with count-key-data (CKD) storage.

    SAID
    See system adapter identification number.

    SAM
    See sequential access method.

    SAN
    See storage area network.

    SBCON
    See Single-Byte Command Code Sets Connection.

    screen
    The physical surface of a display device upon which information is shown to users.

    SCSI
    See Small Computer System Interface.

    SCSI device
    A disk drive connected to a host through a an I/O interface using the SCSI protocol. A SCSI device is either an initiator or a target. See initiator and Small Computer System Interface.

    SCSI host systems
    Host systems that are attached to the ESS with a SCSI interface. Such host systems run on UNIX, OS/400, Windows NT, Windows 2000, or Novell NetWare operating systems.

    SCSI ID
    A unique identifier assigned to a SCSI device that is used in protocols on the SCSI interface to identify or select the device. The number of data bits on the SCSI bus determines the number of available SCSI IDs. A wide interface has 16 bits, with 16 possible IDs.

    SCSI-FCP
    Short for SCSI-to-fibre-channel protocol, a protocol used to transport data between a SCSI adapter on an open-systems host and a fibre-channel adapter on an ESS. See fibre-channel protocol and Small Computer System Interface.

    Seascape architecture
    A storage system architecture developed by IBM for open-systems servers and S/390 and zSeries host systems. It provides storage solutions that integrate software, storage management, and technology for disk, tape, and optical storage.

    serial connection
    A method of device interconnection for determining interrupt priority by connecting the interrupt sources serially.

    self-timed interface (STI)
    An interface that has one or more conductors that transmit information serially between two interconnected units without requiring any clock signals to recover the data. The interface performs clock recovery independently on each serial data stream and uses information in the data stream to determine character boundaries and inter-conductor synchronization.

    sequential access
    A mode of accessing data on a medium in a manner that requires the storage device to access consecutive storage locations on the medium.

    sequential access method (SAM)
    An access method for storing, deleting, or retrieving data in a continuous sequence based on the logical order of the records in the file.

    serial storage architecture (SSA)
    An IBM standard for a computer peripheral interface. The interface uses a SCSI logical protocol over a serial interface that configures attached targets and initiators in a ring topology. See SSA adapter.

    server
    (1) A type of host that provides certain services to other hosts that are referred to as clients.
    (2) A functional unit that provides services to one or more clients over a network.

    service boundary
    A category that identifies a group of components that are unavailable for use when one of the components of the group is being serviced. Service boundaries are provided on the ESS, for example, in each host bay and in each cluster.

    service information message (SIM)
    A message sent by a storage server to service personnel through an S/390 operating system.

    service personnel
    A generalization referring to individuals or companies authorized to service the ESS. The terms "service provider", "service representative", and "IBM service support representative (SSR)" refer to types of service personnel. See service support representative.

    service processor
    A dedicated processing unit used to service a storage facility.

    service support representative (SSR)
    Individuals or a company authorized to service the ESS. This term also refers to a service provider, a service representative, or an IBM service support representative (SSR). An IBM SSR installs the ESS.

    shared storage
    Storage within an ESS that is configured so that multiple homogeneous or divergent hosts can concurrently access the storage. The storage has a uniform appearance to all hosts. The host programs that access the storage must have a common model for the information on a storage device. The programs must be designed to handle the effects of concurrent access.

    shortwave laser adapter
    A connector used between host and ESS to support shortwave fibre-channel communication.

    SIM
    See service-information message.

    Simple Network Management Protocol (SNMP)
    In the Internet suite of protocols, a network management protocol that is used to monitor routers and attached networks. SNMP is an application layer protocol. Information on devices managed is defined and stored in the application's Management Information Base (MIB). See management information base.

    simplex volume
    A volume that is not part of a FlashCopy, XRC, or PPRC volume pair.

    Single-Byte Command Code Sets Connection (SBCON)
    The ANSI standard for the ESCON or FICON I/O interface.

    Small Computer System Interface (SCSI)
    (1) An ANSI standard for a logical interface to computer peripherals and for a computer peripheral interface. The interface uses a SCSI logical protocol over an I/O interface that configures attached initiators and targets in a multidrop bus topology.
    (2) A standard hardware interface that enables a variety of peripheral devices to communicate with one another.

    SMIT
    See System Management Interface Tool.

    SMP
    See symmetric multi-processor.

    SNMP
    See simple network management protocol.

    software transparency
    Criteria applied to a processing environment that states that changes do not require modifications to the host software in order to continue to provide an existing function.

    spare
    A disk drive on the ESS that can replace a failed disk drive. A spare can be predesignated to allow automatic dynamic sparing. Any data preexisting on a disk drive that is invoked as a spare is destroyed by the dynamic sparing copy process.

    spatial reuse
    A feature of serial storage architecture that enables a device adapter loop to support many simultaneous read/write operations. See serial storage architecture.

    Specialist
    See IBM TotalStorage Enterprise Storage Server Specialist.

    SSA
    See serial storage architecture.

    SSA adapter
    A physical adapter based on serial storage architecture. SSA adapters connect disk drive modules to ESS clusters. See serial storage architecture.

    SSID
    See subsystem identifier.

    SSR
    See service support representative.

    stacked status
    In Enterprise Systems Architecture/390, the condition when the control unit is holding status for the channel, and the channel responded with the stack-status control the last time the control unit attempted to present the status.

    stage operation
    The operation of reading data from the physical disk drive into the cache.

    staging
    To move data from an offline or low-priority device back to an online or higher priority device, usually on demand of the system or on request of the user.

    STI
    See self-timed interface.

    storage area network
    A network that connects a company's heterogeneous storage resources.

    storage complex
    Multiple storage facilities.

    storage device
    A physical unit that provides a mechanism to store data on a given medium such that it can be subsequently retrieved. See disk drive module.

    storage facility
    (1) A physical unit that consists of a storage server integrated with one or more storage devices to provide storage capability to a host computer.
    (2) A storage server and its attached storage devices.

    storage server
    A physical unit that manages attached storage devices and provides an interface between them and a host computer by providing the function of one or more logical subsystems. The storage server can provide functions that are not provided by the storage device. The storage server has one or more clusters.

    striping
    A technique that distributes data in bit, byte, multi-byte, record, or block increments across multiple disk drives.

    subchannel
    A logical function of a channel subsystem associated with the management of a single device.

    subsystem identifier (SSID)
    A number that uniquely identifies a logical subsystem within a computer installation.

    support catcher
    A server to which a machine sends a trace or a dump package.

    support catcher telephone number
    The telephone number that connects the support catcher server to the ESS to receive a trace or dump package. See support catcher. See remote technical assistance information network.

    switched fabric
    One of three a fibre-channel connection topologies supported by the ESS. See arbitrated loop and point-to-point.

    symmetric multi-processor (SMP)
    An implementation of a multi-processor computer consisting of several identical processors configured in a way that any subset of the set of processors is capable of continuing the operation of the computer. The ESS contains four processors set up in SMP mode.

    synchronous write
    A write operation whose completion is indicated after the data has been stored on a storage device.

    System/390
    See S/390.

    system adapter identification number (SAID)

    System Management Interface Tool (SMIT)
    An interface tool of the AIX operating system for installing, maintaining, configuring, and diagnosing tasks.

    System Modification Program (SMP)
    A program used to install software and software changes on MVS systems.

    T

    TAP
    See Telocator Alphanumeric Protocol.

    target
    A SCSI device that acts as a slave to an initiator and consists of a set of one or more logical units, each with an assigned logical unit number (LUN). The logical units on the target are typically I/O devices. A SCSI target is analogous to an S/390 control unit. A SCSI initiator is analogous to an S/390 channel. A SCSI logical unit is analogous to an S/390 device. See Small Computer System Interface.

    TB
    See terabyte.

    TCP/IP
    See Transmission Control Protocol/Internet Protocol.

    Telocator Alphanumeric Protocol (TAP)
    An industry standard protocol for the input of paging requests.

    terabyte (TB)
    (1) Nominally, 1 000 000 000 000 bytes, which is accurate when speaking of bandwidth and disk storage capacity.
    (2) For ESS cache memory, processor storage, real and virtual storage, a terabyte refers to 240 or 1 099 511 627 776 bytes.

    thousands of power-on hours (KPOH)
    A unit of time used to measure the mean time between failures (MTBF).

    time sharing option (TSO)
    An operating system option that provides interactive time sharing from remote terminals.

    TPF
    See transaction processing facility.

    track
    A unit of storage on a CKD device that can be formatted to contain a number of data records. See home address, track-descriptor record, and data record.

    track-descriptor record (R0)
    A special record on a track that follows the home address. The control program uses it to maintain certain information about the track. The record has a count field with a key length of zero, a data length of 8, and a record number of 0. This record is sometimes referred to as R0.

    transaction processing facility (TPF)
    A high-availability, high-performance IBM operating system, designed to support real-time, transaction-driven applications. The specialized architecture of TPF is intended to optimize system efficiency, reliability, and responsiveness for data communication and database processing. TPF provides real-time inquiry and updates to a large, centralized database, where message length is relatively short in both directions, and response time is generally less than three seconds. Formerly known as the Airline Control Program/Transaction Processing Facility (ACP/TPF).

    Transmission Control Protocol/Internet Protocol (TCP/IP)
    (1) Together, the Transmission Control Protocol and the Internet Protocol provide end-to-end connections between applications over interconnected networks of different types.
    (2) The suite of transport and application protocols that run over the Internet Protocol. See Internet Protocol.

    transparency
    See software transparency.

    TSO
    See time sharing option.

    U

    UFS
    UNIX filing system.

    ultra-SCSI
    An enhanced Small Computer System Interface.

    unit address
    The ESA/390 term for the address associated with a device on a given control unit. On ESCON or FICON interfaces, the unit address is the same as the device address. On OEMI interfaces, the unit address specifies a control unit and device pair on the interface.

    unprotected volume
    An AS/400 term that indicates that the AS/400 host recognizes the volume as an unprotected device, even though the storage resides on a RAID array and is therefore fault tolerant by definition. The data in an unprotected volume can be mirrored. Also referred to as an unprotected device.

    upper-layer protocol
    The layer of the Internet Protocol (IP) that supports one or more logical protocols (for example, a SCSI-command protocol and an ESA/390 command protocol). Refer to ANSI X3.230-199x.

    UTC
    See Coordinated Universal Time.

    utility device
    The ESA/390 term for the device used with the Extended Remote Copy facility to access information that describes the modifications performed on the primary copy.

    V

    virtual machine (VM)
    A virtual data processing machine that appears to be for the exclusive use of a particular user, but whose functions are accomplished by sharing the resources of a real data processing system.

    vital product data (VPD)
    Information that uniquely defines the system, hardware, software, and microcode elements of a processing system.

    VM
    See virtual machine.

    volume
    In Enterprise Systems Architecture/390, the information recorded on a single unit of recording medium. Indirectly, it can refer to the unit of recording medium itself. On a nonremovable-medium storage device, the term can also indirectly refer to the storage device associated with the volume. When multiple volumes are stored on a single storage medium transparently to the program, the volumes can be referred to as logical volumes.

    VPD
    See vital product data.

    W

    Web Copy Services
    See IBM TotalStorage Enterprise Storage Server Copy Services.

    worldwide node name (WWNN)
    A unique 64-bit identifier for a host containing a fibre-channel port. See worldwide port name.

    worldwide port name (WWPN)
    A unique 64-bit identifier associated with a fibre-channel adapter port. It is assigned in an implementation- and protocol-independent manner.

    write hit
    A write operation in which the requested data is in the cache.

    write penalty
    The performance impact of a classical RAID 5 write operation.

    WWPN
    See worldwide port name.

    X

    XRC
    See Extended Remote Copy.

    xSeries
    An IBM e(logo)server product that emphasizes architecture.

    Z

    zSeries
    An IBM e(logo)server product that emphasizes near-zero downtime.

    zSeries storage
    See S/390 and zSeries storage.

    Numerics

    2105
    The machine number for the IBM Enterprise Storage Server (ESS). 2105-100 is an ESS expansion enclosure typically referred to as the Model 100. See IBM TotalStorage Enterprise Storage Server and Model 100.

    3390
    The machine number of an IBM disk storage system. The ESS, when interfaced to IBM S/390 or zSeries hosts, is set up to appear as one or more 3390 devices, with a choice of 3390-2, 3390-3, or 3390-9 track formats.

    3990
    The machine number of an IBM control unit.

    7133
    The machine number of an IBM disk storage system. The Model D40 and 020 drawers of the 7133 can be installed in the 2105-100 expansion enclosure of the ESS.

    8-pack
    See disk drive module group.

    Index

    Special Characters
    A B C D E F G H I J K L M N O P R S T U V W
    Special Characters A B C D E F G H I J K L M N O P R S T U V W

    1
    Form Z125-4144