new hitachi nas platform f1000 series · 2019. 4. 23. · preface this user's guide describes...

Hitachi NAS Platform F1000 SeriesEnterprise Array Features Administrator's Guide

MK-92NAS094-02

Product Version

Getting Help

Contents

© 2015 Hitachi, Ltd. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means,electronic or mechanical, including photocopying and recording, or stored in a database or retrievalsystem for any purpose without the express written permission of Hitachi, Ltd.

Hitachi, Ltd., reserves the right to make changes to this document at any time without notice andassumes no responsibility for its use. This document contains the most current information availableat the time of publication. When new or revised information becomes available, this entiredocument will be updated and distributed to all registered users.

Some of the features described in this document might not be currently available. Refer to the mostrecent product announcement for information about feature and product availability, or contactHitachi Data Systems Corporation at https://portal.hds.com.

Notice: Hitachi, Ltd., products and services can be ordered only under the terms and conditions ofthe applicable Hitachi Data Systems Corporation agreements. The use of Hitachi, Ltd., products isgoverned by the terms of your agreements with Hitachi Data Systems Corporation.

Hitachi is a registered trademark of Hitachi, Ltd., in the United States and other countries. HitachiData Systems is a registered trademark and service mark of Hitachi, Ltd., in the United States andother countries.

Archivas, Essential NAS Platform, HiCommand, Hi-Track, ShadowImage, Tagmaserve, Tagmasoft,Tagmasolve, Tagmastore, TrueCopy, Universal Star Network, and Universal Storage Platform areregistered trademarks of Hitachi Data Systems Corporation.

AIX, AS/400, DB2, Domino, DS8000, Enterprise Storage Server, ESCON, FICON, FlashCopy, IBM,Lotus, OS/390, RS6000, S/390, System z9, System z10, Tivoli, VM/ESA, z/OS, z9, zSeries, z/VM, z/VSE are registered trademarks and DS6000, MVS, and z10 are trademarks of International BusinessMachines Corporation.

All other trademarks, service marks, and company names in this document or website areproperties of their respective owners.

Microsoft product screen shots are reprinted with permission from Microsoft Corporation.

iiHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

https://portal.hds.com

Contents

Preface.................................................................................................viiIntended Audience...................................................................................................viiiProduct version.......................................................................................................viiiRelease Notes.........................................................................................................viiiOrganization of HNAS F manuals...............................................................................viiiReferenced Documents..............................................................................................ixDocument Conventions...............................................................................................xConvention for storage capacity values.......................................................................xiGetting help..............................................................................................................xiComments ...............................................................................................................xi

1 Introduction.........................................................................................1-1Supported Program Products...................................................................................1-2

2 Replication Functions............................................................................2-1Overview................................................................................................................2-2

ShadowImage .................................................................................................2-2TrueCopy ........................................................................................................2-2Universal Replicator .........................................................................................2-2

Command Control Interface ....................................................................................2-2Overview of CCI Operation in HNAS F................................................................2-3About Command Devices..................................................................................2-3Scope Related to CCI........................................................................................2-3

How to use ShadowImage.......................................................................................2-6Overview of ShadowImage Operation on the HNAS F..........................................2-6Scope of ShadowImage Operations on the HNAS F.............................................2-8

Scope Related to ShadowImage...............................................................2-8Scope Related to Backup Restore.............................................................2-9Scope Related to the file snapshot functionality.........................................2-9

Prerequisites for Using ShadowImage on the HNAS F..........................................2-9Hardware Prerequisites............................................................................2-9Software Prerequisites.............................................................................2-9

Notes on Operating the Volume Replication Function.........................................2-10Preparing for ShadowImage Operation.............................................................2-15

Preparing for ShadowImage Volume Pair Operation.................................2-15

iiiHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Registering Public Key Used for SSH.......................................................2-15Configuration Settings of CCI.................................................................2-16File System Creation on P-VOL of ShadowImage .....................................2-29

Overview of Using ShadowImage.....................................................................2-30Starting ShadowImage Operations and Creating ShadowImage Pair..........2-30Splitting a ShadowImage Volume Pair.....................................................2-31Resynchronizing a ShadowImage Volume Pair.........................................2-36Deleting a ShadowImage Volume Pair.....................................................2-38

P-VOL Data Recovery from S-VOL....................................................................2-41Recovering Data When the P-VOL File System is in Normal Status.............2-41Recovering Data When the P-VOL File System is Blocked.........................2-45Recovering Data When an Error Occurs on the Device File Used by the P-VOL............................................................................................................2-50

Data Backup Cooperated with a Tape Device....................................................2-56Notes on Failures of External Volumes..............................................................2-57

How to use TrueCopy............................................................................................2-57Overview of TrueCopy Operations in the HNAS F...............................................2-57Scope of TrueCopy Function with the HNAS F...................................................2-58

Scope of TrueCopy Function with the HNAS F..........................................2-58Scope Related to Features in Backup Restore..........................................2-59Scope Related to the file snapshot functionality.......................................2-59

Prerequisites..................................................................................................2-59Hardware Prerequisites..........................................................................2-59Software Prerequisites...........................................................................2-59

Notes on Operations.......................................................................................2-60Preparing for TrueCopy Operations..................................................................2-63

Preparing for TrueCopy Volume Pair Operation........................................2-63Registration of Public Key Used for SSH .................................................2-63Configuring the CCI Environment............................................................2-63Create File System in TrueCopy P-VOL....................................................2-76

Overview of TrueCopy Operations....................................................................2-77Volume Replication................................................................................2-77Disaster Recovery Operations.................................................................2-87

Notes on Failures of External Volumes..............................................................2-92How to use Universal Replicator on the HNAS F.......................................................2-93

Overview of Universal Replicator Operations in HNAS F system..........................2-93Scope of Universal Replicator Function in the HNAS F system.............................2-94

Scope Related to Universal Replicator.....................................................2-94Scope Related to Features in Backup Restore..........................................2-95Scope Related to the file snapshot functionality.......................................2-95

Prerequisites..................................................................................................2-95Hardware Prerequisites..........................................................................2-95Software Prerequisites...........................................................................2-95

Notes on Operations.......................................................................................2-96Preparing for Universal Replicator Operations...................................................2-99

Preparing for Universal Replicator Volume Pair Operation.........................2-99Registration of Public Key Used for SSH .................................................2-99Configuring the CCI Environment............................................................2-99Create File System in Universal Replicator P-VOL...................................2-112

Overview of Universal Replicator Operations...................................................2-113Volume Replication..............................................................................2-113Disaster Recovery Operations...............................................................2-124

ivHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Notes on Failures of External Volumes............................................................2-129Operations of the Log Files..................................................................................2-130

Operations of the Command Control Interface Log Files...................................2-130Format of the Command Control Interface Log Files...............................2-130Downloading the Command Control Interface Log Files to the Client.......2-131Notes on the Operations of the Command Control Interface Log Files......2-131

Commands that HNAS F provides.........................................................................2-132

3 Volume Management Functions.............................................................3-1Dynamic Provisioning..............................................................................................3-2Dynamic Tiering......................................................................................................3-2Universal Volume Manager......................................................................................3-3

Troubleshooting for the HNAS F Including External Storage System.....................3-3Stopping and Restarting External Storage System on Purpose ...................3-3Recovery Procedure in Case of Error in External Storage System................3-5

Volume Migration..................................................................................................3-14Volume Shredder..................................................................................................3-14Encryption License Key..........................................................................................3-14

4 Resource Management Functions...........................................................4-1Storage Navigator...................................................................................................4-2LUN Manager.........................................................................................................4-2Configuration File Loader.........................................................................................4-3Virtual Partition Manager.........................................................................................4-3

5 Performance Management Functions......................................................5-1Performance Monitor...............................................................................................5-2

A Details of ShadowImage Operations on the HNAS F................................A-1Execution of Command Operation on HNAS F ..........................................................A-2

Creating Pairs...................................................................................................A-2Splitting Pairs...................................................................................................A-3Re-synchronizing Pairs....................................................................................A-12Restoring Pairs...............................................................................................A-17Deleting Pairs.................................................................................................A-26

Pairs Recovery from Failures on the HNAS F...........................................................A-35Commands for Pairs Recovery.........................................................................A-35Recovery Procedure from Failures....................................................................A-36

B Details of TrueCopy and Universal Replicator Operations on the HNAS F...B-1Execution of Command Operation on HNAS F...........................................................B-2

Creating Pairs...................................................................................................B-2Splitting Pairs...................................................................................................B-3Re-synchronizing Pairs....................................................................................B-10Deleting Pairs.................................................................................................B-13

C Operation when Failures Occurred on the HNAS F...................................C-1Isolation when Failures Occur .................................................................................C-3

vHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Operation when the Main Site or Storage System Went Down....................................C-4Operation when the Cluster of the Main Site Went Down.........................................C-16Operation when Multiple Failures Occurred in All the Storages on the Main Site.........C-28Operation when Multiple Failures Occurred in a Part of Storages on the Main Site.....C-43Operation when Network Failures Occurred Between the Main Site and the Remote SiteCausing Journal Overflow......................................................................................C-58Operation when Network Failures Occurred Between the Main Site and the Remote SiteWithout Causing Journal Overflow..........................................................................C-60Operation when Network Failures Occurred in the Main Site.....................................C-61Operation when Multiple Failures Occurred in Some or All Journals on the Main Site. .C-72

D System Option Settings........................................................................D-1mode495...............................................................................................................D-2

E Acronyms.............................................................................................E-1Acronyms used in the HNAS F manuals....................................................................E-2

Index

viHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Preface

This user's guide describes and provides instructions for using Hitachi UnifiedStorage VM Series products with the Hitachi NAS Platform F.

Please read this document carefully to understand how to use this product,and maintain a copy for reference purposes.

Notice: The use of Hitachi NAS Platform F and all other Hitachi Data Systemsproducts is governed by the terms of your agreement(s) with Hitachi DataSystems.

This manual is not applicable to single-node configurations.

This preface includes the following information:

□ Intended Audience

□ Product version

□ Release Notes

□ Organization of HNAS F manuals

□ Referenced Documents

□ Document Conventions

□ Convention for storage capacity values

□ Getting help

□ Comments

Preface viiHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Intended AudienceThis document is intended for system administrators, Hitachi Data Systemsrepresentatives, and Authorized Service Providers who are involved ininstalling, configuring, and operating the storage system.

This document assumes the following:

• The user has a background in data processing and understands direct-access storage device (DASD) systems and their basic functions.

• The user is familiar with the Hitachi Unified Storage VM storage systemand has read the Hitachi Unified Storage VM Block Module Hardware UserGuide.

• The user is familiar with the operating system and web browser softwareon the system hosting the Storage Navigator software. For details on theapplicable operating systems and web browser software, please refer toHitachi Storage Navigator User's Guide(User Guide).

• The user has read and understands the Hitachi Command ControlInterface (CCI) User's Guide (User Guide).

• The user has read the Installation and Configuration Guide and is familiarwith how to back up and restore file system data used in a Hitachi NASPlatform F system.

Product versionThis document revision applies to Hitachi NAS Platform F version 5.1.1 orlater.

Release NotesRelease notes can be found on the documentation CD or on the Hitachi DataSystems Support Portal: https://extranet.hds.com/http:/aim.hds.com/portal/dtRelease notes contain requirements and more recent product information thatmay not be fully described in this manual. Be sure to review the release notesbefore installation.

Organization of HNAS F manualsHNAS F manuals are organized as shown below:

Manual name Description

Hitachi NAS Platform F1000 SeriesInstallation and Configuration Guide,MK-92NAS061

You must read this manual first to use an HNASF system.

This manual contains the information that youmust be aware of before starting HNAS F system

viii PrefaceHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

https://extranet.hds.com/http:/aim.hds.com/portal/dt

https://extranet.hds.com/http:/aim.hds.com/portal/dt

Manual name Description

operation, as well as the environment settingsfor an external server.

Hitachi NAS Platform F1000 SeriesCluster Getting Started Guide,MK-92NAS076

This manual explains how to set up an HNAS Fsystem in a cluster configuration.

To operate HNAS F on a virtual server, see theCluster Getting Started Guide for Virtual NAS.

Hitachi NAS Platform F1000 SeriesCluster Getting Started Guide forVirtual NAS, MK-92NAS073

This manual explains how to set up virtualservers for HNAS F systems in a clusterconfiguration.

Hitachi NAS Platform F1000 SeriesCluster Administrator's Guide,MK-92NAS084

This manual provides procedures for using HNASF systems in a cluster configuration, as well asprovides GUI references.

Hitachi NAS Platform F1000 SeriesCluster Troubleshooting Guide,MK-92NAS066

This manual provides troubleshootinginformation for HNAS F systems in a clusterconfiguration.

Hitachi NAS Platform F1000 SeriesSingle Node Getting Started Guide,MK-92NAS079

This manual explains how to set up an HNAS Fsystem in a single-node configuration.

Hitachi NAS Platform F1000 SeriesSingle Node Administrator's Guide,MK-92NAS089

This manual explains the procedures for usingHNAS F systems in a single-node configuration,as well as provides GUI references.

Hitachi NAS Platform F1000 SeriesSingle Node Troubleshooting Guide,MK-92NAS078

This manual provides troubleshootinginformation for HNAS F systems in a single-nodeconfiguration.

Hitachi NAS Platform F1000 Series CLIAdministrator's Guide, MK-92NAS085

This manual describes the syntax of thecommands that can be used for HNAS F systemsin a cluster configuration or a single-nodeconfiguration.

Hitachi NAS Platform F1000 Series APIReferences, MK-92NAS064

This manual explains how to use the API forHNAS F systems in a cluster configuration or asingle-node configuration.

Hitachi NAS Platform F1000 SeriesError Codes, MK-92NAS065

This manual contains messages for HNAS Fsystems in a cluster configuration or a single-node configuration.

Hitachi NAS Platform F1000 Series FileSystem Protocols (CIFS/NFS)Administrator's Guide, MK-92NAS086

This manual contains the things to keep in mindbefore using the CIFS or NFS service of an HNASF system in a cluster configuration or a single-node configuration from a CIFS or NFS client.

Referenced Documents

Hitachi Unified Storage VM

• Hitachi Encryption License Key User's Guide (User Guide)

• Hitachi Performance Manager User's Guide (User Guide)

Preface ixHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

• Hitachi ShadowImage User's Guide (User Guide)

• Hitachi Storage Navigator User's Guide (User Guide)

• Hitachi System Operations Using Spreadsheets

• Hitachi TrueCopy User's Guide (User Guide)

• Hitachi Universal Volume Manager User's Guide (User Guide)

• Hitachi Virtual Partition Manager User's Guide (User Guide)

• Hitachi Volume Migration User Guide

• Hitachi Volume Shredder User Guide

• Provisioning Guide

• Provisioning Guide for Open Systems

Document ConventionsThe terms "Hitachi Unified Storage VM" and "HUS VM" refer to all models ofthe Hitachi Unified Storage Platform VM, unless otherwise noted.

This document uses the following typographic conventions:

TypographicConvention Description

Bold Indicates text on a window, other than the window title,including menus, menu options, buttons, fields, and labels.Example: Click OK.

Italic Indicates a variable, which is a placeholder for actual textprovided by the user or system. Example: copy source-filetarget-file

Note: Angled brackets (< >) are also used to indicatevariables.

screen/code Indicates text that is displayed on screen or entered by theuser. Example: # pairdisplay -g oradb

< > angled brackets Indicates a variable, which is a placeholder for actual textprovided by the user or system. Example: # pairdisplay -g<group>Note: Italic font is also used to indicate variables.

[ ] square brackets Indicates optional values. Example: [ a | b ] indicates that youcan choose a, b, or nothing.

{ } braces Indicates required or expected values. Example: { a | b }indicates that you must choose either a or b.

| vertical bar Indicates that you have a choice between two or more optionsor arguments. Examples:

[ a | b ] indicates that you can choose a, b, or nothing.

{ a | b } indicates that you must choose either a or b.

x PrefaceHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Convention for storage capacity valuesStorage capacity values (e.g., drive capacity) are calculated based on thefollowing values:

Capacity Unit Physical Value Logical Value

1 KB 1,000 bytes 1,024 (210) bytes

1 MB 1,000 KB or 1,0002 bytes 1,024 KB or 1,0242 bytes

1 GB 1,000 MB or 1,0003 bytes 1,024 MB or 1,0243 bytes

1 TB 1,000 GB or 1,0004 bytes 1,024 GB or 1,0244 bytes

1 PB 1,000 TB or 1,0005 bytes 1,024 TB or 1,0245 bytes

1 EB 1,000 PB or 1,0006 bytes 1,024 PB or 1,0246 bytes

1 block - 512 bytes

Getting helpThe Hitachi Data Systems customer support staff is available 24 hours a day,seven days a week. If you need technical support, log on to the Hitachi DataSystems Portal for contact information: https://portal.hds.com

CommentsPlease send us your comments on this document: [email protected] the document title, number, and revision, and refer to specificsection(s) and paragraph(s) whenever possible.

Thank you! (All comments become the property of Hitachi Data Systems.)

Preface xiHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

https://portal.hds.com

xii PrefaceHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

1Introduction

The Hitachi NAS Platform F (HNAS F) is network-attached storage thatconnects to a storage system via a Fibre Channel interface in order to providefile-sharing services over NFS or CIFS protocol to clients on the network.

The HNAS F can be used in conjunction with the rich variety of functionsprovided by the program products on the storage system.

This manual describes the features of and provides precautions andrestrictions regarding HNAS F, when it is used in conjunction with theprogram product functions provided by HUS VM enterprise-class storagesystems.

For details about the functions and operation of the program productssupplied with storage systems, see the appropriate User's Guide.

□ Supported Program Products

Introduction 1-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Supported Program ProductsThe file system service functionality provided by HNAS F can be used inconjunction with the program product functions supported by storagesystems connected to the system.

This section describes the program products that can be used in conjunctionwith the HNAS F. For notes and instructions on using these programs in thismanner, read from Chapter 2, Replication Functions on page 2-1 of thismanual.

Program products for using replication functions

• CCI

• ShadowImage

• TrueCopy

• Universal Replicator

Program products for using volume management functions

• Dynamic Provisioning

• Dynamic Tiering

• Encryption License Key

• Universal Volume Manager

• Volume Migration

• Volume Shredder

Program products for using resource management functions

• Storage Navigator

• LUN Manager

• Virtual LVI

• Configuration File Loader

• Virtual Partition Manager

Program products for using performance management functions

• Performance Monitor

1-2 IntroductionHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

2Replication Functions

This chapter describes the replication functions in storage systems that canbe used in a HNAS F system.

□ Overview

□ Command Control Interface

□ How to use ShadowImage

□ How to use TrueCopy

□ How to use Universal Replicator on the HNAS F

□ Operations of the Log Files

□ Commands that HNAS F provides

Replication Functions 2-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

OverviewThe following program products provide replication functions compatible withthe HNAS F

• ShadowImage

• TrueCopy


ShadowImage ShadowImage is a program product for backing up and duplicating data instorage systems. Using ShadowImage in conjunction with the HNAS F, youcan easily back up and replicate user data. ShadowImage lets you managecopies of logical units (LUs) within the same storage system. Copies of thesource LU (secondary volumes or S-VOLs ) can be created in the samestorage system without impairing the RAID redundancy of the data on thesource LU (primary volume or P-VOL ).

TrueCopy TrueCopy is a program product for backing up and replicating data in storagesystems. Using TrueCopy in conjunction with the HNAS F, you can create aremote backup or disaster recovery system for user data. TrueCopy lets youmanage copies of the logical units (LUs) across storage systems connected bya Fibre Channel interface. Copies of the source LU (secondary volumes or S-VOLs )can be created on a different storage system without impairing theRAID redundancy of the data on the source LU (primary volume or P-VOL ).

TrueCopy uses synchronous mode to transmit data written from the HNAS Fto a different storage system.

Universal Replicator Similar to TrueCopy, Universal Replicator is a program product for backing upand duplicating data in storage systems. Using Universal Replicator inconjunction with the HNAS F, you can create a remote backup or disasterrecovery system for user data. Universal Replicator allows for a reliablereplication system by temporarily storing the data to be copied in a logicalvolume called a journal volume. This minimizes the frequency of interruptionsto the copy process between volume pairs caused by restrictions imposed ondata transfer from the primary site to the secondary site.

Command Control Interface This section describes how to use CCI on the HNAS F.

2-2 Replication FunctionsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Overview of CCI Operation in HNAS FCCI operations need to be performed when using ShadowImage, TrueCopy,or Universal Replicator in a HNAS F system. When you use a HNAS F system,CCI is installed with the OS. As such, there is no need to install CCIseparately.

If you intend to use ShadowImage, TrueCopy, or Universal Replicator in yourHNAS F system, first read the Hitachi Command Control Interface (CCI) Userand Reference Guide and make sure that you have a thorough understandingof how to use CCI. Then read the relevant sections in this user's guide.

About Command DevicesTo use CCI, a command device needs to be configured in the storage system.Before starting operation, make sure that a command device is configured.

If HNAS F system operation is started without first configuring a commanddevice, configure a command device, and then restart both nodes.

For details about how to configure a command device and check the settings,see the Storage Navigator User's Guide (User Guide).

Do not set an authentication mode for the command devices used by a HNASF system.

Scope Related to CCI

CCI and failover/failback by the HNAS F system

CCI is not subject to failover or failback by the HNAS F system. Prepare thesame configuration definition file as was prepared for the main node, with theexception of the ip_address parameter in the HORCM_MON section. If a failureoccurs while CCI is running, you can start CCI on the standby node andcontinue operating ShadowImage, TrueCopy, or Universal Replicator.

If a virtual server fails over, you can continue operation by starting CCI afterthe failover.

Protect feature

When using ShadowImage, TrueCopy, or Universal Replicator in a HNAS Fsystem, you need to create and resynchronize pairs using CCI, without theHNAS F being aware of the S-VOLs. If the HNAS F recognizes the file system,Data Retention Utility automatically gives volumes "S-VOL Disable attribute".Therefore, the protect feature of CCI which prohibits the operation of volumepairs not recognized by the system is unavailable.

Executing Command Control Interface (CCI) commands

To execute CCI commands using the nasroot account, which is used forlogging in to a node or virtual server via SSH, use the sudo command asshown below.


$ sudo pairdisplay -g VG01

Example 2-1 Example of Executing the CCI 'pairdisplay' Command Usingthe sudo Command

$ ssh –2 [email protected] sudo pairdisplay -g VG01

Example 2-2 Example of Executing the 'pairdisplay' Command using aShell Script created on the SSH Client

Note:Create a shell script on the OS used by the system administrator. Fordetails about creating a shell script, see the documentation for your OS.

Special file names given from standard input to CCI commands

When passing a special file name from standard input to the raidscan,inqraid, or mkconf.sh commands, specify ls /dev/sdu*u.$ ls /dev/sdu*u | sudo inqraid –CLI

Example 2-3 Example of Executing the 'inqraid' Command Using the sudoCommand

$ ssh –2 [email protected] "ls /dev/sdu*u | sudo inqraid –CLI"

Example 2-4 Executing the 'inqraid' Command Using a Shell Script Createdon the SSH Client

/dev/sdu00u to /dev/sduFFu corresponds to /dev/enas/lu00 to /dev/enas/luFF of user LU.

Note on scanning on HNAS F ports using the raidscan command

When you use the raidscan command to scan HNAS F ports, informationabout virtual server OS disks and shared LU sometimes appears in thecommand output. However, because virtual server OS disks and shared LUcannot be specified as a P-VOL or S-VOL in ShadowImage, TrueCopy, orUniversal Replicator, you cannot define them in the HORCM_DEV section of theconfiguration definition file.

Note on the CCI configuration definition file

Immediately after HNAS F is installed, a temporary node name is used as thehost name in the CCI configuration definition file. Please confirm whether thenode name is right before starting CCI, and change manually if the nodename is wrong.

The CCI configuration definition file used by a virtual server must be createdand edited on both of the nodes on which the virtual server operates. Makesure that the host name and IP address in the configuration file areappropriate for the node where the file is located.

When the FC channel path connection mode between the node and thestorage system is Active-Active, and if you use the inqraid or mkconf.sh


commands, the ports that are paired by the alternative paths will bedisplayed inconsistently.

On CCI instance numbers

HNAS F makes the following instance numbers available.

Two instance numbers are allocated to each node, which are made availableafter the installation of HNAS F.

Table 2-1 Instance numbers allocated by default

Type Instancenumber

Service value of configurationdefinition file

Instance numbers allocated toeach node by default

16

17

20331

20332

To use three or more instances on each node, or to use instance numberswith virtual servers, you must configure an environment that can handle theadditional instance numbers. You can use a maximum of 18 instancenumbers (including the two default instance numbers) on each node.

Table 2-2 Additional instance numbers

Type Instancenumber

Service value of configurationdefinition file

Additional instance numbers(virtual servers)

Virtual-server-ID + 1000

Virtual-server-ID + 1500

30000 + instance-number

Additional instance numbers(common)

20 to 499 30000+instance number

To use additional instance numbers, use the horcsetconf command to createa CCI configuration definition file.

After configuring a command device, you can execute this command.

If a CCI configuration definition file is no longer required (with the exceptionsof the configuration instance files for instance numbers 16 and 17), delete itby using the horcunsetconf command. If the same CCI configurationdefinition file might be re-created, create a backup of the CCI configurationdefinition file by the scp command. It is also possible to confirm the alreadyallocated instance numbers by the horcconflist command.

Operation examples are described below. Follow these examples to addinstance numbers and create a CCI configuration definition file.


For creating a CCI configuration definition file using instancenumbers 16 and 20:

1. By the "ls /dev/sdu*u | sudo mkconf.sh –gg device-group-name –i16" command, add the HORCM_MON,HORCM_CMD,HORCM_DEV and HORCM_INSTdefinition to the CCI configuration definition file (instance number 16).

2. Use the sudo horcconfedit horcm16.conf command to change theHORCM_CMD definition in the CCI configuration definition file (instancenumber 16).

3. Delete the unnecessary devices from HORCM_DEV of the CCI configurationdefinition file (instance number 16).

4. By the "sudo horcsetconf –i 20" command, create a CCI configurationdefinition file with the instance number 20.

5. Use the ls /dev/sdu*u | sudo mkconf.sh –gg device-group-name –i20 command to add the HORCM_MON, HORCM_CMD, HORCM_DEV, andHORCM_INST definitions to the CCI configuration definition file (instancenumber 20).

6. Use the sudo horcconfedit horcm20.conf command to change theHORCM_CMD definition in the CCI configuration definition file (instancenumber 20).

7. Delete the unnecessary devices from HORCM_DEV of the CCI configurationdefinition file (instance number 20).

How to use ShadowImageThis section describes how to use ShadowImage on the HNAS F.

Overview of ShadowImage Operation on the HNAS FThe HNAS F system enables data consolidation on storage system andheterogeneous platform data sharing by using existing LAN environment, asfile system servers. The HNAS F makes copied volumes of original volumes inthe same storage system by using the ShadowImage function at hardwarespeed without reducing RAID redundancy. From these copied volumes usercan obtain backup of user data without interrupting current transactions. Theuser may also perform additional transactions.

CCI is used to operate ShadowImage in the HNAS F system. Because CCI isinstalled with the OS on the HNAS F system, there is no need to install CCIseparately.

Figure 2-1 Figure 2-1 Concept of ShadowImage Operation on the HNAS F onpage 2-7shows an overview of ShadowImage operation.


Figure 2-1 Concept of ShadowImage Operation on the HNAS F

The following table shows the main features provided by ShadowImage onthe HNAS F.

Table 2-3 Main ShadowImage Features

Item ShadowImage support

ShadowImage user interface CLI (Backup Restore commands and CCIcommands)

Relationship between P-VOL connectednode and S-VOL connected node

Can be used between different nodes andbetween virtual servers provided they arein the same disk array

Cascade configuration (TrueCopy-ShadowImage,

ShadowImage/ShadowImage)

Applicable

Duplicating file systems and differential-data snapshots used by the file snapshotfunctionality

Applicable

Expanding or deleting file systems duringShadowImage operation

Applicable to P-VOL. Applicable to S-VOLonly when its status is SMPL or SSUS.


Item ShadowImage support

Starting or ending to use the file snapshotfunctionality or expanding the device whichstores the differential data

If you want to continue using the filesnapshot functionality, you need to changethe configuration definition file.

Handling of the file system obtained on theS-VOL after ShadowImage operation ends

Operation can continue.

Support for NDMP-compatible backupmanagement software

No

ShadowImage is suitable for data replication when the replicated data is to beused for a different purpose than the purpose for the original data, for clusterfailure recovery, or for disaster recovery of data center utilizing TrueCopy,because it allows for the creation of duplicate file systems and the placementof volumes in a cascade configuration.

Scope of ShadowImage Operations on the HNAS F

Scope Related to ShadowImage

Volume type

Only a user LU can be defined as P-VOL or S-VOL of ShadowImage on theHNAS F system.

The OS disks and shared LU of a virtual server cannot be specified as aShadowImage P-VOL or S-VOL.

Platform which can access P-VOL or S-VOL of ShadowImage

A P-VOL or S-VOL of ShadowImage created by the HNAS F system may beaccessed by the HNAS F system or clients connected to the HNAS F through anetwork. It cannot be accessed by hosts connected through serial ports orfibre channel ports.

Filesystem which can be placed on P-VOL of ShadowImage

When Logical Volume Manager (LVM) on OS is utilized, a file system thatconsist of 129 or more LUs each, but are still not used by the file snapshotfunctionality, cannot be placed on P-VOL of ShadowImage.

File systems used by the file snapshot functionality that consist of 129 ormore LUs each, including the file system and the differential devices, cannotbe placed on P-VOL of ShadowImage.


Scope Related to Backup Restore

Limitations on the functionality of the backup managementsoftware

If you resynchronize a ShadowImage pair defined by Volume Replicationfunction after performing a backup of the S-VOL by using the backupmanagement software, a full backup will be acquired next time.

Scope Related to the file snapshot functionality

The file snapshot functionality

When you copy a file system managed by the file snapshot functionality usingShadowImage, you must copy both the LUs that constitute the file systemand the LUs which constitute the file system for differential devices. If youcopy only the LUs which constitute the file system, you will not be able toconnect them to the HNAS F system on the remote site.

The setting of an automatic creation schedule for the file snapshotfunctionality is not copied. The differential-data snapshot is not mounted atthe copy destination.

Prerequisites for Using ShadowImage on the HNAS F

Hardware Prerequisites

You need a workstation or PC to log in to the HNAS F system using secureshell (SSH) in addition to the prerequisite hardware for ShadowImage, CCI,and the HNAS F system described in the following guides:

Manuals related to HNAS F

¢ Installation and Configuration Guide

Manuals related to storage systems

¢ Hitachi ShadowImage User's Guide (User Guide)

¢ Hitachi Command Control Interface (CCI) User and Reference Guide

¢ Hitachi TrueCopy User and Reference Guide (if TrueCopy is used inconjunction with forming a cascaded configuration)

¢ Universal Replicator User's Guide (User Guide) (if Universal Replicatoris used in a cascaded configuration)

Software Prerequisites

To use the HNAS F system, the program products in the HNAS F must beproperly installed and have valid licenses.

In addition, to use ShadowImage in a HNAS F system, the following programproducts must be installed in the storage system to which HNAS F isconnected, and their licenses must be valid:


• ShadowImage

• TrueCopy (required for cascaded configurations between TrueCopy)

• Universal Replicator (if Universal Replicator is used in a cascadedconfiguration)

Notes on Operating the Volume Replication Function

Mounting the NFS client for a file system whose data is backed uponline

If you perform online backup for a file system accessed by the NFS client byvolume replication, you must specify the NFS version 3 before mounting a filesystem on the NFS client. If you specify the NFS version 2, you must specifythe hard option before mounting a file system on the NFS client.

Limitation for operation of ShadowImage due to cluster, node,resource group status

When a cluster is not configured, the cluster is stopped, the nodes arestopped, the resource groups are offline, or the resource groups are split,connecting the device file or creating and mounting a file system is restricted.Due to these restrictions, the following operations performed in the targetShadowImage operation will also end in error. You therefore should notoperate the cluster, nodes, and resource groups during the ShadowImageoperation. Should any problems occur with the cluster, nodes, or resourcegroups, fix them immediately.

• Unmount and mount the source file system during the splitting of aShadowImage pair volume.

• Connect the target file system to the node.

• Unmount and delete the target file system before the ShadowImage pairis resynchronized.

Notes on occurrence of a failover during operation on node

During the operation on node, to continue the operation at the failoverdestination when a failover occurs, execute the command making the remotehost specify the virtual IP address. When you connect the S-VOL file systemof the copy destination, execute the horcimport or horcvmimport commandby adding the -r option and specifying a resource group name.

Notes on changing system configuration during the ShadowImageoperations

When one of following system configuration changes is performed, the CCIconfiguration definition files on the node must be modified:

• Changing the fixed IP addresses of the node

• Expanding or deleting the source file system


• Defining, expanding, or deleting the file snapshot functionalitydifferential-data storage device to or from the source file system

When the host name is specified in the CCI configuration definition file andyou change the following system configurations, you must change theconfiguration definition file:

• Editing the "/etc/hosts" file (when resolving the host name using the "/etc/hosts" file)

• Changing registration information on NIS server, or changing a setting forNIS server (when resolving the host name using NIS)

• Changing registration information on DNS server, or changing a settingfor DNS server (when resolving the host name using DNS)

• Changing a name of node or a name of host

Notes on using external volumes

When using external volumes for P-VOLs or S-VOLs, do not perform the quickrestore pairresync operation. For example, if the P-VOL is an internal volumeand the S-VOL is an external volume, the quick restore pairresync operationswaps the contents of the P-VOL and S-VOL. Consequently, the P-VOLbecomes an external volume and the S-VOL becomes an internal volume.

See Hitachi ShadowImage User's Guide (User Guide) for further informationon backward pairresync operation.

In case of periodic maintenance of external volumes, you must delete allShadowImage pairs once.

Notes on splitting ShadowImage pairs by online backup

After the horcfreeze command is executed, if it takes long time until thehorcunfreeze command is executed, the timeout may occur in some client.In addition, if the file snapshot functionality is used on the copy source filesystem, the timeout tends to occur since the horcfreeze command takeslong time to be executed.

Confirming whether the access to the file system is suspended

After you execute the horcfreeze command and the horcunfreezecommand, you can check whether the access from the client to the filesystem is suspended or not by using the fsctl command.

Notes on using volume replication with virtual servers

Operations that take place when CCI is not running, such as creating, editing,and deleting the CCI configuration definition file and log file, must beperformed on a node. The table below lists the commands that cannot beperformed on a virtual server. These commands must be executed on a node.


Table 2-4 Commands that Cannot Be Used on a Virtual Server

Command name Description

horclogremove Deletes log files and trace files stored in the CCI log directoriesnasroot/logn (where n is the instance number)

horcconflist Displays a list of instance numbers that are in use.

horcsetconf Provides a model CCI configuration definition file and configuresthe CCI log directory for a specific instance number.

horcunsetconf Deletes the CCI configuration definition file and CCI log directoryfor a specific instance number.

horcconfedit Assists the user in editing the CCI configuration definition file.

mkconf.sh Defines the CCI configuration definition file.

CCI configuration definition files must be created and edited on both of thenodes where the virtual server operates. Make sure that the host name andIP address in the configuration file are appropriate for the node where the fileis located. The instance numbers used by CCI on a virtual server are virtual-server-ID + 1000 and virtual-server-ID + 1500, and the numbers 20 to 499.You can use the vnasinfo command to check the virtual server ID.

When you execute the horcmstart.sh command to start CCI on a virtualserver, CCI starts on the node where the virtual server is running. If thevirtual server fails over, you can continue operation by starting CCI after thefailover.

Different administrators are responsible for creating the CCI configurationdefinition file on the node and operating CCI on the virtual server. To make iteasier for both parties to be notified of pertinent events, we recommend thatyou incorporate the file system name into the volume group name. (Example:vg_fs01_001)

Table 2-5 Starting CCI Operation on a Virtual Server

StepLocation

OperationNode Virtual server

1 -- Y Notify the node administrator of the filesystem and volume to be used.

2 Y -- Determine the instance number to be usedby the virtual server.

3 Y -- Use the mkconf.sh command or someother means to create and edit CCIconfiguration definition files on both nodes.Make sure that the host names and IPaddresses in the files correspond to theirrespective nodes. For the IP address of thelocal node, use the IP address of the nodewhere you edited the CCI configurationdefinition file.


StepLocation


4 Y -- Notify the virtual server administrator ofthe instance number and volume groupname.

5 -- Y Start a CCI instance using the instancenumber reported by the nodeadministrator, and begin operation.

Legend:Y: Perform the operation.--: Do not perform the operation.

Table 2-6 Procedure after Restarting a Virtual Server

StepLocation


1 -- Y Start a CCI instance. Ignore any errors thatindicate an instance is already running.

2 -- Y Continue operation.


Table 2-7 Ending CCI Operation on a Virtual Server

StepLocation

OperationNote Virtual server

1 -- Y Notify the node administrator that youintend to terminate CCI.

2 Y -- On both nodes, stop the CCI instance thatwas being used on the virtual server.

3 Y -- From both nodes, delete the configurationdefinition files for the CCI instance that wasbeing used on the virtual server.

4 Y -- Delete the virtual server.



Notes on file systems managed by the file snapshot functionality

The following settings and status of a copy source file system managed bythe file snapshot functionality are copied to a target file system.

• Warning threshold

• Operation threshold

• Overflow prevention

• Status of the differential data storage device

If a ShadowImage pair is split and connected to the node while thedifferential data storage device in the copy source file system does not havesufficient capacity, certain measures must be taken in the copy source filesystem to solve the lack of the capacity.

In splitting a ShadowImage pair, confirm that the differential data storagedevice in the copy source file system is in the normal status.

Note on file systems that support single-instancing

The single-instancing setting is not copied over to copy-destination filesystems. If the single-instancing setting is enabled for a copy-source filesystem, after splitting a ShadowImage pair and connecting to a node, use thefsedit command to enable the single-instancing setting for the copy-destination file system. Also, set the policy for duplicate file capacityreduction.

Before splitting a ShadowImage pair, make sure that the duplicate filecapacity reduction policy has not yet been executed on the copy source filesystem. If a ShadowImage pair is split during policy execution, resynchronizethe ShadowImage pair and perform its subsequent operations again.

Notes on tiered file systems

If the copy-source file system is a tiered file system, the LUs in all the filesystems that make up the tier must be assigned the same device group namein the configuration definition file.

To connect the copy-source file system to a node or a virtual server, specifythe --tier1 and --tier2 options for the horcimport command or thehorcvmimport command if the copy file system is a tiered file system.Otherwise, specify the -d option.

If the tiered file system is already connected to a node or a virtual server,you need to set up a tier policy schedule. For details, see the Hitachi NASPlatform F1000 Series Cluster Administrator's Guide.

Notes on pair recovery in using the replication functions

If the HNAS F system creates a file system, the S-VOL Disable attribute ofData Retention Utility is given to the LUs of the storage system. This is forpreventing the HNAS F volumes from becoming S-VOLs.


If the copy-destination file system is connected to the node, a pair recoverycannot be performed because the S-VOL Disable attribute is assigned. Beforeperforming a pair recovery, execute the horcexport command to separatethe copy-destination file system from the node. Separating the copy-destination file system from the node releases the S-VOL Disable attribute,which makes it possible to perform a pair recovery.

Notes on dynamic recognition in the FC path

HNAS F automatically recognizes LUs connected to FC paths.

A user LU (device file number) is determined when the device file to be usedfor the copy target of the file system is reserved by using the horcvmdefinecommand but, if the OS is rebooted without having the device file numberdetermined, the device file number might change from the one before thereboot. In the following cases, immediately reserve the device file to be usedfor the copy target of the file system by using the horcvmdefine command sothat the device file number will not be changed.

• In starting to use the replication functions.

• In case the file system is deleted and then the LU of the deleted filesystem is to be used as the S-VOL.

If the OS is rebooted before reserving the device file to be used for the copytarget of the file system, find out the device file number and the LU numberthat constitute the file system by using the horcdevlist command.

Note on WORM file systems

A node or virtual server cannot be connected to if ShadowImage is used tocopy a WORM file system.

If the file system was encrypted by using the HNAS F functionality

If the local data encryption functionality is being used, the copy-destinationfile system can be connected to node only when the copy-source file systemand the copy-destination file system are in the same cluster. If the filesystems are in different clusters, the copy-destination file system cannot beconnected to the node.

Preparing for ShadowImage Operation

Preparing for ShadowImage Volume Pair Operation

Prepare the ShadowImage volume pair. For details about the preparationrequired on the storage system side, refer to the Hitachi ShadowImage User'sGuide (User Guide).

Registering Public Key Used for SSH

Before issuing the commands described in this document, the public key usedfor SSH needs to be registered in the node or virtual server and also in the


node or virtual server which will use the S-VOL. You can register the publickey from the Access Protocol Configuration window, in the Add Public Keypage.

Configuration Settings of CCI

Logging in to a node via SSH

Using the nasroot account via SSH, log in both the node in which theShadowImage P-VOL is connected, and the node which will use the S-VOL.(For information about login, refer to the appropriate documentation of theSSH communication software)

Setting the environment of instance numbers to be used

For using the instance numbers allocated by default, the usage environmentis already configured, and no operation is required here.

For using additional instance numbers, the environment for the instancenumbers to be used must be configured by using the horcsetconf command.$ sudo horcsetconf -i instance number

Example 2-5 Configuring the usage environment for additional instancenumbers

The instance numbers which are already set can be confirmed by thehorcconflist command. By this command, search for instance numbers notused.$ sudo horcconflistinstance node number or virtual server name 16 node 0(D000000000), node 1(D000000001) 17 node 0(D000000000), node 1(D000000001) 499 node 0(D000000000), node 1(D000000001)

Example 2-6 Confirming already set instance numbers

The usage environment of additional instance numbers can be deleted ifnecessary by the horcunsetconf command. Note that the usage environmentof the instance numbers allocated by default cannot be deleted.$ sudo horcunsetconf -i instance number

Example 2-7 Deleting the usage environment of additional instancenumbers

Configuring the CCI Configuration Definition Files

To control a ShadowImage pairs using CCI, you must first define theShadowImage pair using the CCI configuration definition file.

When HVFP is installed, a template is provided for CCI configuration definitionfiles.


The template of the CCI configuration definition file:/home/nasroot/horcm<instance-number>.conf For nodes, <instance-number> is 16 or 17. Additional instance numbers are from 20 to 499.Use the CCI horcconfedit and mkconf.sh commands to define theHORCM_MON, HORCM_CMD, HORCM_DEV, and HORCM_INST sections in thetemplate. Then edit the file and create the CCI configuration definition file.

You must perform these operations in both nodes. You therefore need toprepare a total of four CCI configuration files, assuming two instances pernode.

In a HNAS F, you can define one or two CCI instances per one node. Tooperate only the ShadowImage pairs in CCI, you need one instance. Tooperate the cascaded pairs in CCI, you need two instances.

In the following section, we explain how to create a CCI configurationdefinition file using the HNAS F system with LUs as an example.


Figure 2-2 Example of Configuration of Pair LU

Note:To enable two node to communicate over a LAN, you must define thenetwork configuration.

Defining configuration definition files by using the CCI mkconf.shcommand

Add the HORCM_MON, HORCM_CMD, HORCM_DEV, and HORCM_INSTsections to the CCI configuration definition file template by using the CCImkconf.sh command.$ ls /dev/sdu*u | sudo mkconf.sh -gg <device-group-name> -i 16

Example 2-8 Defining configuration definition files (for instance number16)


Note:Be sure to specify the -gg option for the mkconf.sh command. Specifyingthe -gg option assigns LU numbers allocated to host groups to pairs whenthe pairs are defined. When the -gg option is not specified, data is copiedto an LU other than the required one because the pair cannot be definedby using an LU number allocated to the host group. In addition, do notspecify the -a option because the command device path of the HORCM_CMDsection is defined by the mkconf.sh command. For details, see the HitachiCommand Control Interface (CCI) User and Reference Guide.

To execute the "mkconf.sh" command, the file system to be paired must becreated. The LU configuration (number and size) in the file system to bepaired needs to be exactly the same in both the P-VOL and the S-VOL.

To create the correct CCI configuration definition file, it is recommended thatyou create a file system to be temporarily paired before executing the"mkconf.sh" command. After creating the CCI configuration definition fileusing the "mkconf.sh" command, you can continue using the P-VOL filesystem is connected. Alternatively, you may also delete it and create a filesystem with the same configuration using the same LU according to theprocedure described in File System Creation on P-VOL of ShadowImage onpage 2-29. However, you must delete the temporarily created file systemfor S-VOL before creating a pair.

We will explain how to create a CCI configuration definition file in the node0to which the P-VOL is connected in the sample configuration below. You mustcreate CCI configuration definition files for the other three nodes using thesame procedure.$ ls /dev/sdu*u | sudo mkconf.sh -gg VG -i 16starting HORCM inst 16HORCM inst 16 starts successfully.HORCM Shutdown inst 16 !!!A CONFIG file was successfully completed.starting HORCM inst 16HORCM inst 16 starts successfully.DEVICE_FILE Group PairVol PORT TARG LUN M SERIAL LDEV/dev/sdu00u VG VG_000 CL1-A-1 0 0 - 62486 70/dev/sdu01u VG VG_001 CL1-A-1 0 1 - 62486 18 : dev/sdu10u VG VG_010 CL1-A-1 0 10 - 62486 10/dev/sdu11u VG VG_011 CL1-A-1 0 11 - 62486 64/dev/sdu12u VG VG_012 CL1-A-1 0 12 - 62486 12/dev/sdu13u VG VG_013 CL1-A-1 0 13 - 62486 66/dev/sdu14u VG VG_014 CL1-A-1 0 14 - 62486 14/dev/sdu15u VG VG_015 CL1-A-1 0 15 - 62486 68/dev/sdu16u VG VG_016 CL1-A-1 0 16 - 62486


16HORCM Shutdown inst 16 !!!Please check '/home/nasroot/horcm16.conf','/home/nasroot/log16/curlog/horcm_*.log', and modify 'ip_address & service'.

Example 2-9 Example of "mkconf.sh" Execution Result (for instancenumber 16)

Then, use the horcconfedit command to change the HORCM_CMD definition inthe CCI configuration definition file to a format that does not depend ondevice file changes (\\.\CMD-<serial-number>:/dev/sd).$ sudo horcconfedit horcm16.conf

Example 2-10 Changing the HORCM_CMD definition in configurationdefinition file (for instance number 16)

Editing the CCI configuration definition file

The following table shows the values that are specified for the items includedin the CCI configuration definition file in the HNAS F system.

Table 2-8 Configuration Definition File Settings (HORCM_MON) andSpecified Values in the HNAS F system

Section Name Item Specified Values in HNAS F system

HORCM_MON ip_address Fixed IP address of the node

service Specify any of the following:

• 20331 (In the case of a node. If the instancenumber is 16.)


• 31032 to 31254 or 31532 to 31754 (In the case of avirtual server. Instance-number +30000)

• 30020 to 30499 (Common. Instance number+30000)

Note:A host name can be specified instead of a fixed IP address if the IPaddress and the corresponding host name are registered into /etc/hosts, on an NIS server, or a DNS server. You can edit /etc/hosts onthe Edit System File page on the Network & System Configurationwindow. You can set the NIS server or DNS server information on theDNS, NIS, LDAP Setup page on the Network & System Configurationwindow.

Using the information in Table 2-8 Configuration Definition File Settings(HORCM_MON) and Specified Values in the HNAS F system on page 2-20,change the entries for service and ip_address to appropriate values.

Change the entries poll and timeout to the appropriate values according tohardware requirements.


HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.78.51 20331 1000 3000

HORCM_CMD#dev_name dev_name dev_name#UnitID 0 (Serial# 62486)\\.\CMD-62486:/dev/sd

HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu00u SER = 62486 LDEV = 70 [ FIBRE FCTBL = 3 ]VG VG_000 CL1-A-1 0 0# /dev/sdu01u SER = 62486 LDEV = 18 [ FIBRE FCTBL = 3 ]VG VG_001 CL1-A-1 0 1 : :# /dev/sdu13u SER = 62486 LDEV = 66 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 13# /dev/sdu14u SER = 62486 LDEV = 14 [ FIBRE FCTBL = 3 ]VG VG_014 CL1-A-1 0 14# /dev/sdu15u SER = 62486 LDEV = 68 [ FIBRE FCTBL = 3 ]VG VG_015 CL1-A-1 0 15# /dev/sdu16u SER = 62486 LDEV = 16 [ FIBRE FCTBL = 3 ]VG VG_016 CL1-A-1 0 16

HORCM_INST#dev_group ip_address serviceVG 127.0.0.1 52323

Example 2-11 Example of CCI Configuration Definition File (for instancenumber 16)

Next, at HORCM_DEV section, remove all unnecessary LU entry (line) otherthan those to be managed by CCI.

LU which makes up a file system and its LDEV ID can be obtained by usingthe following command.$ sudo horcdevlist | grep ':filesystem-name'ForExample 2-12 Example of an Explorer Command of LU which Makes up aFile System Sample on page 2-21, LU 11, 12, 13 shown in left make up afile system "sample", and their LDEV ID are 64, 12, 66 shown on the thirdcolumn to the left.$ sudo horcdevlist | grep ':sample$'11 62486 64 OPEN-V 3.906GB -- -- - Normal File:sample12 62486 12 OPEN-V 3.906GB -- -- - Normal File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal File:sample

Example 2-12 Example of an Explorer Command of LU which Makes up aFile System Sample


Next, modify the device file name and the device name, which are to behandled by CCI. There are some restrictions on modifying the device filename or device name.

• Volumes on both P-VOL and S-VOL must be assigned the same devicegroup name and same device name.

• Volumes making up one file system must be assigned the same devicegroup name.When the file snapshot functionality is utilized, volumes forfile system and differential-data storage device must be assigned thesame device group name.

• For a tiered file system, all of the LUs used to configure the tier (includingdifferential-data storage devices) must be assigned the same devicegroup name.



HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu11u SER = 62486 LDEV = 64 [ FIBRE FCTBL = 3 ]VG VG_011 CL1-A-1 0 11# /dev/sdu12u SER = 62486 LDEV = 12 [ FIBRE FCTBL = 3 ]VG VG_012 CL1-A-1 0 12# /dev/sdu13u SER = 62486 LDEV = 66 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 13


Example 2-13 Example of CCI Configuration Definition File - 2 (forinstance number 16)

For one LU, two lines of the HORCM_DEV section information are output: thecomment line starting with a # and the definition line right below it. Use the /dev/sdu**u (where ** is the LU number) and LDEV = ** (where ** is theLDEV number) text output in the comment lines to identify the necessaryentries.

Next, as an item to be set in the HORCM_INST section, specify the ip addressof the S-VOL. As a preparation for failover, specify the ip_address of theinstance for each node.

Table 2-9 Sections and Items on HORCM_INST Section of CCIConfiguration Definition File and Required Values for HNAS F system

Section Item Values for HNAS F system System

HORCM_INST ip_address Fixed IP address of the other ShadowImage node


Section Item Values for HNAS F system System

service Specify any of the following:




• 30020 to 30499 (Common. Instance number+30000)





HORCM_INST#dev_group ip_address serviceVG 123.45.78.51 20332VG 123.45.78.52 20332

Example 2-14 Example of CCI Configuration Definition File - 3 (forinstance number 16)

Checking the contents of the CCI configuration definition file

By combining the following commands, you can check whether an appropriateLU is specified in the HORCM_DEV section in the CCI configuration definitionfile.

Start CCI on both the node or virtual server connected to the ShadowImageP-VOL and the node or virtual server that will use the S-VOL.


$ sudo horcsetenv HORCMINST 16 (for instance number 16) OR sudo horcsetenv HORCMINST 17 (for instance number 17)$ sudo horcsetenv HORCC_MRCF 1IF THESE SETTINGS ARE DONE THROUGH SSH, LOG OUT ONCE THEN LOG IN AGAIN TO VALIDATE THE MODIFICATIONS.$ sudo horcmstart.sh

Example 2-15 Procedure for Starting CCI

To obtain the LDEV ID of the LU specified in the HORCM_DEV section, executethe pairdisplay command on the node or virtual server connected to theShadowImage P-VOL or the node or virtual server connected to the volumethat either is or will be used for the ShadowImage S-VOL.$ sudo pairdisplay -g device-group-name

Example 2-16 How to Check LDEV Number of LUs Specified in theHORCM_DEV Section

You can check the device file number and LDEV ID that make up the filesystem by executing the horcdevlist command on the node or virtual serverconnected to the ShadowImage P-VOL. Verify the result shown on Example2-16 How to Check LDEV Number of LUs Specified in the HORCM_DEVSection on page 2-24.$ sudo horcdevlist | grep ':sample$'11 62486 64 OPEN-V 3.906GB -- -- - Normal File:sample12 62486 12 OPEN-V 3.906GB -- -- - Normal File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal File:sample

Example 2-17 How to Check Device File Numbers and LDEV Numbers of P-VOL

You can also check the device file numbers and the LDEV numbers availablefor a ShadowImage S-VOL by executing the horcdevlist command on thenode or virtual server connected to the ShadowImage S-VOL. Verify theresult shown in Example 2-16 How to Check LDEV Number of LUs Specified inthe HORCM_DEV Section on page 2-24$ sudo horcdevlist | grep ' Free$'21 62486 74 OPEN-V 3.906GB -- -- - Normal Free22 62486 22 OPEN-V 3.906GB -- -- - Normal Free23 62486 76 OPEN-V 3.906GB -- -- - Normal Free

Example 2-18 How to Check Device File Numbers and LDEV Numbers forUnused Device Files Which Can be S-VOLs

When specifying the port name in the HORCM_DEV section of the CCIconfiguration definition file, use the name of the storage system FibreChannel port connected to the node or virtual server.

After the CCI definition files has been validated, stop CCI on the nodesconnected to the ShadowImage P-VOL and S-VOL.$ sudo horcmshutdown.sh

Example 2-19 Procedure for Stopping CCI


To save CCI configuration definition files, save the system settingsinformation. For details about saving system settings information, see theHitachi NAS Platform F1000 Series Cluster Administrator's Guide.

An example of the configuration definition file for a tiered system

An example of the configuration definition file for a tiered system is providedbelow.

Figure 2-3 Example of a Tiered File System$ sudo horcdevlist | grep ':sample$'10 62486 10 OPEN-V 3.906GB -- -- - Normal Tier1,File:sample11 62486 64 OPEN-V 3.906GB -- -- - Normal Tier1,File:sample12 62486 12 OPEN-V 3.906GB -- -- - Normal Tier2,File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal Tier2,File:sample

Example 2-20 How To Check the Device File Number and the LU Number ofa P-VOL



HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10u SER = 62486 LDEV =10 [ FIBRE FCTBL = 3 ]VG VG_010 CL1-A-1 0 10# /dev/sdu11u SER = 62486 LDEV =64 [ FIBRE FCTBL = 3 ]


VG VG_011 CL1-A-1 0 11# /dev/sdu12u SER = 62486 LDEV =12 [ FIBRE FCTBL = 3 ]VG VG_012 CL1-A-1 0 12# /dev/sdu13u SER = 62486 LDEV =66 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 13


Example 2-21 Example of the CCI Configuration Definition File (for P-VOLs)



HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu20u SER = 62486 LDEV = 20 [ FIBRE FCTBL = 3 ]VG VG_010 CL1-A-1 0 20# /dev/sdu21u SER = 62486 LDEV = 74 [ FIBRE FCTBL = 3 ]VG VG_011 CL1-A-1 0 21# /dev/sdu22u SER = 62486 LDEV = 22 [ FIBRE FCTBL = 3 ]VG VG_012 CL1-A-1 0 22# /dev/sdu23u SER = 62486 LDEV = 76 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 23


Example 2-22 Example of a CCI Configuration Definition File (for S-VOLs)

Cascaded Configuration

Pairs of TrueCopy and ShadowImage, or multiple pairs of ShadowImage canbe cascaded on the HNAS F system. On cascaded configuration of TrueCopypair and ShadowImage pair, ShadowImage can periodically back up S-VOLfile system of TrueCopy, which is copied from P-VOL of TrueCopy, to preparesome kind of disasters such as S-VOL of TrueCopy cannot restore the filesystem. On cascaded configuration of multiple ShadowImage pair file system,up to 9 S-VOLs can be created to 1 P-VOL.

Figure 2-3 shows an example of cascaded configuration of multipleShadowImage pair file system. For cascaded configuration of TrueCopy pairfile system and ShadowImage pair file system, please refer to HitachiTrueCopy User and Reference Guide.


Figure 2-4 Example of a Cascaded Configuration of ShadowImage Pairs

Preparing following CCI configuration definition file, cascaded multipleShadowImage pairs can be managed by CCI.HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.80.51 20331 1000 3000


HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu20u SER = 62486 LDEV =20 [ FIBRE FCTBL = 3 ]VG_SI2 VG_020 CL1-A-1 0 20 0# /dev/sdu20u SER = 62486 LDEV =20 [ FIBRE FCTBL = 3 ]VG_SI2 VG_021 CL1-A-1 0 20 1

HORCM_INST#dev_group ip_address serviceVG_SI2 123.45.80.51 20332VG_SI2 123.45.80.115 20332

Example 2-23 Example of CCI Configuration Definition File on the Cluster(Instance 16)




HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu30u SER =62486 LDEV =30 [ FIBRE FCTBL = 3 ]VG_SI2 VG_020 CL1-A-1 0 30 0# /dev/sdu31u SER =62486 LDEV =31 [ FIBRE FCTBL = 3 ]VG_SI2 VG_021 CL1-A-1 0 31 0

HORCM_INST#dev_group ip_address serviceVG_SI2 123.45.80.51 20331VG_SI2 123.45.80.115 20331

Example 2-24 Example of CCI Configuration Definition File of Cluster(Instance 17)

Setting the CCI User Environmental Variables

The HNAS F system allows you to set the following user environmentalvariables.

• HORCMINST

• HORCC_MRCF

• HORCC_SPLT

• HORCC_RSYN

• HORCC_REST

Modify HORCMINST, HORCC_MRCF, HORCC_SPLT, HORCC_RSYN, and HORCC_RESTaccording to following procedure to meet current system configuration. Thismust be done on the nodes or virtual servers connected to the ShadowImageP-VOL and the volume that will be the S-VOL.

1. Set environmental variable for CCI instance(s).sudo horcsetenv HORCMINST 16 (in case of instance number 16)

or sudo horcsetenv HORCMINST 17 (in case of instance number 17)

2. Set environmental variable for CCI HOMRCF command as "use asShadowImage".sudo horcsetenv HORCC_MRCF 1

3. Set environmental variable HORCC_SPLT, HORCC_RSYN, and HORCC_REST asneeded.sudo horcsetenv <environmental variable name> <value>In case of setting "NORMAL" to HORCC_SPLT is as follows.sudo horcsetenv HORCC_SPLT NORMAL


The following table shows the value that can be set as each environmentvariable and the operation when setting up the value. If you do not setthese environment variables, each operation will be performed accordingto the settings of the storage system. See Hitachi ShadowImage User'sGuide (User Guide) for further information on each operation.

Table 2-10 Value that Can be Set as CCI User Environment Variableand the Operation When Setting Up the Value

EnvironmentVariable Name Value Operation

HORCC_SPLT NORMAL Pair split operation performs as steady split.

QUICK Pair split operation performs as quick split.

HORCC_RSYN NORMAL Pairresync operation performs as normal copy.

QUICK Pairresync operation performs as quick resync.

HORCC_REST NORMAL Backward pairresync operation performs asreverse copy.

QUICK Backward pairresync operation performs asquick restore.

4. If procedure 1., 2., or 3 are done through SSH, log out once and log inagain to validate these settings.exitssh nasroot@<fixed ip address or virtual IP address>

To confirm the defined variables, type the following command.$ sudo horcprintenvThe following table shows the settings of the environment variables just afterHNAS F is installed.

Table 2-11 Settings of Environment Variables just after the Installation ofHNAS F

Environment Variables Setting

HORCMINST 16 (for a node)

HORCC_MRCF No setting.

File System Creation on P-VOL of ShadowImage

Create file system through the "Create New File System" window of FileServices Manager, or using the fscreate command on P-VOL ofShadowImage. Without creating file system on P-VOL of ShadowImage, S-VOL of ShadowImage cannot be accessed from the node or virtual server towhich S-VOL is connected even volume pair is successfully created and split.


Overview of Using ShadowImageThis section describes the overview of ShadowImage operations, CommandControl Interface (CCI) commands, and the commands provided by the HNASF for using ShadowImage for volume replication.

We describe only the arguments of the CCI commands, which are required forthe basic ShadowImage operations. For other arguments, please refer toHitachi Command Control Interface (CCI) User and Reference Guide. For thecommands provided by the HNAS F products, see Commands that HNAS Fprovides on page 2-132.

For node operation examples in this section, instance numbers 16 and 17 areused. If you are using additional instance numbers, replace 16 and 17 in theexamples with those numbers.

This chapter describes the procedure for backup of file systems. Overview ofthe operation and the corresponding sections are shown in the figure below.When the pair is in the PSUS state, the S-VOL can be accessed from the nodeor virtual server connected to the S-VOL.

Figure 2-5 Overview of the Operations (and the Related DocumentSections)

Starting ShadowImage Operations and Creating ShadowImage Pair

If the ShadowImage S-VOL contains a file system, you need to delete the filesystem using the fsdelete command in File Services Manager before startingthe ShadowImage operation. If the ShadowImage S-VOL has a device thatstores the differential data of the file snapshot functionality, you need torelease the device using the syncstop command before starting theShadowImage operation.

To start ShadowImage operation and to create ShadowImage volume pair:

1. Start CCI from the nodes or virtual servers connected to the P-VOL andS-VOL.


sudo horcmstart.sh (1-instance configuration)

or sudo horcmstart.sh 16 17 (2-instance configuration)

2. In the node or virtual server to which the S-VOL is connected, reserve adevice file used as a target file system.$ sudo horcvmdefine -d Device file number,...

3. Create a ShadowImage volume pair in the node or virtual server to whichthe P-VOL is connected.$ sudo paircreate {-g group-name|-d volume-name} -vl

4. Check if the ShadowImage volume has been created in the node or virtualserver to which the P-VOL is connected.$ sudo pairvolchk {-g group-name |-d volume-name }Note:

You can also use the pairevtwait command, which waits for thevolumes to be paired.

a. pairvolchk : Volstat is P-VOL.[status = COPY] => Creating

b. pairvolchk : Volstat is P-VOL.[status = PAIR] => Created

Splitting a ShadowImage Volume Pair

The method for splitting a ShadowImage volume pair depends on whetheryou perform an offline backup or an online backup. For an offline backup, theP-VOL is unmounted to split the pair. For an online backup, instead ofunmounting the P-VOL, updates to the file system are temporarily stopped tosplit the pair.

In the offline backup method, a ShadowImage volume pair is split after NFSshares and CIFS shares are deleted and accesses from the client arecompletely stopped. An I/O error is reported to the application if the userdeletes NFS/CIFS shares while the application is writing data to P-VOL, or ifthe application tries to write data to the P-VOL after the user deletes theshares. As a result, the application is able to see which application's data isreflected to the ShadowImage volume pair. This method can be used foralmost all applications.

With online backup, the ShadowImage volume pair is split without deletingthe NFS/CIFS shares. Since an I/O error is not reported to the applicationthat is writing data to the P-VOL, the application cannot see how much data isreflected to the S-VOL. Therefore, this method can be used for only theapplications that are able to keep track of the progress of writing data usingjournal files etc.

ShadowImage Volume Pair Split by Offline Backup

When the file system is not the one operated by the file snapshotfunctionality or when the differential-data snapshot is not publishedin the shared file system even though the file system is operated bythe file snapshot functionality


1. Stop the program which accesses the P-VOL in the node or virtual serverto which the P-VOL is connected. Then, unmount the NFS shares on theclient side.

2. In the node or virtual server to which the P-VOL is connected, delete theNFS/CIFS shares using the nfsdelete command and the cifsdeletecommand in File Services Manager, and unmount all file systems in thepair using the fsumount command.

3. In the node or virtual server to which the P-VOL is connected, stopoperations from the file snapshot functionality to the P-VOL.$ sudo horcfreeze -f source-file-system-name

4. In the node or virtual server to which the P-VOL is connected, split theShadowImage volume pair.$ sudo pairsplit {-g group-name |-d volume-name }

5. In the node or virtual server to which the P-VOL is connected, confirmthat the ShadowImage volume is split.$ sudo pairvolchk {-g group-name |-d volume-name }Note:

You can also use the pairevtwait command, which waits for the pairto be split (PSUS).

a. pairvolchk : Volstat is P-VOL.[status = COPY] => Splitting

b. pairvolchk : Volstat is P-VOL.[status = PSUS] => Split

6. In the node or virtual server to which the P-VOL is connected, permit thestopped operations from the file snapshot functionality to the P-VOL.$ sudo horcunfreeze -f source-file-system-name

7. In the node or virtual server to which the P-VOL is connected, mount theP-VOL using the fsmount command in File Services Manager and set theNFS/CIFS shares using the nfscreate and the cifscreate command.

8. In the node or virtual server to which the P-VOL is connected, mount theNFS shares of the P-VOL on the client side. Then, restart the programwhich accesses the P-VOL.

9. In the node or virtual server to which the S-VOL is connected, connect thetarget file system to the node or virtual server.

For a non-LVM and non-tiered file systemsudo horcimport -f target-file-system-name -d device-file-number

For a non-LVM and tiered file systemsudo horcimport -f target-file-system-name --tier1 device-file-number --tier2 device-file-number

For a LVM and non-tiered file systemsudo horcvmimport -f target-file-system-name -d device-file-number,...

For a LVM and tiered file systemsudo horcvmimport -f target-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...


10. In the node or virtual server to which the S-VOL is connected, mount theS-VOL using the fsmount command. After that, set the NFS/CIFS sharesusing the nfscreate command and the cifscreate command.

Note:In the file snapshot functionality target file system, mount thedifferential-data snapshot using the syncmount command ifnecessary, and set the NFS/CIFS shares using the nfscreatecommand and the cifscreate command.

11. In the node or virtual server to which the S-VOL is connected, start theprogram which accesses the S-VOL.

When the file system is operated by the file snapshot functionalityand the differential-data snapshot is published in the shared filesystem

1. In the node or virtual server to which the P-VOL is connected, stop theprogram which accesses the P-VOL and un-mount the NFS shares at aclient-side.

2. In the node or virtual server to which the P-VOL is connected, un-mountthe differential-data snapshot using the syncumount command, delete theNFS/CIFS shares using the nfsdelete command and the cifsdeletecommand in File Services Manager, and un-mount all file systems in thepair using the fsumount command.

3. In the node or virtual server to which the P-VOL is connected, stopoperations from the file snapshot functionality to the P-VOL.$ sudo horcfreeze -f source-file-system-name

4. In the node or virtual server to which the P-VOL is connected, split theShadowImage volume pair.$ sudo pairsplit {-g group-name | -d volume-name }

5. In the node or virtual server to which the P-VOL is connected, confirmthat the ShadowImage volume is split.$ sudo pairvolchk {-g group-name | -d volume-name }Note:




6. In the node or virtual server to which the P-VOL is connected, permit thestopped operations from the file snapshot functionality to the P-VOL.$ sudo horcunfreeze -f source-file-system-name

7. In the node or virtual server to which the P-VOL is connected, mount theP-VOL using the fsmount command in File Services Manager and set theNFS/CIFS shares using the nfscreate command and the cifscreatecommand then mount the differential-data snapshots using thesyncmount command.


8. In the node or virtual server to which the P-VOL is connected, mount theP-VOL NFS shares at a client-side and restart the program which accessesthe P-VOL.


For a non-tiered file systemsudo horcvmimport -f target-file-system-name -d device-file-number,...

For a tiered file systemsudo horcvmimport -f target-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...

10. From the node or virtual server that is connected to the S-VOL, executethe fsmount command to mount the S-VOL. After that, set the NFS/CIFSshares using the nfscreate command and the cifscreate command.

11. In the node or virtual server to which the S-VOL is connected, mount thedifferential-data snapshot using the syncmount command, and will bemade public the NFS/CIFS shares.


ShadowImage Volume Pair Split by Online Backup

When the file system is not the one operated by the file snapshotfunctionality or when the differential-data snapshot is not publishedin the shared file system even though the file system is operated bythe file snapshot functionality

1. In the node or virtual server to which the P-VOL is connected, stop theoperation from the file snapshot functionality to the P-VOL, hold theaccess from the client, and write the unreflected data to the disk.$ sudo horcfreeze -f source-file-system-name

2. In the node or virtual server to which the P-VOL is connected, split theShadowImage volume pair.$ sudo pairsplit {-g group-name|-d volume-name }

3. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair is split.$ sudo pairvolchk {-g group-name|-d volume-name }Note:




4. In the node or virtual server to which the P-VOL is connected, permit thewrite request to P-VOL, and permit the stopped access from the filesnapshot functionality.$ sudo horcunfreeze -f source-file-system-name








Note:In the file snapshot functionality target file system, mount thedifferential-data snapshot using the syncmount command ifnecessary, and set the NFS/CIFS shares using the nfscreatecommand and the cifscreate command.



1. In the node or virtual server to which the P-VOL is connected, stop theoperation from the file snapshot functionality to the P-VOL, stop the clientaccess request, and write the un-reflected data to the disk.$ sudo horcfreeze -f source-file-system-name


3. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair is split.$ sudo pairvolchk {-g group-name | -d volume-name }Note:





4. In the node or virtual server to which the P-VOL is connected, permit theclient access stop request to P-VOL, and permit the stopped operationsfrom the file snapshot functionality.$ sudo horcunfreeze -f source-file-system-name


For a non-tiered file systemsudo horcvmimport -f target-file-system-name -d device-file-number





Resynchronizing a ShadowImage Volume Pair

You need to separate the target file system using the horcexport commandbefore resynchronizing the pair. However, in spite of separating the target filesystem, resynchronizing the volume pair is not initial copy.

Note:In the file snapshot functionality target file system, unmount thedifferential-data snapshot using the syncumount command beforeexecuting the horcexport command.

To resynchronize the ShadowImage volume pair:

When the file system is not the one operated by the file snapshotfunctionality or when the differential-data snapshot is not madepublic in the shared file system even though the file system isoperated by the file snapshot functionality

1. Stop the program which accesses the S-VOL in the node or virtual serverto which the S-VOL is connected. Then, unmount the NFS shares on theclient side.

2. In the node or virtual server to which the S-VOL is connected, use thenfsdelete command and the cifsdelete command to delete the NFS/CIFS share. Then, use the fsumount command to unmount.

Note:In the file snapshot functionality target file system, delete the NFS/CIFS shares for the differential-data snapshot by using the nfsdelete


command and the cifsdelete command, unmount the differential-data snapshot by using the syncumount command.

3. In the node or virtual server to which the S-VOL is connected, separatethe file system of the S-VOL by using the horcexport command.$ sudo horcexport -f target-file-system-name

4. In the node or virtual server to which the P-VOL is connected, resume theShadowImage volume pair.$ sudo pairresync {-g group-name |-d volume-name }

5. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair has been resumed.$ sudo pairvolchk {-g group-name |-d volume-name }Note:


a. pairvolchk : Volstat is P-VOL.[status = COPY] => Recovering

b. pairvolchk : Volstat is P-VOL.[status = PAIR] => Recovered

When the file system is operated by the file snapshot functionalityand the differential-data snapshot is make public in the shared filesystem

1. In the node or virtual server to which the S-VOL is connected, stop theprogram which accesses and un-mount the NFS shares at a client-side.

2. In the node or virtual server to which the S-VOL is connected, un-mountthe differential-data snapshot using the syncumount command.

Note:When the differential-data snapshot is not only make public in theshared file system but also create the NFS/CIFS shares to thedifferential-data snapshot, it is required to cancel the NFS/CIFSshares for the differential-data snapshot using the nfsdelete andcifsdelete commands before un-mounting the differential-datasnapshot.

3. In the node or virtual server to which the S-VOL is connected, release theNFS/CIFS shares for the differential-data snapshot using the nfsdeleteand cifsdelete commands and un-mount all file systems in the pairusing the fsumount command.

4. In the node or virtual server to which the S-VOL is connected, separatethe S-VOL file system using the fsdelete command.$ sudo horcexport -f target-file-system-name

5. In the node or virtual server to which the P-VOL is connected,resynchronize the ShadowImage volume pair.$ sudo pairresync {-g group-name | -d volume-name }

6. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair has been resynchronized.


$ sudo pairvolchk {-g group-name | -d volume-name }Note:


a. pairvolchk : Volstat is P-VOL.[status = COPY]=> Recovering

b. pairvolchk : Volstat is P-VOL.[status = PAIR] => Recovered

Deleting a ShadowImage Volume Pair

Procedure of finishing the ShadowImage operations by deleting theShadowImage volume pair in the PSUS state differs depending on whetheryou will continue using the file system in S-VOL, or will destroy the filesystem.

When deleting the ShadowImage volume pair in the state other than PSUS,you cannot use the file system because the consistency of data in the S-VOLis not guaranteed as a file system.

To delete the ShadowImage volume pair in the PSUS status, and to continueto use the S-VOL file system after that:

1. In the node or virtual server to which the P-VOL is connected, delete theShadowImage volume pair.$ sudo pairsplit {-g group-name |-d volume-name } -S

2. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair has been deleted.$ sudo pairvolchk {-g group-name |-d volume-name}a. pairvolchk : Volstat is P-VOL.[status = COPY] => Deleting

b. pairvolchk : Volstat is P-VOL.[status = SMPL] => Deleted

3. In the node or virtual server to which the P-VOL is connected and thenode to which the S-VOL is connected, stop CCI.$ sudo horcmshutdown.sh (1-instance configuration)

or $ sudo horcmshutdown.sh 16 17 (2-instance configuration)

Note:If you have started the task of splitting the ShadowImage volume pair viaan offline backup but have not yet completed steps 1 through 11,complete steps 1 through 11 before proceeding with the above steps.When using an online backup, you need to have completed steps 1through 7 of splitting the ShadowImage volume pair before proceedingwith the above steps (see Splitting a ShadowImage Volume Pair on page2-31).

To delete the ShadowImage volume pair in PSUS state without using the S-VOL file system:

When the file system is not the one operated by the file snapshotfunctionality or when the differential-data snapshot is not make


public in the shared file system even though the file system isoperated by the file snapshot functionality

1. In the node or virtual server to which the S-VOL is connected, terminatethe program which accesses the S-VOL.

2. In the node or virtual server to which the S-VOL is connected, deleteNFS/CIFS shares using the nfsdelete command and the cifsdeletecommand, and unmount S-VOL using the fsumount command.

Note:In the file snapshot functionality target file system, delete the NFS/CIFS shares for the differential-data snapshot using the nfsdeletecommand and the cifsdelete command, unmount the differential-data snapshot using the syncumount command, and release thedifferential-data storage device using the syncstop command.

3. In the node or virtual server to which the S-VOL is connected, delete thefile system in the S-VOL using the fsdelete command.

4. In the node or virtual server to which the P-VOL is connected, delete theShadowImage volume pair.$ sudo pairsplit {-g group-name|-d volume-name} -S

5. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair has been deleted.$ sudo pairvolchk {-g group-name |-d volume-name }a. pairvolchk : Volstat is P-VOL.[status = COPY] => Deleting


6. Stop CCI on the nodes or virtual servers connected to the P-VOL and theS-VOL.$ sudo horcmshutdown.sh (1-instance configuration)


Note:If you have started the task of splitting the ShadowImage volume pair viaan offline backup but have not yet completed steps 1 through 11,complete steps 1 through 11 before proceeding with the above steps.When using an online backup, you need to have completed steps 1through 7 of splitting the ShadowImage volume pair before proceedingwith the above steps (See Splitting a ShadowImage Volume Pair on page2-31).


1. In the node or virtual server to which the S-VOL is connected, terminatethe program which accesses the S-VOL.




3. In the node or virtual server to which the S-VOL is connected, release thedevice which stores the differential data using the syncstop command.

4. In the node or virtual server to which the S-VOL is connected, release theNFS/CIFS shares for the differential-data snapshot using the nfsdeleteand cifsdelete commands and un-mount all file systems in the pairusing the fsumount command.

5. In the node or virtual server to which the S-VOL is connected, delete theS-VOL file system using the fsdelete command.

6. In the node or virtual server to which the P-VOL is connected, delete theShadowImage volume pair.$ sudo pairsplit {-g group-name | -d volume-name } -S

7. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair has been deleted.$ sudo pairvolchk {-g group-name | -d volume-name}You can also use the pairevtwait command, which waits for the pair tobe deleted.

a. $ pairvolchk : Volstat is P-VOL.[status = COPY] => Deleting

b. $ pairvolchk : Volstat is P-VOL.[status = SMPL] => Deleted

8. In the node or virtual server to which the P-VOL is connected and thenode to which the S-VOL is connected, stop CCI.$ sudo horcmshutdown.sh (1-instance configuration)


To delete the ShadowImage volume pair in the state other than PSUS state:

1. In the node or virtual server to which the P-VOL is connected, delete theShadowImage volume pair.$ sudo pairsplit {-g group-name |-d volume-name } -S

2. In the node or virtual server to which the P-VOL is connected, check if theShadowImage volume pair has been deleted.$ sudo pairvolchk {-g group-name |-d volume-name }a. pairvolchk : Volstat is P-VOL.[status = COPY] => Deleting


3. In the node or virtual server to which the S-VOL is connected, release thedevice file used in the target file system.$ sudo horcvmdelete -d Device-file-number,...

4. In the node or virtual server to which the P-VOL is connected and thenode to which the S-VOL is connected, stop CCI.


$ sudo horcmshutdown.sh (1-instance configuration)


P-VOL Data Recovery from S-VOLThis section describes how to recover the P-VOL data to the status at thepoint of the backup when the ShadowImage pair is in PSUS status, by usingthe data that is backed up in the S-VOL. The operations to recover the P-VOLdata differ according to the following condition:

• When the P-VOL file system is in normal status, see Recovering DataWhen the P-VOL File System is in Normal Status on page 2-41.

• When the P-VOL file system is blocked, see Recovering Data When the P-VOL File System is Blocked on page 2-45.

• When an error occurs on the device file used by the P-VOL, seeRecovering Data When an Error Occurs on the Device File Used by the P-VOL on page 2-50.

Note that if the ShadowImage pair is not in PSUS status, you cannot recoverthe P-VOL data since the consistency of the S-VOL data as the file system isnot guaranteed.

Recovering Data When the P-VOL File System is in Normal Status

When the P-VOL file system is in normal status, separate the source filesystem and then recover the data. Assuming the P-VOL is in PSUS status andthe S-VOL is in SSUS status, to recover the data:

When the file system is not the one operated by the file snapshotfunctionality or when the differential-data snapshot is not makepublic in the shared file system even though the file system isoperated by the file snapshot functionality

1. Unmount the NFS share of P-VOL and S-VOL on the client side.

2. In the node or virtual server to which the P-VOL is connected, delete NFS/CIFS shares by using the nfsdelete command and the cifsdeletecommand, and unmount P-VOL by using the fsumount command.

Note:In the file snapshot functionality file system, separate the NFS/CIFSshares for the differential-data snapshot by using the nfsdeletecommand and the cifsdelete command, unmount the differential-data snapshot by using the syncumount command.

3. In the node or virtual server to which the P-VOL is connected, separatethe file system in the P-VOL by using the horcexport command.$ sudo horcexport -f source-file-system-name

4. In the node or virtual server to which the S-VOL is connected, deleteNFS/CIFS shares by using the nfsdelete command and the cifsdeletecommand, and unmount S-VOL by using the fsumount command.



5. In the node to which the S-VOL is connected, separate the file system ofthe S-VOL by using the horcexport command.$ sudo horcexport -f target-file-system-name

6. In the node or virtual server to which the P-VOL is connected,resynchronize the ShadowImage pair in reverse direction.$ sudo pairresync {-g group-name |-d volume-name } -restore

7. In the node or virtual server to which the P-VOL is connected, check if theShadowImage pair has been resynchronized.$ sudo pairvolchk {-g group-name |-d volume-name }Note:

You may use the pairevtwait command to check whether theresynchronization is finished or not. Execute the pairevtwaitcommand and if the result is pairvolchk : Volstat is P-VOL.[status = PAIR], resynchronization is finished.

8. In the node or virtual server to which the P-VOL is connected, split theShadowImage pair.$ sudo pairsplit {-g group-name |-d volume-name }

9. In the node or virtual server to which the P-VOL is connected, connect thesource file system to the node or virtual server.

For a non-LVM and non-tiered file systemsudo horcimport -f source-file-system-name -d device-file-number

For a non-LVM and tiered file systemsudo horcimport -f source-file-system-name --tier1 device-file-number --tier2 device-file-number

LVM and non-tiered file systemsudo horcvmimport -f source-file-system-name -d device-file-number,...

For a LVM and non-tiered file systemsudo horcvmimport -f source-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...

10. In the node or virtual server to which the P-VOL is connected, mount theP-VOL by using the fsmount command, and set the NFS/CIFS shares byusing the nfscreate command and the cifscreate command.

Note:In the file snapshot functionality file system, mount the differential-data snapshot by using the syncmount command if necessary, and setthe NFS/CIFS shares by using the nfscreate command and thecifscreate command.


11. In the node to which the S-VOL is connected, connect the copydestination file system to the node.





12. In the node or virtual server to which the S-VOL is connected, mount theS-VOL by using the fsmount command, and set the NFS/CIFS shares byusing the nfscreate command and the cifscreate command.


13. Mount the NFS share of P-VOL and S-VOL on the client side.


1. Un-mount the NFS shares of the P-VOL and S-VOL at a client-side.

2. In the node or virtual server to which the P-VOL is connected, un-mountthe differential-data snapshot using the syncumount command.


3. In the node or virtual server to which the P-VOL is connected, delete theNFS/CIFS shares using the nfsdelete command and the cifsdeletecommand, and un-mount the P-VOL file systems using the fsumountcommand.

4. In the node or virtual server to which the P-VOL is connected, separatethe P-VOL file system using the horcexport command.$ sudo horcexport -f source-file-system-name




6. In the node or virtual server to which the S-VOL is connected, delete theNFS/CIFS shares using the nfsdelete command and the cifsdeletecommand, and un-mount the S-VOL using the fsumount command.


8. In the node or virtual server to which the P-VOL is connected,resynchronize the ShadowImage volume pair in the reverse direction.$ sudo pairresync {-g group-name | -d volume-name } -restore

9. In the node or virtual server to which the P-VOL is connected, check thecompletion of the resynchronization of the ShadowImage volume pair.$ sudo pairvolchk {-g group-name | -d volume-name}Note:



11. In the node or virtual server to which the P-VOL is connected, connect thecopy source file system to the node or virtual server.

For a non-tiered file systemsudo horcvmimport -f source-file-system-name -d device-file-number,...

For a tiered file systemsudo horcvmimport -f source-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...

12. In the node or virtual server to which the P-VOL is connected, mount theP-VOL using the fsmount command, and set the NFS/CIFS shares usingthe nfscreate command and the cifscreate command.

13. In the node or virtual server to which the P-VOL is connected, mount thedifferential-data snapshot using the syncmount command, and will bemade public the NFS/CIFS shares.



For a non-tiered file systemsudo horcvmimport -f target-file-system-name -d device-file-number,,...


15. In the node or virtual server to which the S-VOL is connected, mount theS-VOL using the fsmount command and set the NFS/CIFS shares usingthe nfscreate command and the cifscreate command.


17. Mount the NFS shares of the P-VOL and S-VOL at a client-side.

Recovering Data When the P-VOL File System is Blocked

When the P-VOL file system is blocked because of a logical failure, first deletethe blocked source file system, and then recover the data.

To delete the file system (when the file system is not the oneoperated by the file snapshot functionality or when the differential-data snapshot is not make public in the shared file system eventhough the file system is operated by the file snapshotfunctionality):

1. Unmount the NFS share of P-VOL and S-VOL on the client side.

2. Ask a service engineer to acquire the Dump of the node in which thefailed file system exists.

3. Perform the failback on the resource group of the failover node or virtualserver in step 2.

4. Delete NFS/CIFS shares of the P-VOL by using the nfsdelete commandand the cifsdelete command, and unmount the P-VOL using thefsumount command.

Note:In the file snapshot functionality file system, delete the NFS/CIFSshares for the differential-data snapshot by using the nfsdeletecommand and the cifsdelete command, unmount the differential-data snapshot by using the syncumount command, and release thedifferential-data storage device by using the syncstop command.

5. Delete the P-VOL file system by using the fsdelete command.

6. Start CCI on the nodes or virtual servers connected to the P-VOL and theS-VOL.

¢ $ sudo horcmstart.sh (1-instance configuration)

¢ $ sudo horcmstart.sh 16 17 (2-instance configuration)


Note:If the failed file system is not unmounted when step 2 is finished, deleteonly the CIFS share in step 3 by using the cifsdelete command.

To delete the file system (when the file system is operated by thefile snapshot functionality and the differential data snapshot ismake public in the shared file system)

1. Un-mount the NFS shares of the P-VOL and S-VOL at a client-side.

2. Get the Dump from the node in which the failed file system exists.

3. Execute a failback for the resource group of the node or virtual server towhich a failover occurred in Step 2.

4. Un-mount the differential-data snapshot using the syncumount command.


5. Release the device which stores the differential data using the syncstopcommand.

6. Delete NFS/CIFS shares of the P-VOL by using the nfsdelete commandand the cifsdelete command, and unmount the P-VOL using thefsumount command.

Note:In the file snapshot functionality file system, delete the NFS/CIFSshares for the differential-data snapshot by using the nfsdeletecommand and the cifsdelete command, unmount the differential-data snapshot by using the syncumount command, and release thedifferential-data storage device by using the syncstop command.

7. Delete the P-VOL file system using the fsdelete command.

8. Start CCI on the nodes or virtual servers connected to the P-VOL and theS-VOL.



Note:After the execution was completed up to Step 3, if a failed file system isnot mounted, perform only the release of the CIFS sharing in Step 6 usingthe cifsdelete command.

To recover P-VOL data (when the file system is not the one operatedby the file snapshot functionality or when the differential-data


snapshot is not make public in the shared file system even thoughthe file system is operated by the file snapshot functionality):




3. In the node or virtual server to which the P-VOL is connected, reserve adevice file to be used as a source file system.$ sudo horcvmdefine -d device-file-number,...

4. In the node or virtual server to which the P-VOL is connected,resynchronize the ShadowImage pair in reverse direction.$ sudo pairresync {-g group-name |-d volume-name } -restore

5. In the node or virtual server to which the P-VOL is connected, check if theShadowImage pair has been resynchronized.$ sudo pairvolchk {-g group-name |-d volume-name }Note:


6. In the node or virtual server to which the P-VOL is connected, split theShadowImage pair.$ sudo pairsplit {-g group-name |-d volume-name }




For a LVM and non-tiered file systemsudo horcvmimport -f source-file-system-name -d device-file-number,...

For a LVM and tiered file systemsudo horcvmimport -f source-file-system-name --tier1 device-file-number --tier2 device-file-number









10. In the node or virtual server to which the S-VOL is connected, mount theS-VOL by using the fsmount command, and set the NFS/CIFS shares byusing the nfscreate command and the cifscreate command.


11. Mount the NFS share of P-VOL and S-VOL on the client side.

To recover P-VOL data (when the file system is operated by the filesnapshot functionality and the differential-data snapshot is makepublic in the shared file system)






4. In the node or virtual server to which the P-VOL is connected, reserve adevice file that was used by the copy source file system.$ sudo horcvmdefine -d device-file-number,...

5. In the node or virtual server to which the P-VOL is connected,resynchronize the ShadowImage volume pair in the reverse direction.$ sudo pairresync {-g group-name | -d volume-name } -restore

6. In the node or virtual server to which the P-VOL is connected, check thecompletion of the resynchronization of the ShadowImage volume pair.$ sudo pairvolchk {-g group-name | -d volume-name}Note:



8. In the node or virtual server to which the P-VOL is connected, connect thecopy source file system to the node or virtual server.

For a non-tiered file systemsudo horcvmimport -f source-file-system-name -d device-file-number

For a tiered file systemsudo horcvmimport -f source-file-system-name --tier1 device-file-number --tier2 device-file-number,...




For a non-tiered file systemsudo horcvmimport -f target-file-system-name -d device-file-number,...



12. In the node or virtual server to which the S-VOL is connected, mount theP-VOL using the fsmount command and set the NFS/CIFS shares usingthe nfscreate command and the cifscreate command.


14. Mount the NFS shares of the P-VOL and S-VOL at a client-side.

Recovering Data When an Error Occurs on the Device File Used by the P-VOL

When an error occurs on the device file, first delete the ShadowImage pair.After that, recover the failure of the device file and recover the data, andthen create the ShadowImage pair again.

To apply maintenance mode:

1. Apply maintenance mode by using the lumapctl command. Applying themaintenance mode prevents that the device file number on the storageside is automatically allocated to the user LU.$ sudo lumapctl -t m --on

To delete the ShadowImage pair:

1. In the node or virtual server to which the P-VOL is connected, delete theShadowImage pair.$ sudo pairsplit {-g group-name |-d volume-name } -S

2. In the node or virtual server to which the P-VOL is connected, check if theShadowImage pair has been deleted.$ sudo pairvolchk {-g group-name |-d volume-name }pairvolchk : Volstat is P-VOL.[status = COPY] => Deleting

pairvolchk : Volstat is P-VOL.[status = SMPL] => Deleted

To recover the failure of the device file and delete file system (whenthe file system is not the one operated by the file snapshotfunctionality or when the differential-data snapshot is not makepublic in the shared file system even though the file system isoperated by the file snapshot functionality):

1. Delete the NFS/CIFS shares using the nfsdelete command and thecifsdelete command, and un-mount the P-VOL using the fsumountcommand.

Note:In the file snapshot functionality file system, before delete the filesystem, delete the NFS/CIFS shares for the differential-data snapshotusing the nfsdelete command and the cifsdelete command, andun-mount the differential-data snapshot using the syncumountcommand, and release the device which stores the differential datausing the syncstop command.


2. Delete the file system of the P-VOL by using fsdelete command.

3. Ask maintenance personnel to recover the failed device file.

4. Perform the following operations in order:

¢ Fail over the resource group.

¢ Stop the node.

¢ Ask a service engineer to restart the OS on the node that wasstopped.

¢ Start the node.

¢ Fail back the resource group.

5. Perform step 4 for the other node.

6. Start CCI on the node or virtual server to which the P-VOL is connectedand in the node to which the S-VOL is connected.



To recover the failure of the device file (when the file system isoperated by the file snapshot functionality and the differential-datasnapshot is published in the shared file system):

1. In the node or virtual server to which the P-VOL is connected, un-mountthe differential-data snapshot using the syncumount command.


2. In the node or virtual server to which the P-VOL is connected, release thedifferential-data storage device by using the syncstop command.

3. Delete the NFS/CIFS shares using the nfsdelete command and thecifsdelete command, and un-mount the P-VOL using the fsumountcommand.

4. Delete the file system of the P-VOL by using fsdelete command.

5. Ask maintenance personnel to recover the failed device file.

6. Perform the following operations in order:

¢ Fail over the resource group.

¢ Stop the node.

¢ Ask a service engineer to restart the OS on the node that wasstopped.

¢ Start the node.

¢ Fail back the resource group.


7. Perform step 6 for the other node.

8. Start CCI in the node or virtual server to which the P-VOL is connectedand in the node or virtual server to which the S-VOL is connected.



To restore the data to the P-VOL (when the file system is not theone operated by the file snapshot functionality or when thedifferential-data snapshot is not make public in the shared filesystem even though the file system is operated by the file snapshotfunctionality):


2. In the node or virtual server to which the P-VOL is connected, reserve adevice file to be used as a source file system.$ sudo horcvmdefine -d device-file-number,...

3. In the node or virtual server to which the S-VOL is connected, create aShadowImage pair whose copy direction is opposite to that of the deletedpair.$ sudo paircreate {-g group-name|-d volume-name} -vl

4. Check if the ShadowImage volume has been created in the node or virtualserver to which the S-VOL is connected.$ sudo pairvolchk {-g group-name |-d volume-name }Note:

You may use the pairevtwait command to check whether thecreation of the pair is finished or not. Execute the pairevtwaitcommand and if the result is pairvolchk : Volstat is P-VOL.[status = PAIR], creation is finished.

5. In the node or virtual server to which the S-VOL is connected, suppressthe file snapshot functionality operations on the S-VOL.$ sudo horcfreeze -f source-file-system-name

6. In the node or virtual server to which the S-VOL is connected, split theShadowImage pair.$ sudo pairsplit {-g group-name |-d volume-name }

7. In the node or virtual server to which the S-VOL is connected, check if theShadowImage pair has been split.$ sudo pairvolchk {-g group-name |-d volume-name }Note:

You may use the pairevtwait command to check whether the splitprocessing is finished or not. Execute the pairevtwait command andif the result is pairvolchk : Volstat is P-VOL.[status = PSUS],split processing is finished.





For a LVM and non-tiered file systemsudo horcvmimport -f source-file-system-name -d device-file-number,...

For a LVM and tiered file systemsudo horcvmimport -f source-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...

9. In the node or virtual server to which the S-VOL is connected, delete theShadowImage pair.$ sudo pairsplit {-g group-name |-d volume-name } -S

10. In the node or virtual server to which the S-VOL is connected, check if theShadowImage pair has been deleted.$ sudo pairvolchk {-g group-name |-d volume-name }Note:

You may use the pairevtwait command to check whether the deleteprocessing is finished or not. Execute the pairevtwait command andif the result is pairvolchk : Volstat is P-VOL.[status = SMPL],delete processing is finished.

11. In the node or virtual server to which the S-VOL is connected, cancel thesuppression of the file snapshot functionality operations.$ sudo horcunfreeze -f source-file-system-name



13. In the node or virtual server to which the S-VOL is connected, separatethe S-VOL file system by using the horcexport command.$ sudo horcexport -f target-file-system-nameNote:

In the file snapshot functionality file system, separate the NFS/CIFSshares for the differential-data snapshot by using the nfsdeletecommand and the cifsdelete command, unmount the differential-data snapshot by using the syncumount command.


To restore the data to the P-VOL (when the file system is operatedby the file snapshot functionality and the differential-data snapshotis published in the shared file system)


Note:When the differential-data snapshot is published in the shared filesystem and the NFS/CIFS shares is created in the differential-datasnapshot, delete the NFS/CIFS shares for the differential-datasnapshot using the nfsdelete command and the cifsdeletecommand before un-mounting the differential-data snapshot.


3. In the node or virtual server to which the P-VOL is connected, reserve adevice file that was used by the copy source file system.$ sudo horcvmdefine -d device-file-number,...

4. In the node or virtual server to which the S-VOL is connected, create aShadowImage volume pair by copying data in the direction reverse to theusual.$ sudo paircreate {-g group-name | -d volume-name } -vl

5. In the node or virtual server to which the S-VOL is connected, check thecompletion of the creation of the ShadowImage volume pair.$ sudo pairvolchk {-g group-name | -d volume-name}Note:


6. In the node or virtual server to which the S-VOL is connected, suppressthe file snapshot functionality operations on the S-VOL.$ sudo horcfreeze -f source-file-system-name

7. In the node or virtual server to which the S-VOL is connected, split theShadowImage volume pair.$ sudo pairsplit {-g group-name | -d volume-name }

8. In the node or virtual server to which the S-VOL is connected, confirmthat the ShadowImage volume is split.$ sudo pairvolchk {-g group-name | -d volume-name }Note:

You may use the pairevtwait command to check whether the splitprocessing is finished or not. Execute the pairevtwait command andif the result is pairvolchk : Volstat is P-VOL.[status = PSUS],split processing is finished.

9. In the node or virtual server to which the P-VOL is connected, connect thecopy source file system to the node.


For a non-tiered file systemsudo horcvmimport -f source-file-system-name -d device-file-number...

For a tiered file systemsudo horcvmimport -f source-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...

10. In the node or virtual server to which the S-VOL is connected, delete theShadowImage volume pair.$ sudo pairsplit {-g group-name | -d volume-name } -S

11. In the node or virtual server to which the S-VOL is connected, check if theShadowImage volume pair has been deleted.$ sudo pairvolchk {-g group-name | -d volume-name}Note

You may use the pairevtwait command to check whether the deleteprocessing is finished or not. Execute the pairevtwait command andif the result is pairvolchk : Volstat is P-VOL.[status = SMPL],delete processing is finished.

12. In the node or virtual server to which the S-VOL is connected, permit thestopped operations from the file snapshot functionality to the S-VOL.$ sudo horcnfreeze -f source-file-system-name



15. In the node or virtual server to which the S-VOL is connected, release thedevice which stores the differential data using the syncstop command.

16. In the node or virtual server to which the S-VOL is connected, separatethe S-VOL file system using the horcexport command.$ sudo horcexport -f target-file-system-name

To restore the ShadowImage volume pair

1. In the node or virtual server to which the P-VOL is connected, create aShadowImage pair.$ sudo paircreate {-g group-name|-d volume-name} -vl

2. Check if the ShadowImage volume has been created in the node or virtualserver to which the P-VOL is connected.$ sudo pairvolchk {-g group-name |-d volume-name }Note:



3. Split the ShadowImage pair.For details about how to split a ShadowImage volume pair, see Overviewof Using ShadowImage on page 2-30.

To apply normal management mode:

1. Apply normal management mode by using the lumapctl command.$ sudo lumapctl -t m --off

Data Backup Cooperated with a Tape DeviceThis section explains how to back up data from an online volume to a tape byusing a tape device connected to a node or virtual server via an FC interface.The backup to the tape media by using ShadowImage is the full backup.When making the incremental backup to a tape media, use the file snapshotfunctionality. Refer to the Hitachi NAS Platform F1000 Series ClusterAdministrator's Guide for the procedure for the backup using the file snapshotfunctionality.

Prerequisites before starting operation

• The ShadowImage volume pair created and it in the pair status of PSUS(refer to Overview of Using ShadowImage on page 2-30).

• The tape device connected to the node or virtual server via the FCinterface and the tape device in the Ready status (refer to the Installationand Configuration Guide).

• The operating environment to be used with Backup Restore is set in thebackup management software (such as NetBackup). (Refer to thesupplemental documentation for Backup Restore provided with the HNASF system.)

The procedure for backing up data:

1. Resynchronize the pair. Change the pair status from PSUS to PAIR byupdating the S-VOL using the differential between the data of the P-VOLand the S-VOL.For details about how to resynchronize a ShadowImage volume pair, seeOverview of Using ShadowImage on page 2-30.

2. Split the pair following the procedure for the online backup.For the procedure on how to use the online backup method to split aShadowImage volume pair, see Overview of Using ShadowImage on page2-30.

3. Copy the updated data from the S-VOL to the tape media.

Note:Concerning the backup to a tape media using ShadowImage,acceptance of a Read/Write I/O instruction by the P-VOL ofShadowImage realizes the online backup. Perform the operationfollowing the procedure for the execution of the offline backup byBackup Restore because the offline S-VOL is updated based on the


data of the online P-VOL and then the backup is made from the S-VOLto the tape media.

Notes on Failures of External VolumesIf a trouble ( including a temporary trouble, for example, the cableconnection has cut out) occurs to the external volume which is currently usedwith ShadowImage volume pair, perform the following processes.

1. If the external volume can not be used succeedingly, deleteShadowImage volume pair by using the pairsplit -S command.

If the S-VOL is not connected to the node or virtual server, release thedevice file used in the target file system by using the horcvmdeletecommand.

2. Confirm all the ShadowImage volume pair statuses that are using theexternal volume with trouble(s) by using the pairdisplay command.

If the status of the ShadowImage volume pair is PSUE, delete the volumepair by using the pairsplit -S command.

If you delete ShadowImage volume pair when the S-VOL is not connectedto the node or virtual server, release the device file used in the target filesystem by using the horcvmdelete command.

How to use TrueCopyThis section describes how to use TrueCopy with the HNAS F.

Overview of TrueCopy Operations in the HNAS FThe HNAS F system enables integrated management of the data in thestorage system. HNAS F utilizes the LAN environment already in place andenables the data within a. HNAS F utilizes the LAN environment already inplace and enables the data within a storage system to be shared acrossheterogeneous platforms. HNAS F also enables you to copy and maintain thedata stored in the storage system using the volume replication featureTrueCopy. With TrueCopy, you can copy data utilizing the file system server(even in the remote center), back up data to guard against a main volumefailure, and perform disaster recovery at the main site.

To use TrueCopy in the HNAS F system, use Command Control Interface(CCI). As CCI is installed automatically with the OS, there is no need forseparate installation of CCI in your HNAS F system.

Figure 2-6 Using TrueCopy in the HNAS F system on page 2-58 illustratesthe TrueCopy operation.


Figure 2-6 Using TrueCopy in the HNAS F system

Scope of TrueCopy Function with the HNAS F

Scope of TrueCopy Function with the HNAS F

Volume type

Only the user LUs may be used as TrueCopy P-VOLs and S-VOLs in an HNASF system

You cannot specify the OS disks and shared LU of a virtual server as aTrueCopy P-VOL or S-VOL.

Platforms which can access the TrueCopy P-VOLs or S-VOLs

A P-VOL or S-VOL of TrueCopy created by the HNAS F system may beaccessed by the HNAS F system or clients connected to the HNAS F through anetwork. Hosts connected via serial port or fibre-channel port will not be ableto access them.

File systems which can be allocated in the TrueCopy P-VOL

You cannot allocate a file system that consists of 129 or more LUs and thatuses Logical Volume Manager (LVM) on the OS.

You cannot allocate a volume group that consists of 129 or more LUs if itcontains a file system managed by the file snapshot functionality anddifferential-data storage devices.

Writing data to S-VOL when a volume pair is split

When you split a TrueCopy volume pair in the HNAS F system, you need tochange the S-VOL to write-enable. Make sure that you always specify the –rwoption instead of the -r option when issuing the CCI pairsplit command.

To protect the data in the split S-VOL from being written to by the client, usethe fsmount command with –r option in File Services Manager whenmounting the file system in S-VOL.


Note on the path settings

If you use the HNAS F system with TrueCopy, you must add paths to theports of both CL1 and CL2 to connect the primary and secondary sites.

Scope Related to Features in Backup Restore


If you resynchronize a TrueCopy pair defined by Volume Replication functionafter performing a backup of the S-VOL by using the backup managementsoftware, a full backup will be acquired next time.



When you copy a file system managed by the file snapshot functionality usingTrueCopy, you must copy both the LUs that constitute the file system and theLUs which constitute the differential-data storage devices. If you copy onlythe LUs which constitute a file system, you will not be able to connect themto the HNAS F system on the remote site.


Prerequisites


You need a workstation or PC to log in to the HNAS F system using secureshell (SSH) in addition to the prerequisite hardware for TrueCopy, CCI, andthe HNAS F system described in the following guides:



Manuals related to storage systems

¢ Hitachi TrueCopy User and Reference Guide


¢ Hitachi ShadowImage User's Guide (User Guide) (when using theShadowImage cascade function)




In addition, to use TrueCopy in a HNAS F system, all of the following programproducts must be installed in the storage system to which the HNAS F systemis connected, and their licenses must be valid:

• TrueCopy

• ShadowImage(Required when configuring the cascade connection of TrueCopy/ShadowImage)

Notes on Operations



Limitations on TrueCopy operations due to the status of cluster,nodes, and resource groups

When a cluster is not configured, the cluster is stopped, the nodes arestopped, and the resource groups are offline, connecting the device file orcreating and mounting a file system is restricted. Due to these restrictions,the following operations performed in the target TrueCopy operation will alsoend in error. You therefore should not operate the cluster, nodes, andresource groups during the TrueCopy operation. Should any problems occurwith the cluster, nodes, or resource groups, fix them immediately.

• Unmount and mount the source file system during the splitting of aTrueCopy pair volume.


• Unmount and delete the target file system before the TrueCopy pair isresynchronized.

Notes on occurrence of a failover during operating on node


Notes on changing system configuration during the TrueCopyoperations

If any of the following are changed during the operation of TrueCopy, youmust change the Command Control Information configuration definition fileson the node:


• Changing the fixed IP address of the node

• Expanding or deleting a source file system

• Setting up, expanding or releasing a differential-data storage device forthe file snapshot functionality

When the host name is specified in the Command Control Interfaceconfiguration definition file and you change the following systemconfigurations, you must change the configuration definition file:

• Editing the /etc/hosts file (when resolving the host name using the /etc/hosts file)



• Changing the node name or the host name


In case of periodic maintenance of external volumes, you must delete allTrueCopy pairs once.

Notes on splitting TrueCopy volume pairs with online backup

After the horcfreeze command is executed, if it takes long time until thehorcunfreeze command is executed, the timeout may occur in some client.In addition, if the file snapshot functionality is used on the copy source filesystem, the timeout tends to occur because the horcfreeze command takeslong time to be executed.

Confirming whether the access to the file system is suspended

After you execute the horcfreeze command and the horcunfreezecommand, you can check whether the access from the client to the filesystem is suspended or not by using the fsctl command.







If a TrueCopy pair is split and connected to the node while the differentialdata storage device in the copy source file system does not have sufficientcapacity, certain measures must be taken in the copy source file system tosolve the lack of the capacity.


In splitting a TrueCopy pair, confirm that the differential data storage devicein the copy source file system is in the normal status.


The single-instancing setting is not copied over copy-destination file systems.If the single-instancing setting is enabled for a copy-source file system, aftersplitting a TrueCopy pair and connecting to a node, use the fsedit commandto enable the single-instancing setting for the copy-destination file system.Also, set the policy for duplicate file capacity reduction.

Before splitting a TrueCopy pair, make sure that the duplicate file capacityreduction policy has not yet been executed on the copy source file system. Ifa TrueCopy pair is split during policy execution, resynchronize the TrueCopypair and perform its subsequent operations again.


If the copy-source file system is a tiered file system, the LUs of all the filesystems that make up the tier must be assigned the same device group namein the configuration definition file.

To connect the copy-source file system to a node or a virtual server, specifythe --tier1 and --tier2 options for the horcimport command or thehorcvmimport command if the file system is a tiered file system. Otherwise,specify the -d option.










A node cannot be connected to if TrueCopy is used to copy a WORM filesystem.


If the local data encryption functionality is being used, the copy-destinationfile system can be connected to the node only when the copy-source filesystem and the copy-destination file system are in the same cluster. If thefile systems are in different clusters, the copy-destination file system cannotbe connected to the node.

Preparing for TrueCopy Operations

Preparing for TrueCopy Volume Pair Operation

Make the preparations necessary for creating TrueCopy volume pairs. Fordetails on the preparations required on the storage system side, see theHitachi TrueCopy User's Guide (User Guide).

Registration of Public Key Used for SSH

Before issuing the commands described in this document, the public key usedfor SSH needs to be registered in the node or virtual server to which theTrueCopy P-VOL is connected, and also in the node or virtual server that willuse the S-VOL. You can register the public key from the Access ProtocolConfiguration window, in the Add Public Key page.

Configuring the CCI Environment

Logging in to a node via SSH

Using the nasroot account via SSH, log in both the node in which theTrueCopy P-VOL is connected, and the node which will use the S-VOL. (Forinformation about login, refer to the appropriate documentation of the SSHcommunication software.)



For using additional instance numbers, the environment for the instancenumbers to be used must be configured by using the horcsetconf command.$ sudo horcsetconf -i instance number








To control a TrueCopy pair by using CCI, you must first define the TrueCopypair by using the CCI configuration definition file.

A Command Control Interface configuration definition file template is madeavailable with the installation of HNAS F.

The template of the CCI configuration definition file:/home/nasroot/horcm<instance-number>.confFor nodes, <instance number> is 16 or 17. Additional instance numbers are from 20 to 499.Next, add the HORCM_MON, HORCM_CMD, HORCM_DEV, and HORCM_INST sectionsto the CCI configuration definition file template using the CCI mkconf.sh andhorcconfedit commands. After that, edit the file as necessary.

To complete the CCI configuration definition file:

1. Add the HORCM_DEV section and the HORCM_INST section to a templatefor the Command Control Interface configuration definition file using themkconf.sh command in Command Control Interface.

2. Edit the file and create the Command Control Interface configurationdefinition file.

You must perform these operations in both nodes of the main site and in bothnodes of the remote site. You therefore need to prepare a total of four CCIconfiguration files, assuming one instance per node. For two instances, youwill need a total of eight configuration files.

In the HNAS F, you can define one or two CCI instances per a node. Tooperate only the TrueCopy pairs in CCI, you need one instance. To operatethe pairs where TrueCopy and ShadowImage are cascaded, you need twoinstances.


In the following section, we explain how to create a CCI configurationdefinition file using the HNAS F system with LUs as an example.



Use the CCI mkconf.sh command to define the HORCM_MON, HORCM_CMD,HORCM_DEV, and HORCM_INST sections in the CCI configuration definition filetemplate.$ ls /dev/sdu*u | sudo mkconf.sh -gg device-group-name -i 16



To execute the mkconf.sh command, the file system to be paired must becreated. The LU configuration (size and number) in the file system to be


paired needs to be exactly the same in both the primary site and thesecondary site.

To create the correct CCI configuration definition file, it is recommended thatyou create a file system to be temporarily paired before executing themkconf.sh command. After creating the CCI configuration definition file usingthe mkconf.sh command, you can continue using the file system in theprimary site. Alternatively, you may also delete it and create a file systemwith the same configuration using the same LU in Create File System inTrueCopy P-VOL on page 2-76. You must delete the temporarily created filesystem in the secondary site before creating a pair.

Example 2-29 Example of mkconf.sh Execution Result (for Instance Number16) on page 2-66 shows an example of how to create a template for a CCIconfiguration definition file. We will explain how to create a CCI configurationdefinition file in node0 in the primary site in the sample configuration above.You must create CCI configuration definition files for the other nodes usingthe previously explained procedure.$ ls /dev/sdu*u | sudo mkconf.sh -gg VG -i 16starting HORCM inst 16HORCM inst 16 starts successfully.HORCM Shutdown inst 16 !!!A CONFIG file was successfully completed.starting HORCM inst 16HORCM inst 16 starts successfully.DEVICE_FILE Group PairVol PORT TARG LUN M SERIAL LDEV/dev/sdu00u VG VG_000 CL1-A-1 0 17 - 62486 70/dev/sdu01u VG VG_001 CL1-A-1 1 18 - 62486 18 : :/dev/sdu10u VG VG_010 CL1-A-1 0 10 - 62486 10/dev/sdu11u VG VG_011 CL1-A-1 0 11 - 62486 64/dev/sdu12u VG VG_012 CL1-A-1 0 12 - 62486 12/dev/sdu13u VG VG_013 CL1-A-1 0 13 - 62486 66/dev/sdu14u VG VG_014 CL1-A-1 0 14 - 62486 14/dev/sdu15u VG VG_015 CL1-A-1 0 15 - 62486 68/dev/sdu16u VG VG_016 CL1-A-1 0 16 - 62486 16HORCM Shutdown inst 16 !!!Please check '/home/nasroot/horcm16.conf','/home/nasroot/log16/curlog/horcm_*.log', andmodify ‘ip_address & service'.#

Example 2-29 Example of mkconf.sh Execution Result (for InstanceNumber 16)


Example 2-30 Changing the HORCM_CMD definition in the CCIconfiguration definition file (for instance number 16)


Editing the CCI configuration definition file

The following table shows the values are specified for the items included inthe CCI configuration definition file in the HNAS F system as shown in Table2-12.

Table 2-12 Configuration Definition File Settings (HORCM_MON) andSpecified Values in HNAS F system


HORCM_MON ip_address Fixed IP address or virtual IP address of the local node.

service Specify one of the following:



• 31032 to 31254 or 31532 to 31754 (In the case of avirtual server. Instance number +30000)

• 30020 to 30499 (Common. Instance-number+30000)


Using the information in Table 2-12, change the service and ip_addressentries.

Change the entries poll and timeout to the appropriate values according tohardware requirements.HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.78.51 20331 1000 3000


HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu00u SER = 62486 LDEV =70 [ FIBRE FCTBL = 3 ]VG VG_000 CL1-A-1 0 0# /dev/sdu01u SER = 62486 LDEV =18 [ FIBRE FCTBL = 3 ]VG VG_001 CL1-A-1 0 1 :# /dev/sdu13u SER = 62486 LDEV =66 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 13# /dev/sdu14u SER = 62486 LDEV =14 [ FIBRE FCTBL = 3 ]


VG VG_014 CL1-A-1 0 14# /dev/sdu15u SER = 62486 LDEV =68 [ FIBRE FCTBL = 3 ]VG VG_015 CL1-A-1 0 15# /dev/sdu16u SER = 62486 LDEV =16 [ FIBRE FCTBL = 3 ]VG VG_016 CL1-A-1 0 16


Example 2-31 Example of CCI Configuration Definition File - 1 (forInstance Number 16)

Next, delete unnecessary LU entries (lines) for the HORCM_DEV sectionexcept for the LU entry (line) that you want to control using CCI.

LUs which constitute a file system and their LDEV numbers can be reviewedusing the following command.$ sudo horcdevlist | grep ':File System Name$'For example, in Example 2-32 Example of the Command Execution whichLists LUs which Constitute the File System Sample on page 2-68, 11, 12,13 are LUs that constitute the file system sample. 64, 12, and 66 are theirLDEV numbers.$ sudo horcdevlist | grep ':sample$'11 62486 64 OPEN-V 3.906GB -- -- - Normal File:sample12 62486 12 OPEN-V 3.906GB -- -- - Normal File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal File:sample

Example 2-32 Example of the Command Execution which Lists LUs whichConstitute the File System Sample

The device file name and the device name of the LU entry (line) you want tocontrol using Command Control Interface are changed to an appropriatename. In changing the device file name and the device name, observe thefollowing:

• The same device group name and the same device name need to bespecified to the volumes to be paired in the primary site and in thesecondary site. Specify the device group name and the device nameaccordingly.

• The same device group name must be specified as LUs which constituteone file system. If the file snapshot functionality manages the file system,the same device group name must be specified as LUs that constitute afile system and as LUs that constitute differential-data storage devices.








Next, specify the IP address of the instance to be paired in the secondary sitefor the HORCM_INST section. As a preparation for failover, specify the IPaddresses of the instances in both nodes.

Table 2-13 Configuration Definition File Settings (HORCM_INST) andSpecified Values in HNAS F

Section Name Item Specified Values in HNAS F

HORCM_INST ip_address Fixed IP address of the other TrueCopy node

Service Specify one of the following:



• 31032 to 31254 or 31532 to 31754 (In the case of avirtual server. Instance number +30000)


Note:A host name can be specified instead of a IP address if the fixed IPaddress and the corresponding host name are registered into /etc/hosts,on an NIS server, or a DNS server. Refer to the Hitachi NAS PlatformF1000 Series Cluster Administrator's Guide for information on how toregister the IP address and the corresponding host name in /etc/hosts,and for information on how to set up the HNAS F system to search thehost name using NIS or DNS.



HORCM_DEV


#dev_group dev_name port# TargetID LU# MU## /dev/sdu11u SER = 62486 LDEV = 64 [ FIBRE FCTBL = 3 ]VG VG_011 CL1-A-1 0 11# /dev/sdu12u SER = 62486 LDEV = 12 [ FIBRE FCTBL = 3 ]VG VG_012 CL1-A-1 0 12# /dev/sdu13u SER = 62486 LDEV = 66 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 13



Checking the contents of the CCI configuration definition file.


Start CCI on both the node or virtual server connected to the TrueCopy P-VOL and the node or virtual server that will use the S-VOL.$ sudo horcsetenv HORCMINST 16 (For instance number 16) or sudo horcsetenv HORCMINST 17 (For instance number 17)$ sudo horcunsetenv HORCC_MRCFWhen you logged in using SSH and performed the above setup, confirm the setup by once logging out and relogging in again.$ sudo horcmstart.sh


By issuing the pairdisplay command in the node or virtual server in whichTrueCopy P-VOL or S-VOL is connected or in which S-VOL will be used, youcan see the LDEV numbers of LUs specified in the HORCM_DEV section (seeExample 2-35 Procedure for Starting CCI on page 2-70).$ sudo pairdisplay –g Device-Group-Name

Example 2-36 How to Check LDEV Numbers of LUs Specified in theHORCM_DEV Section

You can check the device file numbers, and the LDEV numbers for the devicefiles that constitute a file system, by issuing the horcdevlist command inthe node or virtual server in which P-VOL is connected. Compare this withExample 2-36 How to Check LDEV Numbers of LUs Specified in theHORCM_DEV Section on page 2-70.$ sudo horcdevlist | grep ':sample$'11 62486 64 OPEN-V 3.906GB -- -- - Normal File:sample12 62486 12 OPEN-V 3.906GB -- -- - Normal File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal File:sample



You can also check the device file numbers and the LDEV numbers for theunused device files which can be S-VOL by issuing the horcdevlistcommand in the node or virtual server in which S-VOL will be used. Comparethis with Example 2-36 How to Check LDEV Numbers of LUs Specified in theHORCM_DEV Section on page 2-70.$ sudo horcdevlist | grep ' Free$'21 62486 74 OPEN-V 3.906GB -- -- - Normal Free22 62486 22 OPEN-V 3.906GB -- -- - Normal Free23 62486 76 OPEN-V 3.906GB -- -- - Normal Free


When specifying the port name in the HORCM_DEV section of the CCIconfiguration definition file, use the name of the storage system FibreChannel adaptor port that connects to the node.

After checking the CCI configuration definition file, stop CCI in both the nodeor virtual server in which TrueCopy P-VOL is connected and in the node orvirtual server in which S-VOL is connected.$ sudo horcmshutdown.sh 16

Example 2-39 Stopping CCI (for Instance Number 16)

To save CCI configuration definition files, save the system settingsinformation. For details on saving system settings information, see theHitachi NAS Platform F1000 Series Cluster Administrator's Guide.

Example of a tiered file system configuration definition file

An example of a configuration definition file for a tiered file system isprovided below.



Example 2-40 How to Check Device File Numbers and LU Numbers of a P-VOL


HORCM_CMD#dev_name dev_name dev_name#UnitID 0 (Serial# 62486)\\.\CMD-62486:/dev/sdHORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10u SER = 62486 LDEV =10 [ FIBRE FCTBL = 3 ]VG_ VG_010 CL1-A-1 0 10# /dev/sdu11u SER = 62486 LDEV =64 [ FIBRE FCTBL = 3 ]VG_ VG_011 CL1-A-1 0 11# /dev/sdu12u SER = 62486 LDEV =12 [ FIBRE FCTBL = 3 ]VG_ VG_012 CL1-A-1 0 12# /dev/sdu13u SER = 62486 LDEV =66 [ FIBRE FCTBL = 3 ]VG_ VG_013 CL1-A-1 0 13

HORCM_INST#dev_group ip_address service


VG_TC 123.45.80.51 20331VG_TC 123.45.80.115 20331

Example 2-41 Example of a CCI Configuration Definition File (for P-VOLs)HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.78.115 20331 1000 3000


HORCM_INST#dev_group ip_address serviceVG_TC 123.45.80.115 20331VG_TC 123.45.80.51 20331


Cascade Configuration of TrueCopy and ShadowImage

In the HNAS F system, you may use the TrueCopy pair and ShadowImagepair in a cascade configuration. In a cascade configuration, you may preparefor disaster scenario where, for example, you would not be able to recoverthe file system from the TrueCopy S-VOL, by backing up the file systemcopied from the TrueCopy P-VOL to S-VOL periodically using ShadowImage.

Figure 2-9 Example of Cascade Configuration of TrueCopy and ShadowImageon page 2-74 shows an example of a cascade configuration of TrueCopy andShadowImage.


Figure 2-9 Example of Cascade Configuration of TrueCopy andShadowImage

By preparing the CCI configuration definition file as shown in Example 2-43Example of CCI Configuration Definition File at Main Site (Instance 16) onpage 2-75, Example 2-44 Example of CCI Configuration Definition File atRemote Site (Instance 16) on page 2-75, and Example 2-45 Example of CCIConfiguration Definition File at Remote Site (Instance 17) on page 2-76, youmay operate the cascade configuration of TrueCopy and ShadowImage fromCCI.HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.78.51 20331 1000 3000


HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10u SER = 62486 LDEV = 10 [ FIBRE FCTBL = 3 ]VG_TC VG_032 CL1-A-1 0 10# /dev/sdu11u SER = 62486 LDEV = 64 [ FIBRE FCTBL = 3 ]VG_TC VG_033 CL1-A-1 0 11# /dev/sdu12u SER = 62486 LDEV = 12 [ FIBRE FCTBL = 3 ]VG_TC VG_034 CL1-A-1 0 12# /dev/sdu13u SER = 62486 LDEV = 66 [ FIBRE FCTBL = 3 ]VG_TC VG_035 CL1-A-1 0 13



VG_TC 123.45.80.51 20331VG_TC 123.45.80.115 20331

Example 2-43 Example of CCI Configuration Definition File at Main Site(Instance 16)



HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10u SER = 62486 LDEV = 20 [ FIBRE FCTBL = 3 ]VG_TC VG_032 CL1-A-1 0 20# /dev/sdu11u SER = 62486 LDEV = 74 [ FIBRE FCTBL = 3 ]VG_TC VG_033 CL1-A-1 0 21# /dev/sdu12u SER = 62486 LDEV = 22 [ FIBRE FCTBL = 3 ]VG_TC VG_034 CL1-A-1 0 22# /dev/sdu13u SER = 62486 LDEV = 76 [ FIBRE FCTBL = 3 ]VG_TC VG_035 CL1-A-1 0 23# /dev/sdu10u SER = 62486 LDEV = 20 [ FIBRE FCTBL = 3 ]VG_SI VG_014 CL1-A-1 0 20# /dev/sdu11u SER = 62486 LDEV = 74 [ FIBRE FCTBL = 3 ]VG_SI VG_015 CL1-A-1 0 21# /dev/sdu12u SER = 62486 LDEV = 22 [ FIBRE FCTBL = 3 ]VG_SI VG_015 CL1-A-1 0 22# /dev/sdu13u SER = 62486 LDEV = 76 [ FIBRE FCTBL = 3 ]VG_SI VG_016 CL1-A-1 0 23

HORCM_INST#dev_group ip_address serviceVG_TC 123.45.78.51 20331VG_TC 123.45.78.115 20331VG_SI 123.45.80.51 20332VG_SI 123.45.80.115 20332

Example 2-44 Example of CCI Configuration Definition File at Remote Site(Instance 16)



HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu30u SER = 62486 LDEV = 30 [ FIBRE FCTBL = 3 ]VG_SI VG_014 CL1-A-1 0 30# /dev/sdu31u SER = 62486 LDEV = 84 [ FIBRE FCTBL = 3 ]VG_SI VG_015 CL1-A-1 0 31# /dev/sdu32u SER = 62486 LDEV = 41 [ FIBRE FCTBL = 3 ]


VG_SI VG_016 CL1-A-1 0 32# /dev/sdu33u SER = 62486 LDEV = 86 [ FIBRE FCTBL = 3 ]VG_SI VG_017 CL1-A-1 0 33

HORCM_INST#dev_group ip_address serviceVG_SI 123.45.80.51 20331VG_SI 123.45.80.115 20331

Example 2-45 Example of CCI Configuration Definition File at Remote Site(Instance 17)

Setting the CCI User Environmental Variable

In the following procedure, the environment variables HORCMINST andHORCC_MRCF are corrected corresponding to a system configuration. Thissetup is performed on both the node or virtual server to which TrueCopy P-VOL is connected, and the node or virtual server in which S-VOL will be used.You must perform these operations in both nodes of the main site and in bothnodes of the remote site.

1. Set up the environment variable of the Command Control Interfaceinstance:sudo horcsetenv HORCMINST 16 (For instance number 16)

or sudo horcsetenv HORCMINST 17 (For instance number 17)

2. Set up the HOMRCF command environment variable of the CommandControl Interface as TrueCopy:sudo horcunsetenv HORCC_MRCF

3. If you have logged in using SSH and set up as explained in steps 1 and 2,confirm the setup by logging out once and then logging back in:exitssh nasroot@fixed-IP-address-or-service-IP-address

Enter the command shown in the following example to check the result ofsetting up the environment variable:$ sudo horcprintenvThe following table shows the settings of the environment variables just afterHNAS F is installed.

Table 2-14 Settings of Environment Variables just after the Installation ofHNAS F

Environment Variables Setting

HORCMINST 16 (for a node)

HORCC_MRCF No setting.

Create File System in TrueCopy P-VOL

Create a file system in TrueCopy P-VOL using the Create New File Systemwindow of File Services Manager or by using the fscreate command. Even if


you create and split the TrueCopy volume pair without creating a file systemin the TrueCopy P-VOL, you cannot access the TrueCopy S-VOL in thesecondary site.

Overview of TrueCopy OperationsThis section describes the overview of operations, the CCI commands, andthe commands provided by the HNAS F for the TrueCopy operations ofvolume replication and disaster recovery.

We describe only the arguments of the CCI commands which are required forthe basic TrueCopy operations. For other arguments, please refer to theHitachi Command Control Interface User and Reference Guide. For thecommands provided by the HNAS F products, see Commands that HNAS Fprovides on page 2-132.


Volume Replication

This section describes the procedure of volume replication. Figure 2-10Overview of Volume Replication and the Corresponding Sections on page2-77 shows an overview of volume replication and the corresponding sections.When the pair is in the PSUS state, the S-VOL can be accessed at the remotesite.

Figure 2-10 Overview of Volume Replication and the CorrespondingSections

Starting TrueCopy Operations and Creating a TrueCopy Pair

If the TrueCopy S-VOL contains a file system, you must delete the file systemusing the fsdelete command in File Services Manager before startingTrueCopy operations. If the differential-data storage device of the file


snapshot functionality is in the TrueCopy S-VOL, you must release it usingthe syncstop command before starting the TrueCopy operations.

To start TrueCopy operations and create a TrueCopy volume pair:

1. At the main site and the remote site, start CCI.sudo horcmstart.sh (1-instance configuration)


2. At the remote site, reserve device file used as a target file system.sudo horcvmdefine –d Device file number,...

3. At the main site, create TrueCopy volume pair.sudo paircreate {-g group-name|-d volume-name} –f never –vl

4. At the main site, check the completion of TrueCopy volume pair creation.sudo pairvolchk {-g group-name|-d volume-name}Note:


a. pairvolchk: Volstat is P-VOL.[status = COPY]=> Creating

b. pairvolchk: Volstat is P-VOL.[status = PAIR] => Created

Splitting a TrueCopy Volume Pair

The method for splitting a TrueCopy volume pair depends on whether youperform an offline backup or an online backup. For an offline backup, the P-VOL is unmounted to split the pair. For an online backup, instead ofunmounting the P-VOL, updates to the file system are temporarily stopped tosplit the pair.

During offline backup, a TrueCopy volume pair is split after completelystopping access from clients by deleting NFS/CIFS shares. Since an I/O erroris reported to the application if NFS/CIFS shares are deleted while theapplication writes data in P-VOL, or if the application tries to write data in P-VOL after deleting NFS/CIFS shares, it can be distinguished by the applicationdata reflected in the TrueCopy volume pair. It is therefore applicable to mostapplications.

During online backup, however, a TrueCopy volume pair is split withoutdeleting NFS/CIFS shares. Since an I/O error is not reported to theapplication when writing data to P-VOL, it cannot be distinguished of whichtime data is reflected in S-VOL. For this reason, it is applicable only to theapplication which can identify where data was updated with a journal file etc.

To split the TrueCopy volume pair during offline backup:

TrueCopy Volume Pair Split by Offline Backup (when the file systemis not operated by the file snapshot functionality or when thedifferential-data snapshot is not published in the shared file system


even though the file system is operated by the file snapshotfunctionality)

1. At the main site, stop the program that accesses P-VOL. Then, unmountthe NFS shares on the client side.

2. At the main site, delete NFS/CIFS shares in the P-VOL using thenfsdelete command and the cifsdelete command in File ServicesManager. Unmount the P-VOL using the fsumount command.

3. At the main site, prevent the file snapshot functionality from performingoperations on the P-VOL:sudo horcfreeze –f source-file-system-name

4. At the main site, split the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} -rw

5. At the main site, verify that the TrueCopy volume pair is split.sudo pairvolchk {-g group-name|-d volume-name}Note:


a. pairvolchk: Volstat is P-VOL.[status = COPY]=> Splitting

b. pairvolchk: Volstat is P-VOL.[status = PSUS]=> Split

6. At the main site, enable the operations from the file snapshotfunctionality on the P-VOL.sudo horcunfreeze –f source-file-system-name

7. At the main site, mount P-VOL using the fsmount command in FileServices Manager, and create NFS/CIFS shares using the nfscreatecommand and the cifscreate command. At the main site, mount theNFS shares in the P-VOL on the client side. Then, restart the programwhich accesses the P-VOL.

8. At the main site, mount the P-VOL NFS shares at a client-side and restartthe program which accesses the P-VOL.

9. At the remote site, connect the target file system to the node or virtualserver.



For an LVM and non-tiered file systemsudo horcvmimport -f target-file-system-name -d device-file-number,...

For an LVM and tiered file systemsudo horcvmimport -f target-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...


10. At the remote site, mount S-VOL using the fsmount command. CreateNFS/CIFS shares using the nfscreate command and the cifscreatecommand.

Note:When the S-VOL is the file system managed by the file snapshotfunctionality, mount the differential-data snapshots using thesyncmount command, and create the NFS/CIFS shares using thenfscreate command and the cifscreate command, if necessary.

11. At the remote site, start the program that accesses the S-VOL.

TrueCopy Volume Pair Split by Offline Backup (when the file systemis operated by the file snapshot functionality and the differential-data snapshot is published in the shared file system)

1. At the main site, stop the programs that access the P-VOL. Then,unmount the NFS shares on the client side.

2. At the main site, unmount the differential-data snapshot using thesyncumount command, delete the NFS/CIFS shares using the nfsdeletecommand and the cifsdelete command in File Services Manager, andunmount the P-VOL using the fsumount command.

3. At the main site, prevent the file snapshot functionality from performingoperations on the P-VOL:sudo horcfreeze –f source-file-system-name

4. At the main site, split the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} -rw

5. At the main site, verify that the TrueCopy volume pair is split.$ sudo pairvolchk {-g group-name | -d volume-name }Note:


a. pairvolchk : Volstat is P-VOL.[status = COPY]=> Splitting

b. pairvolchk : Volstat is P-VOL.[status = PSUS]=> Split

6. At the main site, enable the operations from the file snapshotfunctionality on the P-VOL.sudo horcunfreeze –f source-file-system-name

7. At the main site, mount the P-VOL using the fsmount command in FileServices Manager, and set the NFS/CIFS shares using the nfscreatecommand and the cifscreate command. Then, mount the differential-data snapshots using the syncmount command.

8. At the main site, mount the P-VOL NFS shares on the client-side andrestart the programs that access the P-VOL.


For a non-tiered file system


$ sudo horcvmimport -f target-file-system-name -d device-file-number,...

For a tiered file system$ sudo horcvmimport -f target-file-system-name --tier1 device-file-number,... --tier2 device-file-number,...

10. At the remote site, mount the S-VOL using the fsmount command. Then,set the NFS/CIFS shares using the nfscreate command and thecifscreate command.

11. At the remote site, mount the differential-data snapshot using thesyncmount command to make it public in the shared file system.

12. At the remote site, start the programs that access the S-VOL.

To split the TrueCopy volume pair during online backup:

TrueCopy Volume Pair Split by Online Backup (when the file systemis not operated by the file snapshot functionality or when thedifferential-data snapshot is not published in the shared file systemeven though the file system is operated by the file snapshotfunctionality)

1. At the main site, prevent the file snapshot functionality from performingoperations on P-VOL, prohibit the access from the client, and write theunreflected data to the disk.sudo horcfreeze –f Source-file-system-name

2. At the main site, split the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} –rw

3. At the main site, verify that the TrueCopy volume pair has been split.sudo pairvolchk {-g group-name|-d volume-name}Note:




4. At the main site, permit the access from the client to P-VOL and enablethe operations from the file snapshot functionality on P-VOL.sudo horcunfreeze –f Source-file-system-name







6. At the remote site, mount the S-VOL using the fsmount command.

Create NFS/CIFS shares using the nfscreate command and thecifscreate command.


7. At the remote site, start the program that accesses the S-VOL.

TrueCopy Volume Pair Split by Online Backup (when the file systemis operated by the file snapshot functionality and the differential-data snapshot is published in the shared file system)

1. At the main site, prevent the file snapshot functionality from performingoperations on P-VOL, prohibit the access from the client, and write theunreflected data to the disk.sudo horcfreeze –f Source-file-system-name

2. At the main site, split the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} –rw

3. At the main site, verify that the TrueCopy volume pair has been split.sudo pairvolchk {-g group-name|-d volume-name}Note:




4. At the main site, permit client access to the P-VOL and enable operationsfrom the file snapshot functionality on the P-VOL.sudo horcunfreeze –f Source-file-system-name


For a non-tiered file system$ sudo horcvmimport -f target-file-system-name -d device-file-number,...

For a tiered file system


$ sudo horcvmimport -f target-file-system-name --tier1 device-file-number,… --tier2 device-file-number,…

6. At the remote site, mount the S-VOL using the fsmount command. Then,set the NFS/CIFS shares using the nfscreate command and thecifscreate command.


8. At the remote site, restart the programs that access the S-VOL.

Resynchronizing a TrueCopy Volume Pair

You must separate the target file system using the horcexport commandbefore resynchronizing the volume pair. However, in spite of separating thetarget file system, resynchronizing the volume pair is not the same asperforming initial copy.

Note:When the S-VOL is the file system managed by the file snapshotfunctionality, unmount the differential-data snapshots using thesyncumount command before using the horcexport command.

To resynchronize the TrueCopy volume pair:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not madepublic in the shared file system even though the file system isoperated by the file snapshot functionality

1. At the remote site, stop the program that accesses the S-VOL. Then,unmount the NFS shares on the client side.

2. At the remote site, use the nfsdelete command and the cifsdeletecommand to delete the NFS/CIFS share in the S-VOL. Then, use thefsumount command to unmount.

Note:In the file snapshot functionality target file system, delete the NFS/CIFS shares for the differential-data snapshot by using the nfsdeletecommand and the cifsdelete command, unmount the differential-data snapshot by using the syncumount command.

3. At the remote site, separate the file system of the S-VOL by using thehorcexport command.sudo horcexport -f target-file-system-name

4. At the main site, resume the TrueCopy volume pair.sudo pairresync {-g group-name|-d volume-name}

5. At the main site, verify that the TrueCopy volume pair has been resumed.sudo pairvolchk {-g group-name|-d volume-name}Note:



a. pairvolchk : Volstat is P-VOL.[status = COPY]=> Resuming

b. pairvolchk : Volstat is P-VOL.[status = PAIR]=> Resumingcompleted

When the file system is operated by the file snapshot functionalityand the differential-data snapshot is made public in the shared filesystem

1. At the remote site, stop the programs that access the S-VOL. Then,unmount the NFS shares on the client side.

2. At the remote site, unmount the differential-data snapshot using thesyncumount command.

Note:When NFS/CIFS shares are created in a differential-data snapshot thatis made public in the shared file system, you must cancel the NFS/CIFS shares for the differential-data snapshot using the nfsdeleteand cifsdelete commands before unmounting it.

3. At the remote site, release the NFS/CIFS shares using the nfsdelete andcifsdelete commands, and unmount the file system using the fsumountcommand.

4. At the remote site, separate the S-VOL file system using the horcexportcommand.sudo horcexport -f target-file-system-name

5. At the main site, restore the TrueCopy volume pair.sudo pairresync {-g group-name|-d volume-name

6. At the main site, verify that the TrueCopy volume pair has been restored.sudo pairvolchk {-g group-name|-d volume-name}Note:



b. pairvolchk : Volstat is P-VOL.[status = PAIR]=> Resumingcompletes

Deleting a TrueCopy Volume Pair

Completing TrueCopy operations by deleting the TrueCopy volume pair in thePSUS state will differ depending upon whether you will continue using the filesystem in S-VOL, or if you will destroy the file system.

When deleting the TrueCopy volume pair in a state other than PSUS, youcannot use that file system because the consistency of data in the S-VOL isnot guaranteed.


To delete the TrueCopy volume pair in PSUS state, and to continue using theS-VOL file system:

1. At the main site, delete the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} –S

2. At the main site, verify that the TrueCopy volume pair has been deleted.sudo pairvolchk {-g group-name|-d volume-name}a. pairvolchk: Volstat is P-VOL.[status = COPY]=> Deleting

b. pairvolchk: Volstat is P-VOL.[status = SMPL]=> Deleted

3. At the main site and the remote site, stop CCI.sudo horcmshutdown.sh (1-instance configuration)

or sudo horcmshutdown.sh 16 17 (2-instance configuration)

Note:If you have started the task of splitting the TrueCopy volume pair via anoffline backup but have not yet completed steps 1 through 11, completesteps 1 through 11 before proceeding with the above steps. When usingan online backup, you need to have completed steps 1 through 7 ofsplitting the TrueCopy volume pair before proceeding with the abovesteps (see Splitting a TrueCopy Volume Pair on page 2-78).

To delete the TrueCopy volume pair in PSUS state without using the S-VOLfile system (when the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not made public in theshared file system even though the file system is operated by the filesnapshot functionality):

1. At the remote site, terminate the program which accesses the S-VOL.

2. At the remote site, delete NFS/CIFS shares using the nfsdeletecommand and the cifsdelete command, and unmount S-VOL using thefsumount command.

Note:When the S-VOL is the file system managed by the file snapshotfunctionality, delete the NFS/CIFS shares under the differential-datasnapshots using the nfsdelete command and cifsdelete command,unmount the differential-data snapshots using the syncumountcommand, and release the differential-data storage devices using thesyncstop command.

3. At the remote site, delete the S-VOL file system using the fsdeletecommand.




6. At the main site and the remote site, stop CCI.


sudo horcmshutdown.sh (1-instance configuration)


Note:If you have started the task of splitting the TrueCopy volume pair via anoffline backup but have not yet completed steps 1 through 11, completesteps 1 through 11 before proceeding with the above steps. When usingan online backup, you need to have completed steps 1 through 7 ofsplitting the TrueCopy volume pair before proceeding with the abovesteps (see Splitting a TrueCopy Volume Pair on page 2-78).

To delete the TrueCopy volume pair in the PSUS status without using the S-VOL file system (when the file system is operated by the file snapshotfunctionality or when the differential-data snapshot is made public in theshared file system):

1. At the remote site, stop the programs that access the S-VOL.

2. At the remote site, unmount the differential-data snapshot using thesyncumount command.


3. At the remote site, release the differential-data storage device using thesyncstop command.

4. At the remote site, release the NFS/CIFS shares using the nfsdelete andcifsdelete commands, and unmount the file system using the fsumountcommand.

5. At the remote site, delete the S-VOL file system using the fsdeletecommand.

6. At the main site, delete the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} -S

7. At the main site, verify that the TrueCopy volume pair has been deleted.sudo pairvolchk {-g group-name|-d volume-name}a. pairvolchk : Volstat is P-VOL.[status = COPY]=> Deleting

b. pairvolchk : Volstat is P-VOL.[status = SMPL]=> Deleted



Note:If you start the split pair operation with offline backup, but you have notyet connected the target file system to the node at the remote site, finishconnecting the target file system to the node, and then start thepreviously described operations. If you start the split pair operation withonline backup, but you have not yet connected the target file system to


the node at the remote site, finish connecting the target file system to thenode, and then start the above operations (see Splitting a TrueCopyVolume Pair on page 2-78).

To delete the TrueCopy volume pair in a state other than PSUS state:




3. At the remote site, release the device file used in the target file system.sudo horcvmdelete –d Device-file-number,...


or sudo horcmshutdown.sh 16 17 (2-instances configuration)

Disaster Recovery Operations

This section describes the procedures for disaster recovery operations.

Switching Operations to the Remote Site

This section describes the procedure of disaster recovery up to the pointwhen operations are switched to the remote site. Perform these operations onthe remote site only.

To switch operations to the remote site (when the file system is notoperated by the file snapshot functionality or when the differential-data snapshot is not made public in the shared file system):

Note:The response of the takeover may change according to the condition ofthe failed main site.

1. Execute SVOL-Takeover using the horctakeover command.sudo horctakeover {-g group-name|-d volume-name}

2. Connect the target file system to the node or virtual server.



For an LVM and non-tiered file system


sudo horcvmimport -f target-file-system-name -d device-file-number,...


3. Mount S-VOL using the fsmount command and create NFS/CIFS sharesusing the nfscreate command and the cifscreate command.

Recover the file system when mounting the S-VOL.


4. Start the program which accesses the S-VOL.

To switch operations to the remote site (when the file system isoperated by the file snapshot functionality and the differential-datasnapshot is made public in the shared file system):

1. Execute SVOL-Takeover using the horctakeover command.sudo horctakeover {-g group-name|-d volume-name}

2. Connect to the node or virtual server of the target file system.




Restore the file system when mounting the S-VOL.


5. Start the programs that access the S-VOL.

Transferring the Data Back to the Main Site

Below is the procedure for transferring data back to the main site duringdisaster recovery.

To transfer the data back to the main site (when the file system isnot operated by the file snapshot functionality or when the


differential-data snapshot is not made public in the shared filesystem):

1. At the main site, set up the environment variable of the CommandControl Interface instance:sudo horcsetenv HORCMINST 16 (For instance number 16)


2. At the main site, set up the HOMRCF command environment variable of theCommand Control Interface as TrueCopy:sudo horcunsetenv HORCC_MRCF

3. After you have logged in using SSH and set up as explained in steps 1and 2, confirm the setup by logging out and relogging in.exitssh nasroot@fixed-IP-address-or-service-IP-address

4. At the main site and at the remote site, start CCI.sudo horcmstart.sh (1-instance configuration)

or sudo horcmstart.sh 16 17 (2-instances configuration)

5. At the main site, unmount the NFS share in the old P-VOL at the clientside.

6. At the main site, delete the NFS/CIFS shares in the old P-VOL, unmountthe old P-VOL, and separate the file system in the old P-VOL.

Note:When the old P-VOL is the file system managed by the file snapshotfunctionality, delete the NFS/CIFS shares under the differential-datasnapshots, unmount the differential-data snapshots before deletingthe file system in the old P-VOL.

7. At the remote site, delete the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} –S

8. At the remote site, confirm the deletion of the TrueCopy volume pair hasbeen completed.sudo pairvolchk {-g group-name|-d volume-name}a. If pairvolchk : Volstat is P-VOL.[status = COPY] is displayed,

the delete operation is in progress.

b. If pairvolchk: Volstat is P-VOL. [status = SMPL] is displayed,the delete operation has been completed.

9. Transfer the data from the remote site to the main site (Create TrueCopyvolume pair).sudo paircreate {-g group-name|-d volume-name} –f never –vl

10. If the state is changed to PAIR, stop the operation in the secondary site.

11. At the remote site, delete the NFS/CIFS shares in the old S-VOL using thenfsdelete command and the cifsdelete command, and unmount theold S-VOL using the fsumount command.

12. At the remote site, prevent the file snapshot functionality from performingoperations on the old S-VOL.


sudo horcfreeze –f source-file-system-name13. At the remote site, split TrueCopy volume pair.

sudo pairsplit {-g group-name|-d volume-name} –rw 14. At the remote site, enable the operations from the file snapshot

functionality on the old S-VOL.sudo horcunfreeze –f source-file-system-name

15. At the main site, connect the target file system to the node.





16. At the remote site, separate the file system in the old S-VOL using thehorcexport command.sudo horcexport -f target-file-system-nameNote:

When the old S-VOL is the file system managed by the file snapshotfunctionality, delete the NFS/CIFS shares under the differential-datasnapshots before deleting the file system in the old S-VOL.

17. Perform reverse resynchronization of data from the main site to theremote site.sudo pairresync {-g group-name|-d volume-name} –swaps

18. At the main site, mount the new P-VOL using the fsmount command, andcreate the NFS/CIFS shares using the nfscreate command and thecifscreate command.

Note:When the new P-VOL is the file system managed by the file snapshotfunctionality, mount the differential-data snapshots using thesyncmount command, and create the NFS/CIFS shares using thenfscreate command and the cifscreate command, if necessary.

19. Mount the NFS share in the P-VOL at the client side.

20. At the main site, resume the operations in the primary site.


To transfer the data back to the main site (when the file system isoperated by the file snapshot functionality and the differential-datasnapshot is made public in the shared file system):

1. At the main site, set the environmental variable for CCI instance(s).sudo horcsetenv HORCMINST 16 (in case of instance number 16)

or sudo horcsetenv HORCMINST 17 (in case of instance number 17)

2. At the main site, set the environmental variable for the CCI HOMRCFcommand to "use as ShadowImage".sudo horcunsetenv HORCC_MRCF

3. If you have logged in using SSH and set up as explained in steps 1 and 2,log out and then log in again to validate the settings.exitssh nasroot@fixed-IP-address-or-service-IP-address

4. Start CCI at the main site and the remote site.sudo horcmstart.sh (1-instance configuration)


5. At the main site, unmount the NFS share in the old P-VOL at the clientside.

6. At the main site, unmount the differential-data snapshot using thesyncumount command.


7. At the main site, release the NFS/CIFS shares of the old P-VOL, unmountthe old P-VOL, and separate the file system.sudo horcexport -f source-file-system-name

8. At the remote site, delete the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name}

9. At the remote site, confirm that the TrueCopy volume pair has beendeleted.sudo pairvolchk {-g group-name|-d volume-name}a. pairvolchk : Volstat is P-VOL.[status = COPY] => Deleting


10. Transfer the data from the remote site to the main site (Create TrueCopyvolume pair).sudo paircreate {-g group-name|-d volume-name} –f never –vl

11. When the state changes to PAIR, stop the operation in the secondary site.

12. Execute the syncumount command to unmount the differential datasnapshot from the node or virtual server that is connected to the old S-VOL at the secondary site.

Note:


When NFS/CIFS shares are created in a differential-data snapshot thatis made public in the shared file system, you must cancel the NFS/CIFS shares for the differential-data snapshot using the nfsdeleteand cifsdelete commands before unmounting it.

13. At the remote site, release the NFS/CIFS shares of the old S-VOL, andunmount the old S-VOL using the fsumount command.

14. At the remote site, prevent the file snapshot functionality from performingoperations on the S-VOL:sudo horcfreeze -f source-file-system-name

15. At the remote site, split the TrueCopy volume pair.sudo pairsplit {-g group-name|-d volume-name} -rw

16. At the remote site, enable the operations from the file snapshotfunctionality on the S-VOL:sudo horcunfreeze -f source-file-system-name

17. At the main site, connect the target file system to the node or virtualserver.



18. At the remote site, delete the file system of the S-VOL by using thehorcexport command.sudo horcexport -f target-file-system-name

19. Perform reverse resynchronization of data from the main site to theremote site.sudo pairresync {-g group-name|-d volume-name} –swaps

20. At the main site, mount the new P-VOL using the fsmount command, andcreate the NFS/CIFS shares using the nfscreate command and thecifscreate command.

21. On the node or virtual server connected to the new P-VOL of the mainsite, mount the differential-data snapshot using the syncmount commandto make it public in the shared file system.

22. Mount the NFS share in the P-VOL at the client side.

23. Resume operations at the main site.

Notes on Failures of External VolumesIf a trouble (including a temporary trouble, for example, the cable connectionhas cut out) occurs to the external volume which is currently used withTrueCopy volume pair, perform the following processes.

1. If the external volume can not be used succeedingly, delete TrueCopyvolume pair by using the pairsplit -S command.


If the S-VOL is not connected to the node or virtual server, release thedevice file used in the target file system by using the horcvmdeletecommand.

2. Confirm all the TrueCopy volume pair statuses that are using the externalvolume with trouble(s) by using the pairdisplay command.

If the status of the TrueCopy volume pair is PSUE, delete the volume pairby using the pairsplit -S command.

If you delete TrueCopy volume pair when the S-VOL is not connected tothe node or virtual server, release the device file used in the target filesystem by using the horcvmdelete command.

How to use Universal Replicator on the HNAS FThis section describes how to use Universal Replicator with the HNAS F.

Overview of Universal Replicator Operations in HNAS F systemThe HNAS F system enables integrated management of the data in thestorage system. HNAS F utilizes the LAN environment already in place andenables the data within a storage system to be shared across heterogeneousplatforms. HNAS F also enables you to copy and maintain the data stored inthe storage system using the volume replication feature Universal Replicator.With Universal Replicator, you can copy data utilizing the file system server(even in the remote center), back up data to guard against failures in primarydata volumes, and perform disaster recovery at the primary site.

To use Universal Replicator in the HNAS F system, use Command ControlInterface (CCI). Note that if you install OS on the HNAS F system, CCI isinstalled at the same time. You do not need to install CCI in the HNAS Fsystem.

Figure 2-11 Using Universal Replicator in the HNAS F on page 2-93illustrates the Universal Replicator operation.

Figure 2-11 Using Universal Replicator in the HNAS F


Scope of Universal Replicator Function in the HNAS F system

Scope Related to Universal Replicator

Volume type

Only the user LUs may be used as Universal Replicator primary data volumes(hereinafter abbreviated as P-VOLs) and secondary data volumes (hereinafterabbreviated as S-VOLs) in a HNAS F system.

You cannot specify the OS disks and shared LU of a virtual server as aUniversal Replicator P-VOL or S-VOL.

Data Overflow Watch

If you use the HNAS F system with Universal Replicator, Specify the timeoutperiod (data overflow watch) within the range of 0 to 20 seconds as a journalgroup option. The I/O timeout may occur earlier than the timeout by the dataoverflow of the journal when data overflow watch is adjusted to 20 secondsor more, and the file system will become blocked.For details of the setting ofdata overflow watch, refer to the Hitachi Universal Replicator User's Guide(User Guide).

Platforms which can access the Universal Replicator P-VOLs or S-VOLs

Only clients that are connected to a network via a HNAS F system can accessthe Universal Replicator P-VOLs and S-VOLs created in an HNAS F system.Hosts connected via serial port or fibre-channel port will not be able to accessthem.

File systems which can be allocated in the Universal Replicator P-VOL

You cannot allocate a file system that consists of 129 or more LUs and thatuses Logical Volume Manager (LVM) on the OS.

You cannot allocate a volume group that consists of 129 or more LUs if itcontains a file system managed by the file snapshot functionality anddifferential-data storage devices.

Writing data to S-VOL when a volume pair is split

When you split a Universal Replicator volume pair in the HNAS F system, youneed to change the S-VOL to write-enable. Make sure that you always specifythe –rw option instead of the -r option when issuing the CCI pairsplitcommand.

To protect the data in the split S-VOL from being written to by the client, usethe fsmount command with –r option in File Services Manager whenmounting the file system in S-VOL.


Note on path settings

If you use the HNAS F system with Universal Replicator, you must add pathsto ports of both CL1 and CL2 to connect the primary and secondary sites.

Scope Related to Features in Backup Restore


If you resynchronize a Universal Replicator pair defined by Volume Replicationfunction after performing a backup of the S-VOL by using the backupmanagement software, a full backup will be acquired next time.



When you copy a file system managed by the file snapshot functionality usingUniversal Replicator, you must copy both the LUs that constitute the filesystem and the LUs which constitute the differential-data storage devices. Ifyou copy only the LUs which constitute a file system, you will not be able toconnect them to the HNAS F system on the secondary site.


Prerequisites


You need a workstation or PC to log in to the HNAS F system using secureshell (SSH) in addition to the prerequisite hardware for Universal Replicator,CCI, and the HNAS F system described in the following guides:



Manuals related to storage system

¢ Hitachi Universal Replicator User and Reference Guide


¢ Hitachi ShadowImage User's Guide (User Guide) (when using theShadowImage cascade function)




To use Universal Replicator on the HNAS F system, all of the followingprogram products must be installed in the storage system to which the HNASF system is connected, and their licenses must be valid.


• TrueCopy

• ShadowImage (Required when configuring the cascade connection ofUniversal Replicator/ShadowImage)

Notes on Operations



Limitations on Universal Replicator operations due to the status ofcluster, nodes, and resource groups

When a cluster is not configured, the cluster is stopped, the nodes arestopped, and the resource groups are offline, connecting the device file orcreating and mounting a file system is restricted. Due to these restrictions,the following operations performed in the target Universal Replicatoroperation will also end in error. You therefore should not operate the cluster,nodes, and resource groups during the Universal Replicator operation. Shouldany problems occur with the cluster, nodes, or resource groups, fix themimmediately.

• Unmount and mount the source file system during the splitting of aUniversal Replicator pair volume.


• Unmount and delete the target file system before the Universal Replicatorpair is resynchronized.

Notes on occurrence of a failover during operating on node


Notes on changing system configuration during the UniversalReplicator operations

When changing the following system configurations during UniversalReplicator operations, you must change the Command Control Information


configuration definition files on the node and on each node in the clusterwhere the S-VOL is used:

• Changing a fixed IP address of the node.

• Expanding or deleting a source file system

• Setting up, expanding or releasing a differential-data storage device forthe file snapshot functionality

When the host name is specified in the Command Control Interfaceconfiguration definition file and you change the following systemconfigurations, you must change the configuration definition file:

• Editing the /etc/hosts file (when resolving the host name using the /etc/hosts file)



• Changing node names (host names)


In case of periodic maintenance of external volumes, you must delete allUniversal Replicator pairs once.

Notes on splitting a Universal Replicator pair with online backup

If it takes a long time between the horcfreeze command execution and thehorcunfreeze command execution, access timeout may occur with someclients. If the file snapshot functionality operates on the source file system,the horcfreeze command execution may take a long time, and this increasesthe possibility of timeout.

Verifying whether the access to the file system is suspended

After executing of the horcfreeze command and the horcunfreezecommand, you can use the fsctl command to verify whether the accessfrom the client to the file system is suspended.







If a Universal Replicator pair is split and connected to the node while thedifferential data storage device in the copy source file system does not have


sufficient capacity, certain measures must be taken in the copy source filesystem to solve the lack of the capacity.

In splitting a Universal Replicator pair, confirm that the differential datastorage device in the copy source file system is in the normal status.


The single-instancing setting is not copied over copy-destination file systems.If the single-instancing setting is enabled for a copy-source file system, aftersplitting a Universal Replicator pair and connecting to a node, use the fseditcommand to enable the single-instancing setting for the copy-destination filesystem. Also, set the policy for duplicate file capacity reduction.

Before splitting a Universal Replicator pair, make sure that the duplicate filecapacity reduction policy has not yet been executed on the copy source filesystem. If a Universal Replicator pair is split during policy execution,resynchronize the Universal Replicator pair and perform its subsequentoperations again.


If the copy-source file system is a tiered file system, the LUs in all the filesystems that make up the tier must be assigned the same device group namein the configuration definition file.

To connect the copy-source file system to a node or a virtual server, specifythe --tier1 and --tier2 options for the horcimport command or thehorcvmimport command if the copy file system is a tiered file system.Otherwise, specify the -d option.




LUs connected to FC paths are automatically recognized.







A node or virtual server cannot be connected to if Universal Replicator is usedto copy a WORM file system.


If the local data encryption functionality is being used, the copy-destinationfile system can be connected to the node only when the copy-source filesystem and the copy-destination file system are in the same cluster. If thefile systems are in different clusters, the copy-destination file system cannotbe connected to the node.

Preparing for Universal Replicator Operations

Preparing for Universal Replicator Volume Pair Operation

Prepare for creating the Universal Replicator volume pairs. For details of thepreparation required on the storage system side, refer to the HitachiUniversal Replicator User's Guide (User Guide).

Registration of Public Key Used for SSH

Before issuing the commands described in this document, the public key usedfor SSH needs to be registered in the node or virtual server in which theUniversal Replicator P-VOL is connected and also in the node or virtual serverwhich will use the S-VOL. You can register the public key from the AccessProtocol Configuration window, in the Add Public Key page.

Configuring the CCI Environment

Login to the Target node

Using the nasroot account via SSH, log in both the node in which theUniversal Replicator P-VOL is connected, and the node which will use the S-VOL. (For information about login, refer to the appropriate documentation ofthe SSH communication software.)



For using additional instance numbers, the environment for the instancenumbers to be used must be configured by using the horcsetconf command.


$ sudo horcsetconf -i instance number







To control a Universal Replicator pair using CCI, you must first define theUniversal Replicator pair using the CCI configuration definition file.

A Command Control Interface configuration definition file template is madeavailable after the installation of HNAS F.

The template of the Command Control Interface configurationdefinition file is:/home/nasroot/horcm<instance number>.confFor nodes, <instance number> is 16 or 17. Additional instance numbers are from 20 to 499.Use the CCI horcconfedit and mkconf.sh commands to add the HORCM_MON,HORCM_CMD, HORCM_DEV, and HORCM_INST sections to the template. After that,finish editing the CCI configuration definition file.

You must perform these operations in both nodes of the main site and in bothnodes of the remote site. You therefore need to prepare a total of four CCIconfiguration files, assuming one instance per node. For two instances, youwill need a total of eight configuration files.

In a HNAS F, you can define one or two CCI instances per one node. Tooperate only the Universal Replicator pairs in CCI, you need one instance. Tooperate the pairs where Universal Replicator and ShadowImage are cascaded,you need two instances.

In the following section, we explain how to create a CCI configurationdefinition file using the HNAS F system with LUs as an example (see Figure2-12 Example of Configuration of Pair LU on page 2-101).




Use the CCI mkconf.sh command to define the HORCM_MON, HORCM_CMD,HORCM_DEV, and HORCM_INST sections in the CCI configuration definition filetemplate.$ ls /dev/sdu*u | sudo mkconf.sh -gg device-group-name -i 16



Before executing the mkconf.sh command, you must create a file system tobe paired. The LU configuration (size and number) in the file system to bepaired needs to be exactly the same in both the primary site and thesecondary site.

To create the correct CCI configuration definition file, it is recommended thatyou create a file system to be temporarily paired before executing the


mkconf.sh command. After creating the CCI configuration definition file usingthe mkconf.sh command, you can continue using the file system in theprimary site. Alternatively, you may also delete it and create a file systemwith the same configuration using the same LU in Create File System inUniversal Replicator P-VOL on page 2-112. You must delete the temporarilycreated file system in the secondary site before creating a pair.

Example 2-50 Example of mkconf.sh Execution Result (for Instance Number16) on page 2-102 shows an example of how to create a template for a CCIconfiguration definition file.We will explain how to create a CCI configurationdefinition file in node0 at the primary site in the sample configuration above.You must create CCI configuration definition files for the other nodes usingthe previously explained procedure.$ ls /dev/sdu*u | sudo mkconf.sh -gg VG -i 16starting HORCM inst 16HORCM inst 16 starts successfully.HORCM Shutdown inst 16 !!!A CONFIG file was successfully completed.starting HORCM inst 16HORCM inst 16 starts successfully.DEVICE_FILE Group PairVol PORT TARG LUN M SERIAL LDEV/dev/sdu00u VG VG_000 CL1-A-1 0 17 - 62486 70/dev/sdu01u VG VG_001 CL1-A-1 1 18 - 62486 18 : /dev/sdu10u VG VG_010 CL1-A-1 0 10 - 62486 10/dev/sdu11u VG VG_011 CL1-A-1 0 11 - 62486 64/dev/sdu12u VG VG_012 CL1-A-1 0 12 - 62486 12/dev/sdu13u VG VG_013 CL1-A-1 0 13 - 62486 66/dev/sdu14u VG VG_014 CL1-A-1 0 14 - 62486 14/dev/sdu15u VG VG_015 CL1-A-1 0 15 - 62486 68/dev/sdu16u VG VG_016 CL1-A-1 0 16 - 62486 16HORCM Shutdown inst 16 !!!Please check '/home/nasroot/horcm16.conf','/home/nasroot/log16/curlog/horcm_*.log', andmodify ‘ip_address & service'.#

Example 2-50 Example of mkconf.sh Execution Result (for InstanceNumber 16)


Example 2-51 Changing the HORCM_CMD definition in the configurationdefinition file (for instance number 16)

Editing the CCI configuration definition file.

The following table shows the values are specified for the items included inthe CCI configuration definition file in the HNAS F system as shown in Table2-15.


Table 2-15 Configuration Definition File Settings (HORCM_MON) andSpecified Values in HNAS F system

Section Name Item Specified Values in HNAS F

HORCM_MON ip_address Fixed IP address of the local node.







Based on Table 2-15, change each entry of ip_address in HORCM_MON to anappropriate value.

Change the entries poll and timeout to the appropriate values according tohardware requirements.HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.78.51 20331 1000 3000


HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu00u SER = 62486 LDEV = 70 [ FIBRE FCTBL = 3 ]VG VG_000 CL1-A-1 0 0# /dev/sdu01u SER = 62486 LDEV = 18 [ FIBRE FCTBL = 3 ]VG VG_001 CL1-A-1 0 1 : :# /dev/sdu13u SER = 62486 LDEV = 66 [ FIBRE FCTBL = 3 ]VG VG_013 CL1-A-1 0 13# /dev/sdu14u SER = 62486 LDEV = 14 [ FIBRE FCTBL = 3 ]VG VG_014 CL1-A-1 0 14# /dev/sdu15u SER = 62486 LDEV = 68 [ FIBRE FCTBL = 3 ]VG VG_015 CL1-A-1 0 15# /dev/sdu16u SER = 62486 LDEV = 16 [ FIBRE FCTBL = 3 ]


VG VG_016 CL1-A-1 0 16



Next, delete unnecessary LU entries (lines) for the HORCM_DEV section exceptfor the LU entry (line) that you want to control using CCI (see Example 2-52Example of CCI Configuration Definition File - 1 (for Instance Number 16) onpage 2-104).

LUs which constitute a file system and their LDEV numbers can be reviewedusing the following command.$ sudo horcdevlist | grep ':File System Name$'For example, in Example 2-53 Example of the Command Execution whichLists LUs which Constitute the File System Sample on page 2-104, 11, 12,13 are LUs that constitute the file system sample. 64, 12, and 66 are theirLDEV numbers.$ sudo horcdevlist | grep ':sample$'11 62486 64 OPEN-V 3.906GB -- -- - Normal File:sample12 62486 12 OPEN-V 3.906GB -- -- - Normal File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal File:sample

Example 2-53 Example of the Command Execution which Lists LUs whichConstitute the File System Sample

The device file name and the device name of the LU entry (line) you want tocontrol using Command Control Interface are changed to an appropriatename. In changing the device file name and the device name, observe thefollowing:

• The same device group name and the same device name need to bespecified to the volumes to be paired in the primary site and in thesecondary site. Specify the device group name and the device nameaccordingly.

• The same device group name must be specified as LUs which constituteone file system. If the file snapshot functionality manages the file system,the same device group name must be specified as LUs that constitute afile system and as LUs that constitute differential-data storage devices.








For one LU, two lines of the HORCM_DEV section information are output: thecomment line starting with a # and the definition line right below it. Use the /dev/sdu**u (where ** is the LU number) and LDEV = ** (where ** is theLDEV number) text output in the comment lines to identify the necessaryentries.

Next, specify the IP address of the instance to be paired in the secondary sitefor the HORCM_INST section. You operate CCI on node, as a preparation forfailover, specify the IP addresses of the instances in both nodes.

Table 2-16 Configuration Definition File Settings (HORCM_INST) andSpecified Values in the HNAS F system


HORCM_INST ip_address Fixed IP address or virtual IP address of the node in thesecondary site of Universal Replicator



• 20332 (If the instance number is 17.)










Checking the contents of the CCI configuration definition file.


First, start CCI in both the node or virtual server in which the UniversalReplicator P-VOL is connected and in the node or virtual server which will usethe S-VOL.$ sudo horcsetenv HORCMINST 16 (For instance number 16) or sudo horcsetenv HORCMINST 17 (For instance number 17)$ sudo horcunsetenv HORCC_MRCFWhen you logged in using SSH and performed the above setup, confirm the setup by once logging out and relogging in again.$ sudo horcmstart.sh


By issuing the pairdisplay command in the node or virtual server in whichUniversal Replicator P-VOL or S-VOL is connected or in which S-VOL will beused, you can see the LDEV numbers of LUs specified in the HORCM_DEVsection.$ sudo pairdisplay –g Device-Group-Name

Example 2-57 How to Check LDEV Numbers of LUs Specified in theHORCM_DEV Section

You can check the device file numbers, and the LDEV numbers for the devicefiles that constitute a file system, by issuing the horcdevlist command inthe node or virtual server in which P-VOL is connected. Compare this withExample 2-57 How to Check LDEV Numbers of LUs Specified in theHORCM_DEV Section on page 2-106.$ sudo horcdevlist | grep ':sample$'11 62486 64 OPEN-V 3.906GB -- -- - Normal File:sample


12 62486 12 OPEN-V 3.906GB -- -- - Normal File:sample13 62486 66 OPEN-V 3.906GB -- -- - Normal File:sample


You can also check the device file numbers and the LDEV numbers for theunused device files which can be S-VOL by issuing the horcdevlistcommand in the node or virtual server in which S-VOL will be used. Comparethis with Example 2-57 How to Check LDEV Numbers of LUs Specified in theHORCM_DEV Section on page 2-106.$ sudo horcdevlist | grep ' Free$'21 62486 74 OPEN-V 3.906GB -- -- - Normal Free22 62486 22 OPEN-V 3.906GB -- -- - Normal Free23 62486 76 OPEN-V 3.906GB -- -- - Normal Free


When specifying port names in the HORCM_DEV section of the CCIconfiguration definition file, use the name of the storage system FibreChannel port that connects to the node.

After the CCI definition files has been validated, stop CCI on the nodes orvirtual servers connected to the Universal Replicator P-VOL and S-VOL.# sudo horcmshutdown.sh 16

Example 2-60 Stopping CCI (for Instance Number 16)

Save the definition information to save the configured CCI configurationdefinition files. For details about saving definition information, please refer tothe Hitachi NAS Platform F1000 Series Cluster Administrator's Guide.

An example of the configuration definition file for a tiered system

An example of the configuration definition file for a tiered system is providedbelow.



Example 2-61 How To Check the Device File Number and the LU Number ofa P-VOL





VG_TC 123.45.80.51 20331VG_TC 123.45.80.115 20331

Example 2-62 Example of the CCI Configuration Definition File (for P-VOLs)


HORCM_CMD#dev_name dev_name dev_name#UnitID 0 (Serial# 62486)\\.\CMD-62486:/dev/sdHORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10 SER = 62486 LDEV =20 [ FIBRE FCTBL = 3 ]VG_ VG_010 CL1-A-1 0 20# /dev/sdu11u SER = 62486 LDEV =74 [ FIBRE FCTBL = 3 ]VG_ VG_011 CL1-A-1 0 21# /dev/sdu12u SER = 62486 LDEV =22 [ FIBRE FCTBL = 3 ]VG_ VG_012 CL1-A-1 0 22# /dev/sdu13u SER = 62486 LDEV =76 [ FIBRE FCTBL = 3 ]VG_ VG_013 CL1-A-1 0 23

HORCM_INST#dev_group ip_address serviceVG_TC 123.45.80.115 20331VG_TC 123.45.80.51 20331


Cascade Configuration of Universal Replicator and ShadowImage

In the HNAS F system, you may use the Universal Replicator pair andShadowImage pair in a cascade configuration. In a cascade configuration, youmay prepare for disaster scenario where, for example, you would not be ableto recover the file system from the Universal Replicator S-VOL, by backing upthe file system copied from the Universal Replicator P-VOL to S-VOLperiodically using ShadowImage.

Figure 2-14 Example of Cascade Configuration of Universal Replicator andShadowImage on page 2-110 shows an example of a cascade configuration ofUniversal Replicator and ShadowImage.


Figure 2-14 Example of Cascade Configuration of Universal Replicator andShadowImage

By preparing the CCI configuration definition file as shown in Example 2-64Example of CCI Configuration Definition File at Primary Site (Instance 16) onpage 2-111, Example 2-65 Example of CCI Configuration Definition File atSecondary Site (Instance 16) on page 2-111, and Example 2-66 Example ofCCI Configuration Definition File at Secondary Site (Instance 17) on page2-112, you may operate the cascade configuration of Universal Replicator andShadowImage from CCI.HORCM_MON#ip_address service poll(10ms) timeout(10ms)123.45.78.51 20331 1000 3000


HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10u SER = 62486 LDEV = 10 [ FIBRE FCTBL = 3 ]VG_TC VG_032 CL1-A-1 0 10# /dev/sdu11u SER = 62486 LDEV = 64 [ FIBRE FCTBL = 3 ]VG_TC VG_033 CL1-A-1 0 11# /dev/sdu12u SER = 62486 LDEV = 12 [ FIBRE FCTBL = 3 ]VG_TC VG_034 CL1-A-1 0 12# /dev/sdu13u SER = 62486 LDEV = 66 [ FIBRE FCTBL = 3 ]VG_TC VG_035 CL1-A-1 0 13



VG_TC 123.45.80.51 20331VG_TC 123.45.80.115 20331

Example 2-64 Example of CCI Configuration Definition File at Primary Site(Instance 16)


HORCM_CMD#dev_name dev_name dev_name#UnitID 0 (Serial# 62490)/dev/sdf

HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu10u SER = 62486 LDEV = 20 [ FIBRE FCTBL = 3 ]VG_TC VG_032 CL1-A-1 0 20# /dev/sdu11u SER = 62486 LDEV = 74 [ FIBRE FCTBL = 3 ]VG_TC VG_033 CL1-A-1 0 21# /dev/sdu12u SER = 62486 LDEV = 22 [ FIBRE FCTBL = 3 ]VG_TC VG_034 CL1-A-1 0 22# /dev/sdu13u SER = 62486 LDEV = 76 [ FIBRE FCTBL = 3 ]VG_TC VG_035 CL1-A-1 0 23# /dev/sdu10u SER = 62486 LDEV = 20 [ FIBRE FCTBL = 3 ]VG_SI VG_014 CL1-A-1 0 20# /dev/sdu11u SER = 62486 LDEV = 74 [ FIBRE FCTBL = 3 ]VG_SI VG_015 CL1-A-1 0 21# /dev/sdu12u SER = 62486 LDEV = 22 [ FIBRE FCTBL = 3 ]VG_SI VG_015 CL1-A-1 0 22# /dev/sdu13u SER = 62486 LDEV = 76 [ FIBRE FCTBL = 3 ]VG_SI VG_016 CL1-A-1 0 23

HORCM_INST#dev_group ip_address serviceVG_TC 123.45.78.51 20331VG_TC 123.45.78.115 20331VG_SI 123.45.80.51 20332VG_SI 123.45.80.115 20332

Example 2-65 Example of CCI Configuration Definition File at SecondarySite (Instance 16)



HORCM_DEV#dev_group dev_name port# TargetID LU# MU## /dev/sdu30u SER = 62486 LDEV = 30 [ FIBRE FCTBL = 3 ]VG_SI VG_014 CL1-A-1 0 30# /dev/sdu31u SER = 62486 LDEV = 84 [ FIBRE FCTBL = 3 ]VG_SI VG_015 CL1-A-1 0 31# /dev/sdu32u SER = 62486 LDEV = 41 [ FIBRE FCTBL = 3 ]VG_SI VG_016 CL1-A-1 0 32


# /dev/sdu33u SER = 62486 LDEV = 86 [ FIBRE FCTBL = 3 ]VG_SI VG_017 CL1-A-1 0 33

HORCM_INST#dev_group ip_address serviceVG_SI 123.45.80.51 20331VG_SI 123.45.80.115 20331

Example 2-66 Example of CCI Configuration Definition File at SecondarySite (Instance 17)

Setting the CCI User Environmental Variable

In the following procedure, the environment variables HORCMINST andHORCC_MRCF are corrected corresponding to a system configuration. Thissetup is performed on both the node or virtual server to which UniversalReplicator P-VOL is connected, and the node or virtual server in which S-VOLwill be used. When you operate CCI on node, you must perform theseoperations in both nodes of the main site and in both nodes of the remotesite.

1. Set up the environment variable of the Command Control Interfaceinstance:sudo horcsetenv HORCMINST 16 (For instance number 16)


2. Set up the HOMRCF command environment variable of the CommandControl Interface as Universal Replicator:sudo horcunsetenv HORCC_MRCF

3. If you have logged in using SSH and set up as explained in steps 1 and 2,confirm the setup by logging and then logging back in:exitssh nasroot@fixed-IP-address-or-service-IP-address

Enter the command shown in the following example to check the result ofsetting up the environment variable:$ sudo horcprintenvThe following table shows the values of environment variables that are setimmediately after installation of HNAS F.

Table 2-17 Values of Environment Variables Immediately After Installationof HNAS F

Environment Variables Value

HORCMINST 16 for node

HORCC_MRCF None

Create File System in Universal Replicator P-VOL

Create a file system in Universal Replicator P-VOL using the Create New FileSystem window of File Services Manager or by using the fscreate command.


Even if you create and split the Universal Replicator volume pair withoutcreating a file system in the Universal Replicator P-VOL, you cannot accessthe Universal Replicator S-VOL in the secondary site.

Overview of Universal Replicator OperationsThis section describes the overview of operations, the CCI commands, andthe commands provided by the HNAS F for the Universal Replicatoroperations of volume replication and disaster recovery.

We describe only the arguments of the CCI commands which are required forthe basic Universal Replicator operations. For other arguments, please referto the Hitachi Command Control Interface User and Reference Guide. For thecommands provided by the HNAS F products, see Commands that HNAS Fprovides on page 2-132.


Volume Replication

This section describes the procedure of volume replication. Figure 2-15Overview of Volume Replication on page 2-114 shows an overview of volumereplication. When the pair is in the PSUS status, the S-VOL can be accessedat the secondary site.

• For information about how to create volume pairs, see Starting UniversalReplicator Operations and Creating a Universal Replicator Pair on page2-114.

• For information about how to split volume pairs, see Splitting a UniversalReplicator Volume Pair on page 2-115.

• For information about how to resynchronize volume pairs, seeResynchronizing a Universal Replicator Volume Pair on page 2-119.

• For information about how to delete volume pairs, see Deleting aUniversal Replicator Volume Pair on page 2-121.


Figure 2-15 Overview of Volume Replication

Starting Universal Replicator Operations and Creating a UniversalReplicator Pair

If the Universal Replicator S-VOL contains a file system, you must delete thefile system using the fsdelete command in File Services Manager beforestarting Universal Replicator operations. If the differential-data storage deviceof the file snapshot functionality is in the Universal Replicator S-VOL, youmust release it using the syncstop command before starting the UniversalReplicator operations.

To start Universal Replicator operations and create a Universal Replicatorvolume pair:

1. At the primary site and the secondary site, start CCI.sudo horcmstart.sh (1-instance configuration)


2. At the secondary site, reserve device file used as a target file system.sudo horcvmdefine -d Device file number,...

3. At the primary site, create Universal Replicator volume pair.sudo paircreate {-g group-name|-d volume-name} -f async -vl -jp master-journal-group-ID -js restore-journal-group-ID

4. At the primary site, check the completion of Universal Replicator volumepair creation.sudo pairvolchk {-g group-name|-d volume-name}Note:


a. pairvolchk: Volstat is P-VOL.[status = COPY] => Creating

b. pairvolchk: Volstat is P-VOL.[status = PAIR] => Created


Splitting a Universal Replicator Volume Pair

The method for splitting a Universal Replicator volume pair depends onwhether you perform an offline backup or an online backup. For an offlinebackup, the P-VOL is unmounted to split the pair. For an online backup,instead of unmounting the P-VOL, updates to the file system are temporarilystopped to split the pair.

During offline backup, a Universal Replicator volume pair is split aftercompletely stopping access from clients by deleting NFS/CIFS shares. Sincean I/O error is reported to the application if NFS/CIFS shares are deletedwhile the application writes data in P-VOL, or if the application tries to writedata in P-VOL after deleting NFS/CIFS shares, it can be distinguished by theapplication data reflected in the Universal Replicator volume pair. It istherefore applicable to most applications.

During online backup, however, a Universal Replicator volume pair is splitwithout deleting NFS/CIFS shares. Since an I/O error is not reported to theapplication when writing data to P-VOL, it cannot be distinguished of whichtime data is reflected in S-VOL. For this reason, it is applicable only to theapplication which can identify where data was updated with a journal file etc.

To split the Universal Replicator volume pair during offline backup:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not publishedin the shared file system even though the file system is operated bythe file snapshot functionality

1. At the primary site, stop the program that accesses P-VOL and unmountNFS shares at the client side.

2. At the primary site, delete NFS/CIFS shares in the P-VOL using thenfsdelete command and the cifsdelete command in File ServicesManager. Unmount the P-VOL using the fsumount command.

3. At the primary site, prevent the file snapshot functionality fromperforming operations on the P-VOL:sudo horcfreeze –f source-file-system-name

4. At the primary site, split the Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} -rw

5. At the primary site, verify that the Universal Replicator volume pair hasbeen split.sudo pairvolchk {-g group-name|-d volume-name}Note:


a. pairvolchk : Volstat is P-VOL.[status = COPY]=>Splitting

b. pairvolchk : Volstat is P-VOL.[status = PSUS]=>Splitcompletes


6. At the primary site, enable the operations from the file snapshotfunctionality on the P-VOL.sudo horcunfreeze –f source-file-system-name

7. At the primary site, mount P-VOL using the fsmount command in FileServices Manager, and create NFS/CIFS shares using the nfscreatecommand and the cifscreate command.

8. At the primary site, mount NFS shares of the P-VOL at the client side andrestart the program which accesses the P-VOL.

9. At the secondary site, connect the target file system to the node or virtualserver.





10. At the secondary site, mount S-VOL using the fsmount command. CreateNFS/CIFS shares using the nfscreate command and the cifscreatecommand.


11. At the secondary site, start the program that accesses the S-VOL.


1. At the primary site, stop the program that accesses P-VOL and unmountNFS shares at the client side.

2. At the primary site, delete NFS/CIFS shares in the P-VOL using thenfsdelete command and the cifsdelete command in File ServicesManager. Unmount the P-VOL using the fsumount command.


4. At the primary site, split the Universal Replicator volume pair.


sudo pairsplit {-g group-name|-d volume-name} -rw5. At the primary site, verify that the Universal Replicator volume pair has

been split.sudo pairvolchk {-g group-name|-d volume-name}Note:





7. At the primary site, mount P-VOL using the fsmount command in FileServices Manager, and create NFS/CIFS shares using the nfscreatecommand and the cifscreate command, and then mount thedifferential-data snapshots using the syncmount command.

8. At the primary site, mount NFS shares of the P-VOL at the client side andrestart the program which accesses the P-VOL.





11. At the secondary site, mount the differential-data snapshot using thesyncmount command, and will be made public the NFS/CIFS shares.


To split the Universal Replicator volume pair during online backup:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not publishedin the shared file system even though the file system is operated bythe file snapshot functionality

1. At the primary site, prevent the file snapshot functionality fromperforming operations on P-VOL, stop access from clients, and write theunreflected data to the disk.sudo horcfreeze –f Source-file-system-name

2. At the primary site, split the Universal Replicator volume pair.


sudo pairsplit {-g group-name|-d volume-name} –rw 3. At the primary site, verify that the Universal Replicator volume pair has

been split.sudo pairvolchk {-g group-name|-d volume-name}Note:




4. At the primary site, permit access from clients to P-VOL and enable theoperations from the file snapshot functionality on P-VOL.sudo horcunfreeze –f Source-file-system-name






6. At the secondary site, mount the S-VOL using the fsmount command.





2. At the primary site, split the Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} -rw


3. At the primary site, verify that the Universal Replicator volume pair hasbeen split.sudo pairvolchk {-g group-name|-d volume-name}Note:







For a tiered file system$ sudo horcvmimport -f target-file-system-name --tier1 device-file-number,… --tier2 device-file-number,…




Resynchronizing a Universal Replicator Volume Pair

You must separate the target file system using the horcexport commandbefore resynchronizing the volume pair. However, in spite of separating thetarget file system, resynchronizing the volume pair is not the same asperforming initial copy.

Note:When the S-VOL is the file system managed by the file snapshotfunctionality, unmount the differential-data snapshots before using thehorcexport command.

To resynchronize the Universal Replicator volume pair:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not made


public in the shared file system even though the file system isoperated by the file snapshot functionality

1. At the secondary site, terminate the program which accesses the S-VOLand unmount NFS shares at the client side.

2. At the secondary site, delete NFS/CIFS shares by using the nfsdeletecommand and the cifsdelete command, and perform unmounting byusing the fsumount command.

Note:For the file snapshot functionality target file system, delete the NFS/CIFS shares for the differential-data snapshots by using thenfsdelete command and the cifsdelete command, unmount thedifferential-data snapshots by using the syncumount command.

3. At the secondary site, delete the S-VOL file system by using thehorcexport command.sudo horcexport -f source-file-system-name

4. At the primary site, resume the Universal Replicator volume pair.sudo pairresync {-g group-name|-d volume-name}

5. At the primary site, verify that the Universal Replicator volume pair hasbeen resumed.sudo pairvolchk {-g group-name|-d volume-name}Note:





1. At the secondary site, terminate the program which accesses the S-VOLand unmount NFS shares at the client side.

2. At the secondary site, un-mount the differential-data snapshot using thesyncumount command.




4. At the secondary site, delete the file system of the secondary data volumeusing the horcexport command.sudo horcexport -f target-file-system-name

5. At the primary site, resume the Universal Replicator volume pair.sudo pairresync {-g group-name|-d volume-name}

6. At the primary site, verify that the Universal Replicator volume pair hasbeen resumed.sudo pairvolchk {-g group-name|-d volume-name}Note:




Deleting a Universal Replicator Volume Pair

Completing Universal Replicator operations by deleting the UniversalReplicator volume pair in the PSUS status will differ depending upon whetheryou will continue using the file system in S-VOL, or if you will destroy the filesystem.

When deleting the Universal Replicator volume pair in a status other thanPSUS, you cannot use that file system because the consistency of data in theS-VOL is not guaranteed.

To delete the Universal Replicator volume pair in PSUS status, and tocontinue using the S-VOL file system:

1. At the primary site, delete the Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} –S

2. At the primary site, verify that the Universal Replicator volume pair hasbeen deleted.sudo pairvolchk {-g group-name|-d volume-name}a. pairvolchk: Volstat is P-VOL.[status = COPY]=> Deleting


3. At the primary site and the secondary site, stop CCI.sudo horcmshutdown.sh (1-instance configuration)


Note:If you have started the task of splitting the Universal Replicator volumepair via an offline backup but have not yet completed steps 1 through 11,complete steps 1 through 11 before proceeding with the above steps.


When using an online backup, you need to have completed steps 1through 7 of splitting the Universal Replicator volume pair beforeproceeding with the above steps (see Splitting a Universal ReplicatorVolume Pair on page 2-115).

To delete the Universal Replicator volume pair in PSUS status without usingthe S-VOL file system:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not madepublic in the shared file system even though the file system isoperated by the file snapshot functionality

1. At the secondary site, terminate the program which accesses the S-VOL.

2. At the secondary site, delete NFS/CIFS shares using the nfsdeletecommand and the cifsdelete command, and unmount S-VOL using thefsumount command.

Note:When the S-VOL is the file system managed by the file snapshotfunctionality, delete the NFS/CIFS shares under the differential-datasnapshots using the nfsdelete command and cifsdelete command,unmount the differential-data snapshots using the syncumountcommand, and release the differential-data storage devices using thesyncstop command.

3. At the secondary site, delete the S-VOL file system using the fsdeletecommand.






Note:If you have started the task of splitting the Universal Replicator volumepair via an offline backup but have not yet completed steps 1 through 11,complete steps 1 through 11 before proceeding with the above steps.When using an online backup, you need to have completed steps 1through 7 of splitting the Universal Replicator volume pair beforeproceeding with the above steps (see Splitting a Universal ReplicatorVolume Pair on page 2-115).



1. At the secondary site, terminate the program which accesses the S-VOL.

2. At the secondary site, un-mount the differential-data snapshot using thesyncumount command.


3. At the secondary site, release the device which stores the differential datausing the syncstop command.


5. At the secondary site, delete the file system of the secondary data volumeusing the fsdelete command.






Note:If you have started the task of splitting the Universal Replicator volumepair via an offline backup but have not yet completed steps 1 through 11,complete steps 1 through 11 before proceeding with the above steps.When using an online backup, you need to have completed steps 1through 7 of splitting the Universal Replicator volume pair beforeproceeding with the above steps (see Splitting a Universal ReplicatorVolume Pair on page 2-115).

To delete the Universal Replicator volume pair in a status other than PSUSstatus:



2. At the primary site, verify that the Universal Replicator volume pair hasbeen deleted.sudo pairvolchk {-g group-name|-d volume-name}pairvolchk: Volstat is P-VOL.[status = COPY]=> Deleting

pairvolchk: Volstat is P-VOL.[status = SMPL]=> Deleted

3. At the secondary site, release the device file used in the target filesystem.sudo horcvmdelete –d Device-file-number,...


or sudo horcmshutdown.sh 16 17 (2-instances configuration)

Disaster Recovery Operations

This section describes the procedures for disaster recovery operations.

Switching Operations to the Secondary Site

To recover the data from a disaster:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not madepublic in the shared file system


1. Execute SVOL-Takeover using the horctakeover command.sudo horctakeover {-g group-name|-d volume-name} -t timeout-period













1. Execute SVOL-Takeover using the horctakeover command.sudo horctakeover {-g group-name|-d volume-name} -t timeout-period








Transferring the Data Back to the Primary Site

To transfer the data back to the primary site:

When the file system is not operated by the file snapshotfunctionality or when the differential-data snapshot is not madepublic in the shared file system

1. At the primary site, set up the environment variable of the CommandControl Interface instance:sudo horcsetenv HORCMINST 16 (For instance number 16)



2. At the primary site, set up the HOMRCF command environment variable ofthe Command Control Interface as Universal Replicator:sudo horcunsetenv HORCC_MRCF


4. At the primary site and at the secondary site, start CCI.sudo horcmstart.sh (1-instance configuration)


5. At the primary site, unmount the NFS shares in the old P-VOL from theclient side.

6. At the primary site, delete the NFS/CIFS shares in the old P-VOL,unmount the old P-VOL, and separate the file system in the old P-VOL.sudo horcexport -f source-file-system-nameNote:

When the old P-VOL is the file system managed by the file snapshotfunctionality, delete the NFS/CIFS shares under the differential-datasnapshots, unmount the differential-data snapshots before deletingthe file system in the old P-VOL.

7. At the secondary site, delete the Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} –S

8. At the secondary site, verify that the Universal Replicator volume pair hasbeen deleted.sudo pairvolchk {-g group-name|-d volume-name}a. pairvolchk: Volstat is P-VOL.[status = COPY]=> Deleting


9. At the secondary site, execute the following command to transfer the datafrom the secondary site to the primary site (Create Universal Replicatorvolume pair).sudo paircreate {-g group-name|-d volume-name} –f async –vl -jp former-restore-journal-group-ID -js former-master-journal-group-ID

10. If the status is changed to PAIR, stop the operation in the secondary site.

11. At the secondary site, delete the NFS/CIFS shares in the old S-VOL usingthe nfsdelete command and the cifsdelete command, and unmountthe old S-VOL using the fsumount command.

12. At the secondary site, prevent the file snapshot functionality fromperforming operations on the old S-VOL.sudo horcfreeze –f Source-file-system-name

13. At the secondary site, split Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} –rw

14. At the secondary site, enable the operations from the file snapshotfunctionality on the old S-VOL.


sudo horcunfreeze –f Source-file-system-name15. At the primary site, connect the target file system to the node.





16. At the secondary site, delete the file system in the old S-VOL using thehorcexport command.sudo horcexport -f target-file-system-nameNote:

When the old S-VOL is the file system managed by the file snapshotfunctionality, delete the NFS/CIFS shares under the differential-datasnapshots, unmount the differential-data snapshots before deletingthe file system in the old S-VOL.

17. At the primary site, enter the following command to reverseresynchronization of data from the primary site to the secondary site.sudo pairresync {-g group-name|-d volume-name} –swaps

18. At the primary site, mount the new P-VOL using the fsmount command,and create the NFS/CIFS shares using the nfscreate command and thecifscreate command.

Note:When the new P-VOL is the file system managed by the file snapshotfunctionality, mount the differential-data snapshots using thesyncmount command, and create the NFS/CIFS shares using thenfscreate command and the cifscreate command, if necessary.

19. Mount the NFS shares of the P-VOL at a client-side.

20. At the primary site, resume the operations in the primary site.


1. At the primary site, set up the environment variable of the CommandControl Interface instance:sudo horcsetenv HORCMINST 16 (For instance number 16)



2. At the primary site, set up the HOMRCF command environment variable ofthe Command Control Interface as Universal Replicator:sudo horcunsetenv HORCC_MRCF


4. At the primary site and at the secondary site, start CCI.sudo horcmstart.sh (1-instance configuration)


5. At the primary site, unmount the NFS shares in the old P-VOL from theclient side.

6. At the primary site, un-mount the differential-data snapshot using thesyncumount command.


7. At the primary site, delete NFS/CIFS shares of the old P-VOL, andperform unmounting, and then, separate the file system.sudo horcexport -f source-file-system-name

8. At the secondary site, delete the Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} –S

9. At the secondary site, verify that the Universal Replicator volume pair hasbeen deleted.sudo pairvolchk {-g group-name|-d volume-name}a. pairvolchk: Volstat is P-VOL.[status = COPY]=> Deleting


10. At the secondary site, execute the following command to transfer the datafrom the secondary site to the primary site (Create Universal Replicatorvolume pair).sudo paircreate {-g group-name|-d volume-name} –f async –vl -jp former-restore-journal-group-ID -js former-master-journal-group-ID

11. If the status is changed to PAIR, stop the operation in the secondary site.

12. Execute the syncumount command to unmount the differential-datasnapshot from the node or virtual server connected to the old S-VOL atthe secondary site..

Note:When the differential-data snapshot is not only make public in theshared file system but also create the NFS/CIFS shares to the


differential-data snapshot, it is required to cancel the NFS/CIFSshares for the differential-data snapshot using the nfsdelete andcifsdelete commands before un-mounting the differential-datasnapshot.

13. At the secondary site, delete the NFS/CIFS shares in the old S-VOL usingthe nfsdelete command and the cifsdelete command, and unmountthe old S-VOL using the fsumount command.

14. At the secondary site, prevent the file snapshot functionality fromperforming operations on the old S-VOL.sudo horcfreeze –f Source-file-system-name

15. At the secondary site, split Universal Replicator volume pair.sudo pairsplit {-g group-name|-d volume-name} –rw

16. At the secondary site, enable the operations from the file snapshotfunctionality on the old S-VOL.sudo horcunfreeze –f Source-file-system-name

17. At the primary site, connect the target file system to the node or virtualserver.



18. At the secondary site, separate the file system in the old S-VOL using thehorcexport command.sudo horcexport -f target-file-system-name

19. At the primary site, enter the following command to reverseresynchronization of data from the primary site to the secondary site.sudo pairresync {-g group-name|-d volume-name} –swaps

20. At the primary site, mount the new P-VOL using the fsmount command,and create the NFS/CIFS shares using the nfscreate command and thecifscreate command.

21. From the node or virtual server connected to the new P-VOL at theprimary site, execute the syncmount command to mount the differential-data snapshot and make file shares on the file system accessible.

22. Mount the NFS shares of the P-VOL at a client-side.

23. At the primary site, resume the operations in the primary site.

Notes on Failures of External VolumesIf a trouble (including a temporary trouble, for example, the cable connectionhas cut out) occurs to the external volume which is currently used withUniversal Replicator volume pair, perform the following processes.


1. If the external volume cannot be used succeedingly, delete the UniversalReplicator volume pair by using the pairsplit -S command.

If the S-VOL is not connected to a node or virtual server, execute thehorcvmdelete command to release the device files of the copy-target filesystem.

2. Confirm all the Universal Replicator volume pair statuses that are usingthe external volume with trouble(s) by using the pairdisplay command.

If the status of the Universal Replicator volume pair is PSUE, delete thevolume pair by using the pairsplit -S command. To delete a UniversalReplicator volume pair, if the S-VOL is not connected to a node or virtualserver, execute the horcvmdelete command to release the device files ofthe copy-target file system.

Operations of the Log FilesThe HVFP system allows you to manage failure information by viewing,downloading, or deleting log files. The log files contain detailed informationabout the execution of commands supported by the HNAS F system. If youwant to investigate the cause of a failure, look at the log files. If you cannotresolve a failure on your own, download the log files and contact maintenancepersonnel.

For details about how to manage failure information, see the ClusterTroubleshooting Guide.

Operations of the Command Control Interface Log FilesThis section explains the format of the Command Control Interface log files inthe HNAS F system, and the method of downloading log files to the SSHclient, and provides notes on the operations of the Command ControlInterface log files.

Format of the Command Control Interface Log Files

This section explains the format of the Command Control Interface log files(which consist of starting logs, error logs, trace logs, and core files). Theselog files are listed in Table 2-18. In Table 2-18, the * (asterisk) shows theinstance number of CCI; HOST shows the host name of the correspondingnode; PID shows the process ID of CCI or the Command Control Interfacecommand; CMD shows the process name (horcmgr in case of CCI, or acommand name in case of the Command Control Interface command); andTIME shows the creation time of the core file.


Table 2-18 Format of the Command Control Interface Log Files in theHNAS F system

Log FileClassification Log File Name Format of Log Files

CCI logs underoperation

CCI starting log /home/nasroot/log*/curlog/horcm_HOST.log

Command log /home/nasroot/log*/horcc_HOST.log

CCI error log /home/nasroot/log*/curlog/horcmlog_HOST/horcm.log

CCI trace log /home/nasroot/log*/curlog/horcmlog_HOST/horcm_PID.trc

Command trace /home/nasroot/log*/curlog/horcmlog_HOST/horccc_PID.trc

Core file /var/core/core-PID-CMD-TIME

CCI logs savedautomatically

CCI starting log /home/nasroot/log*/tmplog/horcm_HOST.log

Command log /home/nasroot/log*/horcc_HOST.log

CCI error log /home/nasroot/log*/tmplog/horcmlog_HOST/horcm.log

CCI trace log /home/nasroot/log*/tmplog/horcmlog_HOST/horcm_PID.trc

Command trace /home/nasroot/log*/tmplog/horcmlog_HOST/horccc_PID.trc

Core file /var/core/core-PID-CMD-TIME

Downloading the Command Control Interface Log Files to the Client

If you want to check log files (except for core files) to identify the causes ofCCI command errors, or if you want to save log files before deleting the logfiles, you can download the log files by specifying the Backup log group onthe List of RAS Information page (for Batch-download) page on the Checkfor Errors window of File Services Manager.

You can download or delete core files output by CCI on the List of RASInformation page (for List of core files) on the Check for Errorswindow provided in File Services Manager. For details, see the Hitachi NASPlatform F1000 Series Cluster Administrator's Guide.

Note:If you read the Command Control Interface log files using Windows™,you must ensure that the text viewer you are using can display the textcontaining the line feed code LF (Line Feed).

Notes on the Operations of the Command Control Interface Log Files

Since the Command Control Interface log files are output to OS disk, ifoperations still show an error, log files will become large and the availablespace on OS disk will decrease. You should periodically check the capacity of


the log files (except the trace files) to which Command Control Interfacerestricts capacity, stop CCI, and delete log files using the horclogremovecommand before exceeding 1 M bytes. Check the capacity of log files usingthe following command.$ ls -l -R /home/nasroot/log*(* means an instance number)

Commands that HNAS F providesThis section describes the commands that are used for replication operationsin an HNAS F system and that are available in HNAS F.

• horcdevlist• horclogremove• horcprintenv, horcsetenv, horcunsetenv• horcfreeze• horcunfreeze• horcvmdefine• horcvmdelete• horcimport, horcvmimport, horcexport• horcsetconf, horcunsetconf, horcconflist• cifscreate, cifsdelete• fscreate, fsdelete, fslist, fsmount, fsumount• nfscreate, nfsdelete• fsctl• lumapctl• synclist, syncmount, syncstop, syncumount


3Volume Management Functions

This section describes the program products providing volume managementfunctions that can be used in conjunction with the HNAS F.

□ Dynamic Provisioning

□ Dynamic Tiering

□ Universal Volume Manager

□ Volume Migration

□ Volume Shredder

□ Encryption License Key

Volume Management Functions 3-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Dynamic ProvisioningDynamic Provisioning is a program product that helps you reduce the cost ofdeployment and volume management in storage system. This is achieved byusing virtual volumes (V-VOLs).

Before using the Dynamic Provisioning functionality, make sure youunderstand it by carefully reading the Provisioning Guide.

The use of Dynamic Provisioning together with HNAS F allows you to reducedeployment costs by virtualizing volumes via HNAS F. Because this eliminatesthe need to re-create a file system each time you run out of storage space,there are fewer interruptions to system operation, meaning lowermanagement costs and less downtime.

Note:

¢ To use Dynamic Provisioning to manage the user LUs and clustermanagement LU in an HNAS F system, the user LUs must be in adifferent pool from the cluster management LU.

¢ You cannot use a virtual volume for an OS LU of an HNAS F virtualserver.

¢ File systems created as V-VOLs by Dynamic Provisioning on the HNASF will become blocked if the pool volume overflows. For this reason,prevent the pool volume from overflowing by setting the appropriatethresholds for the pool volume. Immediately provide additional drivesto increase the pool volume capacity when the thresholds areexceeded.

¢ To monitor the thresholds of pool volumes, you need to set up SNMPtrap notification to be sent out whenever the thresholds are exceeded.

Dynamic TieringDynamic Tiering is a program that helps reduce storage costs. You canimprove storage cost performance by using Dynamic Tiering to configurevolumes with different types of storage drives.

Before using the functionality of Dynamic Tiering, make sure you understandit by carefully reading the Provisioning Guide.

When Dynamic Tiering is used with an HNAS F system, data is automaticallyand optimally placed in storage tiers depending on access frequency. Thisconsiderably reduces the burden on administrators to design systems toincrease storage performance. Storage costs can also be reduced by usinginexpensive disks, while still maintaining storage performance.

Note:

¢ To use Dynamic Tiering to manage the user LUs and clustermanagement LU in an HNAS F system, the user LUs must be in adifferent pool from the cluster management LU.

3-2 Volume Management FunctionsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

¢ You cannot use a virtual volume for an OS LU of an HNAS F virtualserver.

Universal Volume ManagerUniversal Volume Manager is a program product that provides storagevirtualization for storage systems. When Universal Volume Manager is used,an external storage system connected to the storage system via a FibreChannel interface can be treated as part of the storage system.

Before you use Universal Volume Manager, carefully read the HitachiUniversal Volume Manager User's Guide (User Guide), and make sure thatyou understand the program functions.

You can also use Universal Volume Manager in conjunction with the HNAS Fto make external storage system available to the storage system.

Note:An external volume controlled by Universal Volume Manager can be usedas a user LU in the HNAS F, but cannot be used as a system LU.

Note:If you create an HNAS F file system in an external volume controlled byUniversal Volume Manager, and the LU of the external storage systembecomes unreachable due to an error in the external storage system orother reason, the file system will become blocked. To restore the blockedfile system, first resolve the error in the external storage system, andthen use Universal Volume Manager to unblock the external volume. Thefile system will be usable again after you restart the OS on the HNAS F.

Troubleshooting for the HNAS F Including External Storage SystemIn this section, the procedure to stop the external storage system on purposefor the maintenance and the procedure to recover the external storagesystem from the failure are described for the HNAS F system including theexternal storage system. This section consists of the following sections:

• Stopping and Restarting External Storage System on Purpose on page3-3

• Recovery Procedure in Case of Error in External Storage System on page3-5

Stopping and Restarting External Storage System on Purpose

Figure 3-1 An Example of HNAS F System Configuration Including an ExternalStorage System on page 3-4 shows an example of an HNAS F configurationincluding an external storage system. The procedure to stop the externalstorage system on purpose for maintenance, and the procedure to restart theexternal storage system after the maintenance to restore the HNAS F systemare described below, referring to the case of Figure 3-1 An Example of HNASF System Configuration Including an External Storage System on page 3-4.


Figure 3-1 An Example of HNAS F System Configuration Including anExternal Storage System

Note:When you need to stop the external storage system on purpose, executethe procedure according to the following description. If you execute thewrong procedure, an error occurs in the HNAS F system (e.g., the filesystem may be blocked, the status of resource group may becomeinappropriate).

To stop the external storage system on purpose:

1. Stop the accesses from the client.

2. Stop the cluster using File Services Manager.

3. Delete all the pairs of ShadowImage or TrueCopy, if you have created thepairs specifying the external volumes.

4. Stop the OS 0 and the OS 1 using File Services Manager.

5. Disconnect the external storage system using Universal Volume Manager.For the procedure to disconnect an external storage system, see theHitachi Universal Volume Manager User's Guide (User Guide).

6. Stop the external storage system.For the procedure to stop an external storage system, see the HitachiUniversal Volume Manager User's Guide (User Guide).


To restart the storage system and restore the HNAS F system

1. Turn the external storage system power supply on.For the procedure to turn an external storage system power supply on,see the Hitachi Universal Volume Manager User's Guide (User Guide).

2. Confirm that the status of the external storage system is normal.

3. Execute the Check Paths & Restore Vol. command using UniversalVolume Manager to restore the path to the external storage system.For the procedure to execute the Check Paths & Restore Vol.command, see the Hitachi Universal Volume Manager User's Guide (UserGuide).

4. Confirm that the connection between the storage system and the externalstorage system has become normal.

5. Start the OS 0 and the OS 1 using File Services Manager.

6. If you have deleted the ShadowImage or TrueCopy pairs as you stoppedthe external storage system, re-create the deleted pairs.

7. Start the cluster using File Services Manager.

8. Resume the accesses from the client.

Recovery Procedure in Case of Error in External Storage System

This section describes the recovery procedures in case when an error occursin the external storage system that is in the HNAS F system. The recoveryprocedures are described using examples as follows.

• In Case of Error in a Disk in the External Storage System on page 3-5

• In Case of Error in a Path to the External Storage System on page 3-8

• In Case of Error in All the Paths to the External Storage System on page3-9

In Case of Error in a Disk in the External Storage System

Figure 3-2 In Case of Error in a Disk in the External Storage System on page3-6 shows an example of the case when an error occurs in a disk in theexternal storage system. The situation of Figure 3-2 In Case of Error in a Diskin the External Storage System on page 3-6 is as follows:

• The data itself that has been stored in the error disk of external storagesystem 1 cannot be restored. If you need to restore the data to theoriginal disk after the recovery of the HNAS F system, the data shouldhave been backed up in the disk other than the error disk.

• The clients cannot access the volumes of the error disk. The clients canonly access the volumes other than the error disk.

• The storage system recognizes that the status of the file system and thevolume of the error disk is blocked.


Figure 3-2 In Case of Error in a Disk in the External Storage System

The recovery procedure of the error such as Figure 3-2 In Case of Error in aDisk in the External Storage System on page 3-6 is as follows:

1. Delete all the pairs of ShadowImage or TrueCopy, if you have created thepairs.

2. Perform the following operations on the node 0.

¢ Release the differential-data storage device using the file snapshotfunctionality.

¢ Delete the NFS® share, CIFS share, and file system using FileServices Manager.

3. Change the execution node of the resource group 0 to the node 1 usingFile Services Manager (failover).If the status of the resource group 0 has been Offline, this operation isnot required.

4. Stop the node 0 using File Services Manager.

5. Restart the OS 0 using File Services Manager.


6. Start the node 0 using File Services Manager.

7. Perform one of the following operations using File Services Manager.

¢ When the status of the resource group 0 has been Online, change theexecution node of the resource group 0 to the node 0 (failback).

¢ When the status of the resource group 0 has been Offline, start theresource group 0.

8. Perform the following operations on the node 1.

¢ Release the differential-data storage device using the file snapshotfunctionality.

¢ Delete the NFS® share, CIFS share, and file system using FileServices Manager.








14. Change the error disk in the external storage system to restore the statusof the external storage system.

15. Execute the Check Paths & Restore Vol. command using UniversalVolume Manager.For the procedure to execute the Check Paths & Restore Vol.command, see the Hitachi Universal Volume Manager User's Guide (UserGuide).

16. Perform the following operations on both of the node 0 and node 1.

¢ Set up the differential-data storage device using the file snapshotfunctionality.

¢ Create the file system and then create the NFS® share and CIFSshare using File Services Manager.

17. Put back the backed up data to the restored disk of the external storagesystem if you have had the backed up data in the disk other than theerror disk.

18. Re-create the pairs of ShadowImage and TrueCopy.


In Case of Error in a Path to the External Storage System

Figure 3-3 In Case of Error in a Path to the External Storage System on page3-8 shows an example of the case when an error occurs in a path betweenthe storage system and an external storage system. In Figure 3-3 In Case ofError in a Path to the External Storage System on page 3-8, an erroroccurs in the path between the storage system and the external storagesystem 1, however, the path between the storage system and the externalstorage system 2 is normal.

The situation of Figure 3-3 In Case of Error in a Path to the External StorageSystem on page 3-8 is as follows:

• The clients cannot access the volumes of the external storage system 1,but they can access the volumes of the external storage system 2.

• The storage system recognizes that the status of all the file systems andvolumes of the external storage system 1 are blocked.

Figure 3-3 In Case of Error in a Path to the External Storage System

The recovery procedure of the error such as Figure 3-3 In Case of Error in aPath to the External Storage System on page 3-8 is as follows:


1. Restore (e.g., checking the connection status of the cable, changingswitch, and so on) the status of the error path between the storagesystem and the external storage system 1.
















In Case of Error in All the Paths to the External Storage System

The following subsections describe the recovery procedures with twoexamples of the case when the error occurs in all the paths connected to theexternal storage system.

• In cases where both nodes use different external storage systems

• In cases where both nodes use the same external storage system

In Case that Each Node Uses Different External Storage System

In the configuration of Figure 3-4 In Case that Error Occurs on the Path tothe External Storage System that is Used for Node 1 on page 3-11, all user


LUs of the HNAS F system are the volumes of the external storage system.Each node uses the volume of different storage system. All user LUs of thenode 0 are the volumes of the external storage system 1. Therefore, if anerror occurs in the path between the storage system and the external storagesystem 1, this means that the error has occurred in all the paths connectedfrom the node 0 to the external storage system 1. In this situation, no userLU is available for the node 0.

The situation of Figure 3-4 In Case that Error Occurs on the Path to theExternal Storage System that is Used for Node 1 on page 3-11 is as follows:

• The HNAS F system attempts to change the execution node of theresource group 0 to the node 1 because no user LU can be used fromnode 0 (failover). However, the node 1 cannot access the volumes of theexternal storage system 1, the processing of the failover function hasfailed, and srmd executable error is displayed as the error informationof the resource group 0.

• The status of the file system is blocked on the node 1 because of thefailure of the processing of the failover of the resource group 0.

• The clients cannot access the volumes of the external storage system 1,but they can access the volumes of the external storage system 2.

• The storage system recognizes that the status of all the file systems andvolumes of the external storage system 1 are blocked.


Figure 3-4 In Case that Error Occurs on the Path to the External StorageSystem that is Used for Node 1

The recovery procedure of the error such as Figure 3-4 In Case that ErrorOccurs on the Path to the External Storage System that is Used for Node 1 onpage 3-11 is as follows:



3. Perform a forced stop operation using File Services Manager for theresource group 0 for which the status is displayed as srmd executableerror.


5. Restart OS 0 using File Services Manager.



7. Release the blocked status of the file system of the node 1.To release the blocked status of the file system of the node 1, performthe operations in the following order:

a. Change the execution node of the resource group 1 to the node 0using File Services Manager (failover).

b. Restart the OS 1 using File Services Manager.

c. Change the execution node of the resource group 1 to the node 1(failback).

8. Start the resource group 0 on the node 0 using File Services Manager.

In Case that Both Nodes Use the Same External Storage System

In the configuration of Figure 3-5 In Case that Both Nodes Use the SameExternal Storage System on page 3-13, all user LUs of the HNAS F systemare the volumes of the external storage system. Only one external storagesystem is connected to the storage system. Also only one path is set betweenthe storage system and the external storage system. In this case, if an errorhas occurred in the only set path, all the volumes in the external storagesystem cannot be used.

Note:In the configuration such as Figure 3-5 In Case that Both Nodes Use theSame External Storage System on page 3-13, setting alternate paths isrecommended to prevent the HNAS F system to be blocked because ofthe path failure. For detailed information about the alternate paths, seethe Hitachi Universal Volume Manager User's Guide (User Guide).

The situation of Figure 3-5 In Case that Both Nodes Use the Same ExternalStorage System on page 3-13 is as follows:

• All user LUs cannot be used neither from the node 0 nor the node 1.

• The clients cannot access the volumes.

• The storage system recognizes that the status of all the file systems andvolumes are blocked.


Figure 3-5 In Case that Both Nodes Use the Same External Storage System

The recovery procedure of the error such as Figure 3-5 In Case that BothNodes Use the Same External Storage System on page 3-13 is as follows:



3. Perform the forced stop operations using File Services Manager for both ofthe resource group 0 and resource group 1.

4. Stop the cluster using File Services Manager.

5. Restart OS 0 and OS 1 using File Services Manager.

6. Start the cluster using File Services Manager.

7. Start both of the resource group 0 and resource group 1 using FileServices Manager.


Volume MigrationVolume Migration is a program product that helps you eliminate systembottlenecks by distributing the load concentrated on a specific disk orprocessor to other disks or processors in a storage system. If the usagestatistics collected by Performance Monitor show that the access load isconcentrated on a specific hard disk drive, the system administrator can useVolume Migration to distribute the load to another drive.

Before you use Volume Migration, carefully read the Hitachi PerformanceManager User's Guide (User Guide) and the Hitachi Volume Migration UserGuide, and make sure that you understand the program functions.

Volume Migration also works with user LUs in the HNAS F.

You can use Volume Migration in conjunction with Universal Volume Manager.When Universal Volume Manager is used to map volumes on an externalstorage system to internal volumes, you can check the usage of the volumeson the external storage system (external volumes), and of the externalvolume groups that contain these volumes.

Note:In the HNAS F, only user LUs can be used with Volume Migration. SystemLUs are excluded.

Volume ShredderVolume Shredder is a software product that erases all of the data in a volumein storage system. Data erased by this software cannot be recovered.

Before using the Volume Shredder functionality, make sure you understand itby carefully reading the Hitachi Virtual Storage Platform Hitachi VolumeShredder User Guide.

Volume Shredder can be used to completely erase volumes used in the HNASF.

Note:When Volume Shredder is used to erase data from a system LU or userLU used in the HNAS F, the erased data can never be recovered. Exerciseadequate caution before erasing the data.

Encryption License KeyEncryption License Key is a program product used to encrypt data on storagesystem volumes. By encrypting data, information leaks can be prevented inthe event that a storage system or hard disk in the storage system isswapped out (and accidentally used for another purpose) or stolen.

Please read the Encryption License Key User's Guide (User Guide) beforeusing any Encryption License Key functionality.


Volumes used by HNAS F can be encrypted by using the Encryption LicenseKey.


4Resource Management Functions

HNAS F can be connected and used with a storage system configured usingthe resource management functionality provided by storage system.

□ Storage Navigator

□ LUN Manager

□ Configuration File Loader

□ Virtual Partition Manager

Resource Management Functions 4-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Storage NavigatorStorage Navigator is a program product for remotely operating storagesystem.

Before you use Storage Navigator, you must first prepare an environmentthat allows use of the Web browser specified for use with storage system.

Using Storage Navigator, you can perform settings and operations in thefollowing program products:

• Configuration File Loader

• Dynamic Provisioning

• Encryption License Key

• LUN Manager

• Performance Monitor

• ShadowImage

• TrueCopy


• Universal Volume Manager

• Virtual LVI

• Virtual Partition Manager

• Volume Migration

• Volume Shredder

LUN ManagerLUN Manager is a program product that helps you build a storageenvironment using storage system.

Before you use LUN Manager, carefully read the Provisioning Guide, andmake sure that you understand the program functions.

LUN Manager can be used with the HNAS F to perform the following tasks:

Installing the HNAS F

• Create a host group for a Fibre-channel port to which a node connects

• Assign an LU to a created host group

Adding an LU used by the HNAS F

• Add a host group to a Fibre-channel port to which a node is connected

• Add an LU to a registered host group

Note:The HNAS F allows you to create an alternate path if the LU path shouldbecome unavailable for some reason. To ensure that the alternate path

4-2 Resource Management FunctionsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

behaves correctly, configure the host group and its LUs in advance instorage system so that they can use the alternate path.

Note:The HNAS F allows you to set a cluster encompassing the two nodesnode0 and node1. To ensure that the cluster works correctly, configurethe host group and its LUs in advance in the storage system so that thecluster can be used.

Configuration File LoaderConfiguration File Loader is a program product that allows you to define theconfiguration information for storage system in a single file, thus allowing forthe batch setting of configuration information.

Before you use the Configuration File Loader functionality, carefully read theHitachi Storage Navigator User's Guide (User Guide) and the Hitachi SystemOperations Using Spreadsheets, and make sure that you understand theprogram functions.

Configuration File Loader outputs the configuration information as aspreadsheet. You can then use spreadsheet software or a text editor to defineor edit the configuration for storage system.

In a storage system to which a node used by the HNAS F connects, you canuse Configuration File Loader to configure the host group used by the nodeand the LUs allocated to that host group as a batch.

The LUN Manager provided with Storage Navigator can also be used toconfigure storage system. LUN Manager is useful when defining or changingindividual settings. However, Configuration File Loader, because it loads theentire configuration as one file, is more useful in situations that require batchprocessing, such as when you create or modify settings on a large scale.

Virtual Partition ManagerVirtual Partition Manager is a program product that enables logicalpartitioning of the resources of storage system. Virtual Partition Managerprovides cache partitioning.

The cache partitioning functionality allows you to create multiple units ofvirtual cache from the storage system cache memory, which can be allocatedamong the hosts in the system. This means that when a specific host has ahigh I/O workload, it does not have a negative impact on the I/Operformance of the other hosts in the system.

Before you use Virtual Partition Manager, carefully read the Hitachi VirtualPartition Manager User's Guide (User Guide), and make sure that youunderstand the program functions.

HNAS F can be used in conjunction with the Virtual Partition Managerprovided by storage system.

Resource Management Functions 4-3Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Cache partitioning

When one storage system is shared among a large number of hosts includingthe host running the HNAS F, a large proportion of the storage system cachememory may be occupied by a specific host as it handles a large amount ofdata. Such a situation may cause a reduction in write speeds, as other hostswill need to wait for their turn to write to the cache.

The cache partitioning functionality of Virtual Partition Manager partitions thestorage system cache memory into multiple units of virtual cache memory.Because each host is allocated a specific amount of cache memory, you canavoid situations in which one host uses more than its fair share.

Each unit of virtual cache memory created by the cache partitioningfunctionality is called a CLPR (Cache Logical Partition).

Note:When using HNAS F in conjunction with the cache partitioningfunctionality of Virtual Partition Manager, make sure that the storagesystem LUs used by the node pair that makes up the HNAS F cluster arein the same CLPR defined on the storage system device.

4-4 Resource Management FunctionsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

5Performance Management Functions

The HNAS F can be used in conjunction with the performance managementfunctionality provided by storage system.

□ Performance Monitor

Performance Management Functions 5-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Performance MonitorPerformance Monitor is a program product that collects information about theutilization of resources such as hard disk drives, logical volumes andprocessors built into the storage system.

Before you use Performance Monitor, carefully read the Hitachi PerformanceMonitor User's Guide (User Guide), and make sure that you understand theprogram functions.

Using Performance Monitor, you can also monitor disk workloads and trafficbetween a host and a storage system. In the Performance Monitor windows,resource utilization, loading, and traffic patterns are displayed in the form ofline graphs. The system administrator can use the information displayed inPerformance Monitor to analyze trends in disk access or identify when I/Oaccess is busiest.

When using the HNAS F, you can use Performance Monitor with storagesystem to view information about utilization of resources such as hard diskdrives, logical volumes, and processors used by the HNAS F.

5-2 Performance Management FunctionsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

ADetails of ShadowImage Operations on

the HNAS F

This appendix describes operating procedures when the ShadowImagefunctionality of storage system is used in conjunction with the HNAS F .

This chapter discusses the following topics:

□ Execution of Command Operation on HNAS F

□ Pairs Recovery from Failures on the HNAS F

Details of ShadowImage Operations on the HNAS F A-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Execution of Command Operation on HNAS FThe following describes the HNAS F commands used in How to useShadowImage on page 2-6.

Note:To continue the operation at the failover destination when managing thesystem with node, execute the command making the remote host specifythe virtual IP address. When you connect the file system, execute thecommand by adding the -r option and specifying a resource group name.

Creating Pairs

Commands for Creating Pairs

When creating a pair, the following commands are used.

Table A-1 Commands for Creating Pairs (HNAS F)

No. Command Description

1 sudo horcvmdefine -d device-file-number[, device-file-number ...]

Reserves device file.

Table A-2 Commands for Creating Pairs (CCI)


1 sudo paircreate {-g group-name | -dvolume-name} -vl

Creates a volume pair.

2 sudo pairvolchk {-g group-name | -dvolume-name}

Checks the pair volume status.

3 sudo pairdisplay {-g group-name | -dvolume-name} -fc

Checks the pair status.

Procedure for Creating Pairs

The pair is usually created in the following procedure. The sudo command,the options of the HNAS F command, and the CCI command are omittedwhen it is described in the following table. Specify the appropriate options forthe actual operation.

Table A-3 Procedure for Creating Pairs

No.P-VOL Site S-VOL Site

Command Processing NFS/

CIFS FS P-VOLStatus

S-VOLStatus FS NFS/

CIFS Processing Command

1 Public Mount SMPL SMPL – –

A-2 Details of ShadowImage Operations on the HNAS FHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


2 Reserves S-VOL

horcvmdefine


4 pairdisplay

Confirm thecopy targetbeforecreating apair


6 paircreate

Creates avolume pair

7 Public Mount COPY COPY – –

8 pairdisplay

Confirm avolume pairstatus


10 pairvolchk

At this pointthe pairstatus isCOPY

11 Execute pairvolchkseveral times

Public Mount COPY COPY – –

12 pairvolchk

When a pairstatuschange toPAIR,crating pairiscompleted

13 Public Mount PAIR PAIR – –

Splitting Pairs

Commands for Splitting Pairs

When splitting a pair, the following commands are used.

Table A-4 Commands for Splitting Pairs (HNAS F)


1 sudo nfsdelete -d shared-directory {-a |-H Host}

Delete an NFS share.



2 sudo nfscreate -d shared-directory -HHost

Create an NFS share.

3 sudo cifsdelete -x CIFS-share-name Delete a CIFS share.

4 sudo cifscreate -x CIFS-share-name -dshared-directory

Create a CIFS share.

5 sudo fsumount file-system-name Un-mount a file system.

6 sudo fsmount {-r | -w} file-system-name

Mount a file system.

7 sudo horcfreeze -f copy-source-file-system-name

Suppresses on the P-VOL and stopsaccess from clients.

8 sudo horcunfreeze -f copy-source-file-system-name

Restarts operations on the P-VOL andaccess from clients.

9 sudo horcimport -f copy-destination-file-system-name -d device-file-number [-rresource-group-name]

When LVM is not used and the filesystem is not tiered, the commandconnects the file system to the node.

10 sudo horcimport -f copy-destination-file-system-name --tier1 device-file-number--tier2 device-file-number [-r resource-group-name]

When LVM is not used and the filesystem is tiered, the command connectsthe file system to the node.

11 sudo horcvmimport -f copy-destination-file-system-name -d device-file-number[, device-file-number...] [-r resource-group-name]

When LVM is used and the file system isnot tiered, the command connects thefile system to the node.

12 sudo horcvmimport -f copy-destination-file-system-name --tier1 device-file-number [, device-file-number...] --tier2device-file-number [, device-file-number...] [-r resource-group-name]

When LVM is used and the file system istiered, the command connects the filesystem to the node.

13 sudo syncumount mount-point-name# Un-mounts the differential-datasnapshot.

14 sudo syncmount file-system-namedifferential-data-snapshot-name mount-point-name

Mounts the differential-data snapshot.

#This information is required when publishing the data in the shared filesystem.

Table A-5 Commands for Splitting Pairs (CCI)


1 sudo pairsplit {-g group-name | -dvolume-name}

Splits a volume pair.




Procedure for Splitting Pairs with P-VOL Mounted

The pair is split with P-VOL mounted in the following procedure. The sudocommand, the options of the HNAS F command, and the CCI command areomitted when it is described in the following table. Specify the appropriateoptions for the actual operation.

Table A-6 Procedure for Splitting Pairs with P-VOL Mounted



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 horcfreeze

Hold theaccess


4 pairsplit

Split a pair


6 pairvolchk




8 pairvolchk

When thepair statuschanges toPSUS, thepair is split

9 Public Mount PSUS SSUS – –

10 horcunfreeze

Cancelsuppressionofoperations


12 Connect thefilesystem#1

horcvmimport(whenLVM is notused,horcimport)

13 Public Mount PSUS SSUS Un-mount

Non-public

14 Mount fsmount




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


15 Public Mount PSUS SSUS Mount Non-public

16 Share#2 nfscreate/cifscreate

17 Public Mount PSUS SSUS Mount Public

18 When usingthe filesnapshotfunctionality

syncmount


20 Start a program that accesses the filesystem at S-VOL site.

#1:Specify it as a file system name different from that in P-VOL site.

#2:Specify a public directory name and CIFS name that differ from those forthe P-VOL.

Procedure for Splitting Pairs with P-VOL Un-mounted

The pair is split with P-VOL un-mounted in the following procedure. The sudocommand, the options of the HNAS F command, and the CCI command areomitted when it is described in the following table. Specify the appropriateoptions for the actual operation.

When LVM is not used:

Table A-7 Procedure for Splitting Pairs when LVM is Not Used (with P-VOL Un-mounted)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the filesystem at P-VOL site.


3 nfsdelete/cifsdelete

Delete NFS/CIFS shares




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


4 Non-public

Mount PAIR PAIR – –

5 fsumount Un-mount

6 Non-public

Un-mount

PAIR PAIR – –

7 pairsplit

Split a pair

8 Non-public

Un-mount

COPY COPY – –

9 pairvolchk



Public Un-mount

COPY COPY – –

11 pairvolchk


12 Non-public

Un-mount

PSUS SSUS – –

13 fsmount Mount

14 Non-public

Mount PSUS SSUS – –

15 nfscreate/cifscreate

Share


17 Restart a program that accesses the filesystem at P-VOL site.



horcimport


Non-public

21 Mount fsmount





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/







When LVM is used and a differential-data snapshot is not madeavailable in a shared file system:

Table A-8 Procedure for Splitting Pairs when LVM is Used (with a differential-datasnapshot not available in the shared file system, and with P-VOL Un-mounted)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/






4 Non-public


5 fsumount Un-mount

6 Non-public

Un-mount

PAIR PAIR – –

7 horcfreeze

Suppressoperations

8 Non-public

Un-mount

PAIR PAIR – –

9 pairsplit

Split a pair




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


10 Non-public

Un-mount

COPY COPY – –

11 pairvolchk



Non-public

Un-mount

PSUS SSUS – –

13 pairvolchk


14 Non-public

Un-mount

PSUS SSUS – –

15 horcunfreeze


16 Non-public

Un-mount

PSUS SSUS – –

17 fsmount Mount

18 Non-public



Share


21 syncmount

When usingthe filesnapshotfunctionality


23 Restart a program that accesses the filesystem at P-VOL site.



horcvmimport


Non-public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


27 Mount fsmount





syncmount





When LVM is used and a differential-data snapshot is madeavailable in a shared file system

Table A-9 Procedure for Splitting Pairs When LVM is Used (made available in a sharedfile system) (with an unmounted P-VOL)

NoP-VOL Site S-VOL Site


CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stops programs that access file systems atthe P-VOL site.


3 syncumount



DeletesNFS/CIFSshares.




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


6 Non-public


7 fsumount Un-mount

8 Non-public

Un-mount

PAIR PAIR – –

9 horcfreeze

Suppressesoperations.

10 Non-public

Un-mount

PAIR PAIR – –

11 pairsplit

Splits apair.

12 Non-public

Un-mount

COPY COPY – –

13 pairvolchk

At thispoint, thepair statusis COPY.

14 Executes pairvolchkseveral times

Non-public

Un-mount

PSUS SSUS – –

15 pairvolchk

When thepair statuschanges toPSUS, thepair is split.

16 Non-public

Un-mount

PSUS SSUS – –

17 horcunfreeze

Cancels thesuppressionofoperations.

18 Non-public

Un-mount

PSUS SSUS – –

19 fsmount Mount

20 Non-public



Share


23 syncmount




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



25 Restarts programs that access file systemsat the P-VOL site.


27 Connects tothe filesystem.#1

horcvmimport


Non-public

29 Mount fsmount




33 syncmount


35 Starts programs that access file systems atthe S-VOL site.

#1:Specify a file system name different from that of the P-VOL site for this.

#2:Specify a public directory name different from that of the P-VOL site forthis.

Re-synchronizing Pairs

Commands for Re-synchronizing Pairs

When resynchronizing a pair, the following commands are used.

Table A-10 Commands for Re-synchronizing Pairs (HNAS F)








4 sudo horcexport -f file-system-name Separate a file system.

5 sudo syncumount Un-mounts the differential-datasnapshot.

Table A-11 Commands for Re-synchronizing Pairs (CCI)


1 sudo pairresync {-g group-name | -dvolume-name}

Resynchronizes the split pairs.



Procedure for Re-synchronizing Pairs When P-VOL Status is PSUSand S-VOL Status is SSUS

Use the following procedure to resynchronize a pair when the P-VOL status isPSUS and the S-VOL status is SSUS. The sudo command, the options of theHNAS F command, and the CCI command are omitted when it is described inthe following table. Specify the appropriate options for the actual operation.


Table A-12 Procedure for Re-synchronizing Pairs when LVM is Not Used (with P-VOL inPSUS status and S-VOL in SSUS status)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the filesystem at S-VOL site.


3 Deleteshares

nfsdelete/cifsdelete


5 Un-mount fsumount


Non-public

7 Separatefile system

horcexport




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



9 pairresync

Resynchronizes thepairs


11 pairvolchk

At this pointa pairstatus isCOPY



13 pairvolchk

When a pairstatuschange toPAIR, re-synchronizing pair iscompleted



Table A-13 Procedure for Re-synchronizing Pairs when LVM is Used (with a differential-data snapshot not available in the shared file system, and with P-VOL in PSUS status

and S-VOL in SSUS status)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




3 Deleteshares#




syncumount




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



7 Un-mount fsumount


Non-public


horcexport


11 pairresync



13 pairvolchk




15 pairvolchk

When thepair statuschanges toPAIR,resynchronization iscompleted


#If a differential-data snapshot is made available to a shared file system,and if NFS shares or CIFS shares are created for the differential-datasnapshot, before executing the syncumount command, you need to usethe nfsdelete command or the cifsdelete command to release the NFSshares or the CIFS shares for the differential-data snapshot.



Table A-14 Procedure for Re-synchronizing Pairs When LVM Is Used (made available inthe shared file system) (with the P-VOL in the PSUS status and the S-VOL in the SSUS

status)

Note

P-VOL Site S-VOL Site


CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stops programs that access file systems atthe S-VOL site.


3 syncumount


5 Deletesshares.#



7 Un-mount fsumount


Non-public

9 Separatesthe filesystem.

horcexport


11 pairresync

Resynchronizes thepairs.


13 pairvolchk


14 Executes pairvolchkseveral times.


15 pairvolchk

When thepair statuschanges toPAIR, theresynchronization ofthe pair iscomplete.


Note



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




Restoring Pairs

Commands for Restoring Pairs

When restoring a pair, the following commands are used.

Table A-15 Commands for Restoring Pairs (HNAS F)























12 sudo syncumount mount-point-name Un-mounts the differential-datasnapshot.

13 sudo syncmount file-system-namedifferential-data-snapshot-name file-system-name


Table A-16 Commands for Restoring Pairs (CCI)


1 sudo pairsplit {-g group-name | -dvolume-name}


2 sudo pairresync {-g group-name | -dvolume-name} -restore

Resynchronizes the P-VOL and S-VOL.



Procedure for Restoring Pairs

The pair is restored with volume status PSUS/SSUS in the followingprocedure. The sudo command, the options of the HNAS F command, and theCCI command are omitted when it is described in the following table. Specifythe appropriate options for the actual operation.

When an LVM is not used

Table A-17 Procedure for Restoring Pairs (LVM is not used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/






4 Non-public

Mount PSUS SSUS Mount Public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


5 fsumount Un-mount

6 Non-public

Un-mount

PSUS SSUS Mount Public

7 horcexport

Separatefile system

8 – – PSUS SSUS Mount Public



11 Delete NFS/CIFS shares


12 – – PSUS SSUS Mount Non-public

13 Un-mount fsumount

14 – – PSUS SSUS Un-mount

Non-public


horcexport

16 – – PSUS SSUS – –

17 pairresync -restore

Restores apairs

18 – – RCPY RCPY – –

19 pairvolchk

At this pointa pairstatus isRCPY


– – RCPY RCPY – –

21 pairvolchk

When a pairstatuschange toPAIR,restoring iscompleted

22 – – PAIR PAIR – –

23 pairsplit

Splits avolume pair

24 – – PSUS COPY – –




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


25 pairvolchk



– – PSUS COPY – –

27 pairvolchk

When thepair statuschanges toPSUS,splitting iscompleted


29 horcimport

Connect thefile system

30 Non-public

Un-mount

PSUS SSUS – –

31 fsmount Mount

32 Non-public



Share


35 Start a program that accesses the filesystem at P-VOL site.

36 Connect thefile system

horcimport


Non-public

38 Mount fsmount


40 Share nfscreate/cifscreate


42 Restart a program that accesses the filesystem at S-VOL site.


When an LVM is used and a differential-data snapshot is not madeavailable in a shared file system

Table A-18 Procedure for Restoring Pairs (LVM is used and a differential-data snapshotis not made available in a shared file system)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




3 syncumount




Delete NFS/CIFSshares#

6 Non-public


7 fsumount Un-mount

8 Non-public

Un-mount


9 horcexport

Separatefile system





syncumount









CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



Non-public


horcexport



Restores apairs


23 pairvolchk

At this pointa pairstatus isRCPY



25 pairvolchk

When a pairstatuschange toPAIR,restoring iscompleted


27 pairsplit

Splits avolume pair


29 pairvolchk




31 pairvolchk

When thepair statuschanges toPSUS,splitting iscompleted


33 horcvmimport


34 Non-public

Un-mount

PSUS SSUS – –




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


35 fsmount Mount

36 Non-public



Share


39 syncmount



41 Start a program that accesses the filesystem at P-VOL site.


horcvmimport


Non-public

44 Mount fsmount





syncmount


50 Restart a program that accesses the filesystem at S-VOL site.

#In the snapshot function target file system, un-mount the differential-datasnapshot using the syncumount command before executing thenfsdelete or cifsdelete command.



Table A-19 Procedure for Restoring Pairs (LVM is used, the differential-data snapshot ismade available)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stops programs that access file systems atthe P-VOL site.


3 syncumount



DeletesNFS/CIFSshares.#

6 Non-public


7 fsumount Un-mount

8 Non-public

Un-mount


9 horcexport

Separatefile system




13 syncumount


15 DeletesNFS/CIFSshares.#





Non-public


horcexport




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




Restores apair.


23 pairvolchk

At thispoint, thepair statusis RCPY.



25 pairvolchk

When a pairstatuschanges toPAIR, therestorationis complete.


27 pairsplit

Splits avolume pair


29 pairvolchk




31 pairvolchk

When thepair statuschanges toPSUS, thesplitoperation iscomplete.


33 horcvmimport

Connects tothe filesystem.

34 Non-public

Un-mount

PSUS SSUS – –

35 fsmount Mount

36 Non-public





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



Share


39 syncmount


41 Starts programs that access file systems atthe P-VOL site.


horcvmimport


Non-public

44 Mount fsmount




48 syncmount


50 Restarts program that access file systemsat the S-VOL site.


Deleting Pairs

Commands for Deleting Pairs

When deleting a pair, the following commands are used.


Table A-20 Commands for Deleting Pairs (HNAS F)












7 sudo fsdelete file-system-name Delete a file system.
















16 sudo syncstop file-system-name Releases the differential-data storagedevice.

17 sudo horcvmdelete -d device-file-number[, device-file-number...]

Releases device files.


Table A-21 Commands for Deleting Pairs (CCI)


1 sudo pairsplit {-g group-name | -dvolume-name} -S

Deletes a volume pair.

2 sudo pairsplit { -g group-name | -dvolume-name}

Split a volume pair.



Procedure for Deleting Pairs when the Pair Status is PAIR (S-VOLContinuously Used-1)

The pair split when S-VOL is continuously used after the pair deletion withstatus PAIR in the following procedure. The sudo command, the options ofthe HNAS F command, and the CCI command are omitted when it isdescribed in the following table. Specify the appropriate options for the actualoperation.

Table A-22 Procedure for Deleting Pairs when the Pair Status is PAIR (S-VOLContinuously Used-1)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 horcfreeze

Suppresstheoperations


4 pairsplit

Splits avolume pair


6 pairvolchk

The pairstatus atthis time isCOPY.



8 pairvolchk

The pairstatuschanges toPSUS andthe splittingoperationcompletes.




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



10 horcunfreeze

Restarts theoperations


12 pairsplit -S

Deletes avolume pair


14 pairvolchk




16 pairvolchk

When a pairstatuschange toSMPL,deleting iscompleted




19 Public Mount SMPL SMPL Un-mount

Non-public

20 Mount fsmount

21 Public Mount SMPL SMPL Mount Non-public


23 Public Mount SMPL SMPL Mount Public


syncmount





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


26 Start a program that accesses the filesystem at the S-VOL site.



Procedure for Deleting Pairs when the Pair Status is PSUS (S-VOLContinuously Used-2)

The pair split when S-VOL is continuously used after the pair deletion withstatus PSUS/SSUS in the following procedure. The sudo command, theoptions of the HNAS F command, and the CCI command are omitted when itis described in the following table. Specify the appropriate options for theactual operation.

Table A-23 Procedure for Deleting Pairs When the Pair Status Is PSUS (S-VOLContinuously Used-2)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Public Mount PSUS PSUS Mount Public

2 pairsplit -S



4 pairvolchk



Procedure for Deleting Pairs when the Pair Status is PAIR (S-VOLNot Used-1)

The pair split when S-VOL is not used after the pair deletion with status PAIRin the following procedure. The sudo command, the options of the HNAS Fcommand, and the CCI command are omitted when it is described in thefollowing table. Specify the appropriate options for the actual operation.


Table A-24 Procedure for Deleting Pairs When the Pair Status Is PAIR (S-VOL NotUsed-1)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 pairsplit -S



4 pairvolchk




6 pairvolchk



8 Releasesdevice files

horcvmdelete


Procedure for Deleting Pairs (S-VOL Not Used-2) when the PairStatus is PSUS

The pair split when S-VOL is not used after the pair deletion with statusPSUS/SSUS in the following procedure. The sudo command, the options ofthe HNAS F command, and the CCI command are omitted when it isdescribed in the following table. Specify the appropriate options for the actualoperation.



Table A-25 Procedure for Deleting Pairs When the Pair Status Is PSUS while LVM is NotUsed (S-VOL Not Used-2)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/






4 Public Mount PSUS PSUS Mount Non-public

5 Un-mount fsumount

6 Public Mount PSUS PSUS Un-mount

Non-public

7 Delete filesystem

fsdelete

8 Public Mount PSUS PSUS – –

9 pairsplit -S



11 pairvolchk



When LVM is used and a differential-data snapshot is not madeavailable in a shared file system:

Table A-26 Procedure for Deleting Pairs When the Pair Status Is PSUS while LVM isUsed (S-VOL Not Used-2)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




syncumount



syncstop


7 Delete NFS/CIFSshares#



9 Un-mount fsumount


Non-public

11 Delete filesystem

fsdelete


13 pairsplit -S



15 pairvolchk

When thepair statuschanges toSMPL,deleting iscompleted


#In the target file system of the file snapshot functionality, un-mount thedifferential-data snapshot using the syncumount command beforeexecuting the nfsdelete or cifsdelete command, and release thedevice which stores the differential data using the syncstop command.


When LVM is used and a differential-data snapshot is madeavailable in a shared file system:

Table A-27 Procedure for Deleting Pairs When the Pair Status Is PSUS while LVM isUsed (the S-VOL Is Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




3 syncumount


5 DeletesNFS/CIFSshares.#



7 syncstop


9 Un-mount fsumount


Non-public

11 Deletes thefile system.

fsdelete


13 pairsplit -S

Deletes avolumepair.


15 pairvolchk

When thepair statuschanges toSMPL, thedeletionoperation iscomplete.


#If a differential-data snapshot is made available to a shared file system,and if NFS shares or CIFS shares are created for the differential-data


snapshot, before executing the syncumount command, you need to usethe nfsdelete command or the cifsdelete command to release the NFSshares or the CIFS shares for the differential-data snapshot.

Pairs Recovery from Failures on the HNAS F

Commands for Pairs RecoveryWhen recover a pair from failures, the following commands are used.

Table A-28 Commands for Recovery (HNAS F)






3 sudo cifsdelete -x CIFS-share-name [-rresource-group-name]

Delete a CIFS share.






7 sudo fscreate file-system-name Create a file system.


9 sudo lumapctl -t m --on Change the mode of allocation of a userLUN to the maintenance mode.

10 sudo lumapctl -t m --off Change the mode of allocation of a userLUN to the normal management mode.





13 sudo horcvmimport -f copy-destination-file-system-name -d device-file-number[, device-file-number ...] [-r resource-group-name]


14 sudo horcvmimport -f copy-destination-file-system-name --tier1 device-file-number [, device-file-number...] --tier2




device-file-number [, device-file-number...] [-r resource-group-name]

15 syncumount mount-point-name Un-mounts the differential-datasnapshot.



17 syncstop file-system-name Releases the differential-data storagedevice.








Table A-29 Commands for Recovery (CCI)


1 sudo paircreate {-g group-name | -dvolume-name} -vl

Creates a volume pair.







5 sudo horcmstart.sh Starts CCI.

Note:To continue the operation at the fail-over destination when managing thesystem with node, execute the command making the remote host specifythe virtual IP address. When you connect the file system, execute thecommand by adding the -r option and specifying a resource group name.

Recovery Procedure from FailuresWhen a pair was suspended (PSUE) due to a failure, recover in the followingprocedure. The sudo command, the options of the HNAS F command, and theCCI command are omitted when it is described in the following table. Specifythe appropriate options for the actual operation.


When LVM is not Used

Table A-30 Procedure for Recovering Pairs when Using a Node (When LVM Is Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 lumapctl-t m --on

Maintenance mode

2 Public Mount PSUE PSUE Mount Public

3 pairsplit -S





6 Non-public

Mount SMPL SMPL Mount Public

7 fsumount Un-mount

8 Non-public

Un-mount

SMPL SMPL Mount Public

9 fsdelete Delete filesystem

10 – – SMPL SMPL Mount Public

11 Failover from the failed node (node #1) tothe normal node (node #2) #


13 Stop the node #1


15 Restart the OS #1


17 Start the node #1


19 Failback the node #1


21 Failover from the node #2 to the node #1


23 Stop the node #2





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/








31 Device file recovery from the failure (Thisoperation should be done by a servicepersonnel)


33 horcmstart.sh

Starts CCI


35 Starts CCI horcmstart.sh






40 – – SMPL SMPL Mount Non-public


42 – – SMPL SMPL Un-mount

Non-public

43 horcvmdefine

Reserves S-VOL


Non-public

45 Create avolume pair

paircreate

46 – – COPY COPY Un-mount

Non-public

47 Confirmthat a

pairdisplay




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


volume pairis created


Non-public

49 Check theprogress ofcopying

pairvolchk


Non-public

Execute pairvolchkseveral times

51 When a pairstatuschange toPAIR,creatingpair iscompleted

pairvolchk

52 – – PAIR PAIR Un-mount

Non-public

53 Splitting avolume pair

pairsplit


Non-public

55 When a pairstatuschange toPSUS,splittingpair iscompleted

pairvolchk

56 – – SSUS PSUS Un-mount

Non-public

57 horcimport


58 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public

59 Deleting avolume pair

pairsplit -S

60 Non-public

Un-mount

COPY COPY Un-mount

Non-public


pairvolchk

62 Non-public

Un-mount

COPY COPY Un-mount

Non-public





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


63 When a pairstatuschange toSMPL,confirm thatsplittingpair iscompleted

pairvolchk

64 Non-public

Un-mount

SMPL SMPL Un-mount

Non-public

65 fsmount Mount

66 Non-public

Mount SMPL SMPL Un-mount

Non-public


Share


Non-public

69 Separatesfile system

fsdelete


71 paircreate



73 pairvolchk

Check theprogress ofcopying



75 pairvolchk

When a pairstatuschange toPAIR,creatingpair iscompleted


77 lumapctl-t m --off

Normalmanagement mode


#


For example, if the file system is blocked because of the failure occurredon the drive, you need to reboot the OS to release the blocked status thatis recognized by the OS site.

Table A-31 Procedure for Recovering Pairs when Using a Virtual Server (When LVM IsNot Used)

No.


Command Processing NFS/CIFS FS

P-VOL

Status

S-VOL

Status

FS NFS/CIFS Processing Command

1 Public Mount

PSUE PSUE Mount

Public

2 pairsplit-S


3 Public Mount

SMPL SMPL Mount

Public



5 Non-public

Mount

SMPL SMPL Mount

Public

6 fsumount Un-mount

7 Non-public

Un-mount

SMPL SMPL Mount

Public


9 – – SMPL SMPL Mount

Public

10 Fails over any virtual servers running on thesame node as the virtual server that failed.


Public

12 Acquires the dump


Public

14 Fails back the virtual servers


Public

16 Recovers the device file from the error (tobe performed by service personnel).


Public


No.



P-VOL

Status

S-VOL

Status


18 horcmstart.sh

Starts CCI


Public



Public

22 Stops the programs that access the filesystem at the S-VOL site.


Public




Non-public



Non-public

28 horcvmdefine

Reserves S-VOL


Non-public


paircreate


Non-public

32 Confirm thata volumepair iscreated

pairdisplay


Non-public


pairvolchk


No.



P-VOL

Status

S-VOL

Status



Non-public


36 When a pairstatuschange toPAIR,creating pairis completed

pairvolchk


Non-public


pairsplit


Non-public

40 When a pairstatuschange toPSUS,splitting pairis completed

pairvolchk


Non-public

42 horcimport Connect thefile system

43 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


pairsplit-S

45 Non-public

Un-mount

COPY COPY Un-mount

Non-public


pairvolchk

47 Non-public

Un-mount

COPY COPY Un-mount

Non-public



No.



P-VOL

Status

S-VOL

Status


48 When a pairstatuschange toSMPL,confirm thatsplitting pairis completed

pairvolchk

49 Non-public

Un-mount

SMPL SMPL Un-mount

Non-public

50 fsmount Mount

51 Non-public

Mount

SMPL SMPL Un-mount

Non-public


Share

53 Public Mount

SMPL SMPL Un-mount

Non-public


horcexport

55 Public Mount

SMPL SMPL - -

56 paircreate Creates avolume pair

57 Public Mount

COPY COPY - -

58 pairvolchk Check theprogress ofcopying


Public Mount

COPY COPY - -

60 pairvolchk When a pairstatuschange toPAIR,creating pairis completed

61 Public Mount

PAIR PAIR - -


When LVM is Used

Table A-32 Recovery Procedure from Failures When Using node (LVM is Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 lumapctl-t m --on

Maintenance mode

2 Public Mount PSUE PSUE Mount Public

3 pairsplit -S



5 syncumount



7 syncstop When usingthe filesnapshotfunctionality



Delete NFS/CIFSshares#1

10 Non-public


11 fsumount Un-mount

12 Non-public

Un-mount




15 Failover from the failed node (node #1) tothe normal node (node #2) #2


17 Stop the node #1







CIFS FS P-VOLStatus

S-VOLStatus FS NFS/








27 Stop the node #2








35 Device file recovery from the failure (Thisoperation should be done by a servicepersonnel)


37 horcmstart.sh

Starts CCI







syncumount







CIFS FS P-VOLStatus

S-VOLStatus FS NFS/





Non-public

49 horcvmdefine

Reserves S-VOL


Non-public


paircreate


Non-public

53 Confirmthat avolume pairis created

pairdisplay


Non-public


pairvolchk


Non-public


57 When a pairstatuschange toPAIR,creatingpair iscompleted

pairvolchk


Non-public

59 Suppresstheoperations

horcfreeze


Non-public


pairsplit


Non-public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



pairvolchk


Non-public

65 Restartsoperations

horcunfreeze


Non-public

67 horcvmimport


68 Non-public

Un-mount

PAIR PAIR Un-mount

Non-public


pairsplit -S

70 Non-public

Un-mount

COPY COPY Un-mount

Non-public


pairvolchk

72 Non-public

Un-mount

COPY COPY – – Execute pairvolchkseveral times

73 When a pairstatuschange toSMPL,confirm thatsplittingpair iscompleted

pairvolchk

74 Non-public

Un-mount

SMPL SMPL – –

75 fsmount Mount

76 Non-public

Mount SMPL SMPL Un-mount

Non-public


Share




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



Non-public

79 syncmount



Non-public


horcexport


83 paircreate



85 pairvolchk

Check theprogress ofcopying



87 pairvolchk

When a pairstatuschange toPAIR,creatingpair iscompleted





#1:In the target file system of the file snapshot functionality, un-mount thedifferential-data snapshot using the syncumount command beforeexecuting the nfsdelete or cifsdelete command, and release thedevice which stores the differential data using the syncstop command.

#2:For example, if the file system is blocked because of the failure occurredon the drive, you need to reboot the OS to release the blocked status thatis recognized by the OS site.


Table A-33 Procedure for Recovering Pairs when Using a Virtual Server (When LVM IsUsed)

No.



P-VOL

Status

S-VOL

Status


1 Public Mount

PSUE PSUE Mount

Public

2 pairsplit-S


3 Public Mount

SMPL SMPL Mount

Public

4 syncumount When usingthe filesnapshotfunctionality

5 Public Mount

PSUE PSUE Mount

Public


7 Public Mount

PSUE PSUE Mount

Public


Delete NFS/CIFSshares#

9 Non-public

Mount

SMPL SMPL Mount

Public


11 Non-public

Un-mount

SMPL SMPL Mount

Public



Public

14 Fails over the virtual servers running on thesame node as the virtual server that failed.


Public

16 Acquire the dump.


No.



P-VOL

Status

S-VOL

Status



Public

18 Fails back the virtual servers.


Public

20 Recovers the device file from the error (tobe performed by service personnel).


Public

22 horcmstart.sh

Starts CCI


Public



Public

26 Stop programs that access the file systemat the S-VOL site.


Public


syncumount


Public




Non-public



Non-public

34 horcvmdefine

Reserves S-VOL


No.



P-VOL

Status

S-VOL

Status



Non-public


paircreate


Non-public

38 Confirm thata volumepair iscreated

pairdisplay


Non-public


pairvolchk


Non-public

Execute pairvolchkseveral times.

42 When a pairstatuschange toPAIR,creating pairis completed

pairvolchk


Non-public

44 Suppresstheoperations

horcfreeze


Non-public


pairsplit


Non-public

48 When a pairstatus

pairvolchk


No.



P-VOL

Status

S-VOL

Status


change toPSUS,splitting pairis completed


Non-public

50 horcvmimport


51 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


pairsplit-S

53 Non-public

Un-mount

COPY COPY Un-mount

Non-public


pairvolchk

55 Non-public

Un-mount

COPY COPY Non-public

Un-mount

Execute pairvolchkseveral times.

56 When a pairstatuschange toSMPL,confirm thatsplitting pairis completed

pairvolchk

57 Non-public

Un-mount

SMPL SMPL Non-public

Un-mount

58 Restartsoperations

horcunfreeze

59 Non-public

Un-mount

SMPL SMPL Non-public

Un-mount

60 fsmount Mount

61 Non-public

Mount

SMPL SMPL Un-mount

Non-public


No.



P-VOL

Status

S-VOL

Status



Share

63 Public Mount

SMPL SMPL Un-mount

Non-public

64 syncmount When usingthe filesnapshotfunctionality

65 Public Mount

SMPL SMPL Un-mount

Non-public


horcexport

67 Public Mount

SMPL SMPL - -

68 paircreate Creates avolume pair

69 Public Mount

COPY COPY - -

70 pairvolchk Check theprogress ofcopying

71 Execute pairvolchkseveral times.

Public Mount

COPY COPY – –

72 pairvolchk When a pairstatuschange toPAIR,creating pairis completed

73 Public Mount

PAIR PAIR – –

#If the target file system uses the file snapshot functionality, beforeexecuting the nfsdelete or cifsdelete command, use the syncumountcommand to unmount the differential-data snapshots, and then use thesyncstop command to release the differential-date storage device.


BDetails of TrueCopy and Universal

Replicator Operations on the HNAS F

This appendix describes operating procedures when the TrueCopy orUniversal Replicator functionality of storage system is used in conjunctionwith the HNAS F.

This chapter discusses the following topics:

□ Execution of Command Operation on HNAS F

Details of TrueCopy and Universal Replicator Operations on the HNAS F B-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Execution of Command Operation on HNAS FThis appendix explains how to use the commands described in How to useTrueCopy on page 2-57 and How to use Universal Replicator on the HNAS Fon page 2-93 on the HNAS F.

Note:To continue the operation at the fail-over destination when managing thesystem with node, execute the command making the remote host specifythe virtual IP address. When you connect the file system, execute thecommand by adding the -r option and specifying a resource group name.

Creating Pairs

Commands for Creating Pairs

When creating a pair, the following commands are used.

Table B-1 Commands for Creating Pairs (HNAS F)


1 sudo horcvmdefine -d device-file-number[, device-file-number...]


Table B-2 Commands for Creating Pairs (CCI)


1 sudo paircreate {-g group-name | -dvolume-name} -f never -vl

Creates a TrueCopy pair.

2 sudo paircreate {-g group-name | -dvolume-name} -f async –vl -jp P-VOL-journal-volume-ID -js S-VOL-journal-volume-ID

Creates a Universal Replicator pair.

3 sudo pairvolchk {-g group-name | -dvolume-name} -ss

Checks the status of a volume pair.


Checks the pair status. Use thiscommand to check if the pair is created.

Procedure for Creating Pairs

The pair is usually created in the following procedure. The sudo command,the options of the HNAS F command, and the CCI command are omittedwhen it is described in the following table. Specify the appropriate options forthe actual operation.

B-2 Details of TrueCopy and Universal Replicator Operations on the HNAS FHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Table B-3 Procedure for Creating Pairs

No.Main Site Remote Site


CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Reserves S-VOL

horcvmdefine


4 pairdisplay

Confirm thecopy targetbeforecreating apair


6 paircreate



8 pairdisplay

Confirm avolume pairstatus


10 pairvolchk




12 pairvolchk

When a pairstatuschange toPAIR,crating pairiscompleted


Splitting Pairs

Commands for Splitting Pairs

When splitting a pair, the following commands are used.


Table B-4 Commands for Splitting Pairs (HNAS F)

























13 sudo syncumount mount-point-name# Un-mounts the differential-datasnapshot.



#This information is required when publishing the data in the shared filesystem.


Table B-5 Commands for Splitting Pairs (CCI)


1 sudo pairsplit {-g group-name | -dvolume-name} -rw




Procedure for Splitting Pairs with P-VOL Mounted

The pair is split with P-VOL mounted in the following procedure. The sudocommand, the options of the HNAS F command, and the CCI command areomitted when it is described in the following table. Specify the appropriateoptions for the actual operation.

Table B-6 Procedure for Splitting Pairs with P-VOL Mounted



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 horcfreeze

Hold theaccess


4 pairsplit -rw

Split a pair


6 pairvolchk




8 pairvolchk

When a pairstatuschange toPSUS,splittingpair iscompleted


10 horcunfreeze






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


12 Connect thefile system#



Non-public

14 Mount fsmount





syncmount


20 Start a program that accesses the filesystem at the S-VOL.

#:Specify the same file system name as the main site for the file systemname connected with node.

Procedure for Splitting Pairs with P-VOL Un-mounted

The pair is split with P-VOL un-mounted in the following procedure. The sudocommand, the options of the HNAS F command, and the CCI command areomitted when it is described in the following table. Specify the appropriateoptions for the actual operation.



Table B-7 Procedure for Splitting Pairs when LVM is Not Used (with P-VOL Un-mounted)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the filesystem at the P-VOL.




4 Non-public


5 fsumount Un-mount

6 Non-public

Un-mount

PAIR PAIR – –

7 pairsplit -rw

Split a pair

8 Non-public

Un-mount

COPY COPY – –

9 pairvolchk



Non-public

Un-mount

COPY COPY – –

11 pairvolchk

When a pairstatuschange toPSUS,splittingpair iscompleted

12 Non-public

Un-mount

PSUS SSUS – –

13 fsmount Mount

14 Non-public



Share





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


17 Restart a program that accesses the filesystem at the P-VOL.



horcimport


Non-public

21 Mount fsmount




25 Start a program that accesses the filesystem at the S-VOL.

#Specify the same file system name as the main site for the file systemname connected with node.

When LVM is used:

Table B-8 Procedure for Splitting Pairs when LVM is Used (with P-VOL Un-mounted)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the filesystem on the P-VOL.


3 syncumount








CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


6 Non-public


7 fsumount Un-mount

8 Non-public

Un-mount

PAIR PAIR – –

9 horcfreeze

Hold theaccess

10 Non-public

Un-mount

PAIR PAIR – –

11 pairsplit -rw

Split a pair

12 Non-public

Un-mount

COPY COPY – –

13 pairvolchk



Non-public

Un-mount

PSUS SSUS – –

15 pairvolchk


16 Non-public

Un-mount

PSUS SSUS – –

17 horcunfreeze


18 Non-public

Un-mount

PSUS SSUS – –

19 fsmount Mount

20 Non-public



Create NFS/CIFS shares


23 Restart the programs that access the filesystem at the P-VOL.




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




horcvmimport


Non-public

27 Mount fsmount


29 Create NFS/CIFS shares

nfscreate/cifscreate



syncmount


33 Start the programs that access the filesystem on the S-VOL.

#Specify the same file system name as that connected to the main site.

Re-synchronizing Pairs

Commands for Re-synchronizing Pairs

When resynchronizing a pair, the following commands are used.

Table B-9 Commands for Re-synchronizing Pairs (HNAS F)










Table B-10 Commands for Re-synchronizing Pairs (CCI)






Procedure for Re-synchronizing Pairs

The pair is resynchronized with volume status PSUS/SSUS in the followingprocedure. The sudo command, the options of the HNAS F command, and theCCI command are omitted when it is described in the following table. Specifythe appropriate options for the actual operation.


Table B-11 Procedure for Re-synchronizing Pairs when LVM is Not Used



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the filesystem at the S-VOL.





5 Un-mount fsumount


Non-public


horcexport


9 pairresync



11 pairvolchk





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




13 pairvolchk



When LVM is used:

Table B-12 Procedure for Re-synchronizing Pairs when LVM is Used



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop the programs that access the S-VOLfile system at the S-VOL.



syncumount





7 Un-mount fsumount


Non-public


horcexport


11 pairresync





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



13 pairvolchk




15 pairvolchk



#In the target file system of the file snapshot functionality, un-mount thedifferential-data snapshot using the syncumount command beforeexecuting the nfsdelete or cifsdelete command.

Deleting Pairs

Commands for Deleting Pairs

When deleting a pair, the following commands are used.

Table B-13 Commands for Deleting Pairs (HNAS F)































16 sudo syncstop file-system-name Releases the differential-data storagedevice.

17 sudo horcvmdelete -d device-file-number[, device-file-number...]

Releases device files.

Table B-14 Commands for Deleting Pairs (CCI)








Procedure for Deleting Pairs when the Pair Status is PAIR (S-VOLContinuously Used-1)

The pair delete when S-VOL is continuously used after the pair deletion withstatus PAIR in the following procedure. The sudo command, the options ofthe HNAS F command, and the CCI command are omitted when it is


described in the following table. Specify the appropriate options for the actualoperation.

Table B-15 Procedure for Deleting Pairs when the Pair Status is PAIR (S-VOLContinuously Used-1)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 horcfreeze

Suppressestheoperations


4 pairsplit

Splitting avolume pair


6 pairvolchk




8 pairvolchk

When a pairstatuschange toPSUS,splitting iscompleted.


10 horcunfreeze

Restarts theoperations


12 pairsplit -S



14 pairvolchk




16 pairvolchk

When a pairstatuschange toSMPL,




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


deleting iscompleted





Non-public

20 Mount fsmount





syncmount


26 Start a program that accesses the file system at theS-VOL.

#Specify the same file system name as the main site for the file systemname connected with node.

Procedure for Deleting Pairs when the Pair Status is PSUS (S-VOLContinuously Used-2)

The pair delete when S-VOL is continuously used after the pair deletion withstatus PSUS/SSUS in the following procedure. The sudo command, theoptions of the HNAS F command, and the CCI command are omitted when itis described in the following table. Specify the appropriate options for theactual operation.


Table B-16 Procedure for Deleting Pairs When the Pair Status Is PSUS (S-VOLContinuously Used-2)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 pairsplit -S



4 pairvolchk



Procedure for Deleting Pairs when the Pair Status is PAIR (S-VOLNot Used-1)

The pair delete when S-VOL is not used after the pair deletion with statusPAIR in the following procedure. The sudo command, the options of the HNASF command, and the CCI command are omitted when it is described in thefollowing table. Specify the appropriate options for the actual operation.

Table B-17 Procedure for Deleting Pairs When the Pair Status Is PAIR (S-VOL NotUsed-1)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 pairsplit -S



4 pairvolchk

At thispoint, thepair statusis COPY



6 pairvolchk

When a pairstatuschange toSMPL,




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


deleting iscompleted


8 Releasesdevice files

horcvmdelete


Procedure for Deleting Pairs (S-VOL Not Used-2) when the PairStatus is PSUS

The pair delete when S-VOL is not used after the pair deletion with statusPSUS/SSUS in the following procedure. The sudo command, the options ofthe HNAS F command, and the CCI command are omitted when it isdescribed in the following table. Specify the appropriate options for the actualoperation.


Table B-18 Procedure for Deleting Pairs When the Pair Status Is PSUS while LVM is NotUsed (S-VOL Not Used-2)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the S-VOL filesystem.





5 Un-mount fsumount


Non-public

7 Delete filesystem

fsdelete


9 pairsplit -S





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



11 pairvolchk



When LVM is used:

Table B-19 Procedure for Deleting Pairs When the Pair Status Is PSUS while LVM isUsed (S-VOL Not Used-2)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


1 Stop a program that accesses the S-VOL filesystem.



syncumount



syncstop





9 Un-mount fsumount


Non-public

11 Delete filesystem

fsdelete




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



13 pairsplit -S



15 pairvolchk

When thepair statuschanges toSMPL,deletion iscompleted


#In the target file system of the file snapshot functionality, un-mount thedifferential-data snapshot using the syncumount command beforeexecuting the nfsdelete or cifsdelete command, and release thedevice which stores the differential data using the syncstop command.


COperation when Failures Occurred on

the HNAS F

This appendix explains eight potential failures when the TrueCopy orUniversal Replicator functionality of storage system is used in conjunctionwith the HNAS F .

- Operation when the Main Site or Storage System Went Down on pageC-4

- Operation when the Cluster of the Main Site Went Down on page C-16

- Operation when Multiple Failures Occurred in All the Storages on the MainSite on page C-28

- Operation when Multiple Failures Occurred in a Part of Storages on the MainSite on page C-43

- Operation when Network Failures Occurred Between the Main Site and theRemote Site Causing Journal Overflow on page C-58

- Operation when Network Failures Occurred Between the Main Site and theRemote Site Without Causing Journal Overflow on page C-60

- Operation when Network Failures Occurred in the Main Site on page C-61

- Operation when Multiple Failures Occurred in Some or All Journals on theMain Site on page C-72

The commands and the important options are described, but specify theappropriate options referring to the following manual.

- Command References

- Hitachi Command Control Interface (CCI) User and Reference Guide

Note:

To continue operation at the failover destination, specify the virtual IPaddress when you execute commands from the remote host. When youconnect the file system, execute the commands by adding the -r option andspecifying a resource group name.

Operation when Failures Occurred on the HNAS F C-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

The operations described in this document are typical operations, and theseare not likely to be able to solve the problem. Please evaluate the operationenough before executing it.

□ Isolation when Failures Occur

□ Operation when the Main Site or Storage System Went Down

□ Operation when the Cluster of the Main Site Went Down

□ Operation when Multiple Failures Occurred in All the Storages on the MainSite

□ Operation when Multiple Failures Occurred in a Part of Storages on theMain Site

□ Operation when Network Failures Occurred Between the Main Site and theRemote Site Causing Journal Overflow

□ Operation when Network Failures Occurred Between the Main Site and theRemote Site Without Causing Journal Overflow

□ Operation when Network Failures Occurred in the Main Site

□ Operation when Multiple Failures Occurred in Some or All Journals on theMain Site

C-2 Operation when Failures Occurred on the HNAS FHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Isolation when Failures OccurPerform the isolation based on the following flow chart when failuresoccurred. Moreover, failures of the case, which is not applied, to thisflowchart may have occurred, so that perform the isolation contacting theservice personnel. It is described by separating the cases when it can beconnected to the HNAS F volume from the client and when it cannot beconnected to the client.

When the failure prevents access to the HNAS F volume


When the failure does not prevent access to the HNAS F volume

Operation when the Main Site or Storage System WentDown

Assumed Scenarios

The main site is struck, the business operation on the main site cannot beperformed, and the business operation is continued on the remote site. Themain site cannot be restored, and a new storage system was introduced.


Commands to be Used

The following commands are used when the main site or storage system wentdown.

Table C-1 Commands for Recovery when the Main Site or Storage SystemWent Down (HNAS F)

























12 sudo horcsetenv HORCMINST-instance-number

Sets up or modifies the CCI environmentvariable.

13 sudo horcunsetenv HORCC_MRCF Deletes the CCI environment variable.










Table C-2 Commands for Recovery when the Main Site or Storage SystemWent Down (CCI)


1 sudo horctakeover {-g group-name | -dvolume-name} [-t time out]

Takeover the pair. When using UniversalReplicator, you can only executetakeover for individual groups. The -toption is mandatory with async.

2 sudo paircreate {-g group-name | -dvolume-name} -f async –vl -jp P-VOL-journal-ID -js journal-ID


3 sudo paircreate { -g group-name | -dvolume-name -f never -vl




4 sudo pairsplit {-g group-name | -dvolume-name} -R

Brings S-VOL into SMPL forcibly.



6 sudo pairresync {-g group-name | -dvolume-name} -swaps



Checks the status of the paired volumeon the applicable site.

8 sudo pairdisplay {-g group-name | -dvolume-name} -fce

Checks the pair status. The commandoutput contains lines corresponding toCTG (CT group ID), JNL (journal groupID), and AP (path number). When usingUniversal Replicator, verify that the JNLin the command output is the onespecified in the paircreate command.


Recovery Procedure from Failures

When the main site is struck, recover it in the following procedure. The sudocommand, the options of the HNAS F command, and the CCI command areomitted when it is described in the following table. Specify the appropriateoptions for the actual operation.

It is a prerequisite that S-VOL is reserved.


Table C-3 Recovery Procedure when the Main Site or Storage System Went Down (whenLVM is Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Main site is down.

3 – – – SSUS – –

4 The takeover execution is decided by thecustomer judgment.

5 Executetakeover

horctakeover

6 – – – SSUS – –




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



horcimport

8 – – – SSUS Un-mount

Non-public

9 Mount fsmount

10 – – – SSUS Mount Non-public

Mountcompleted


12 – – – SSUS Mount Public

13 Start the business operation at the remotesite. #2

14 Main site is recovered.


16 cluster in the main site was started.


18 horcsetenv

Set CCIenvironment


20 horcunsetenv

Set CCIenvironment


22 Deletes avolume pair

pairsplit -R

23 – – – SMPL Mount Public

24 When a pairstatuschange toSMPL,deleting iscompleted

pairvolchk


26 horcmstart.sh

Start CCI





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


28 horcvmdefine

Reservesthe old P-VOL


30 Confirm theconfiguration definitionfile beforecrating pair

pairdisplay


32 Creates avolume pair

paircreate

33 – – COPY COPY Mount Public

34 Confirm avolume pairstatus

pairdisplay


36 At this pointthe pairstatus isCOPY

pairvolchk

37 – – COPY COPY Mount Public Execute pairvolchkseveral times

38 When a pairstatuschange toPAIR,crating pairiscompleted

pairvolchk

39 – – PAIR PAIR Mount Public

40 Stop the business operation at the remotesite.



42 – – PAIR PAIR Mount Non-public



Non-public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


45 Splits avolume pair

pairsplit -rw

46 COPY COPY Un-mount

Non-public


pairvolchk


Non-public



pairvolchk


Non-public

51 horcimport

Connect thefilesystem#1

52 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

54 Non-public

Un-mount

SSUS PSUS – –

55 pairresync -swaps

Reverseresynchronize

56 Non-public

Un-mount

COPY COPY – –

57 pairvolchk



Non-public

Un-mount

COPY COPY – –

59 pairvolchk

When a pairstatuschange toPAIR, re-synchronizi




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


ng pair iscompleted

60 Non-public

Un-mount

PAIR PAIR – –

61 fsmount Mount

62 Mountcompleted

Non-public



Share

64 Sharecreationcompleted

Public Mount PAIR PAIR – –

65 Resume business operations at the mainsite. #2

#1:Specify the same file system name as the main site for the file systemname connected with node.

#2:When the takeover is performed, the IP address cannot be taken overfrom the primary site to the secondary site. When starting a job at thesecondary site, un-mount the client from the primary site, change the IPaddress of the site on which the client is to be mounted to that of thesecondary site, and then mount the client again. When the job is resumedat the primary site because the primary site has been restored, un-mountthe client from the secondary site, return the IP address of the site onwhich the client is to be mounted to that of the primary site, and thenmount the client again.

When LVM is used:

Table C-4 Recovery Procedure when the Main Site or Storage System Went Down (whenLVM is Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Main site is down.

3 – – – SSUS – –




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


4 The user determines whether to executetakeover.

5 Executetakeover

horctakeover

6 – – – SSUS – –


horcvmimport

8 – – – SSUS Un-mount

Non-public

9 Mount fsmount

10 – – – SSUS Mount Non-public

Mountcompleted





syncmount


15 Start business operations at the remotesite. #2

16 Main site is recovered.


18 Cluster in the main site was started.


20 horcsetenv

Set CCIenvironment


22 horcunsetenv

Set CCIenvironment



pairsplit -R




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



26 When thepair statuschanges toSMPL,deletion iscompleted

pairvolchk


28 horcmstart.sh

Start CCI


30 horcvmdefine



32 Confirm thecopy targetbeforecreatingpair

pairdisplay



paircreate


36 Checkwhethervolume pairhas beencreated

pairdisplay



pairvolchk


40 When thepair statuschanges toPAIR,creation iscompleted

pairvolchk





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


42 Stop business operations at the remotesite.



syncumount


46 Delete NFS/CIFSshares.





Non-public

50 Suppressfile systemoperations

horcfreeze

51 PAIR PAIR Un-mount

Non-public

52 Splits avolume pair

pairsplit -rw


Non-public


pairvolchk


Non-public


56 When thepair statuschanges toPSUS,splitting iscompleted

pairvolchk


Non-public

58 Cancelsuppressionof file

horcunfreeze




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


systemoperations


Non-public

60 horcvmimport


61 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

63 Non-public

Un-mount

SSUS PSUS – –



65 Non-public

Un-mount

COPY COPY – –

66 pairvolchk



Non-public

Un-mount

COPY COPY – –

68 pairvolchk


69 Non-public

Un-mount

PAIR PAIR – –

70 fsmount Mount

71 Mountcompleted

Non-public









CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


74 syncmount




#1:Specify the same file system name as that connected to the main site.

#2:When a takeover is performed, the IP address cannot be taken over fromthe main site to the remote site. When starting a job at the remote site,unmount the client from the primary site, change the IP address of thesite on which the client is to be mounted to that of the remote site, andthen mount the client again. When the main site is restored and businessoperations are resumed at the main site, unmount the client from theremote site, set the IP address of the site on which the client is to bemounted back to that of the main site, and then mount the client again.

Operation when the Cluster of the Main Site Went Down

Assumed Scenarios

Both two nodes of OS on the main side went down. Because OS does notstart, the business operation on the main site cannot be performed, and thebusiness operation is continued on the remote site.


Commands to be Used

The following commands are used when the cluster of the main site wentdown.

Table C-5 Commands for Recovery when the Cluster of Main Site WentDown (HNAS F)
































Table C-6 Commands for Recovery when the Cluster of Main Site WentDown (CCI)






3 sudo paircreate { -g group-name | -dvolume-name -f never} -vl



Deletes a pair.













When the cluster of the main site is struck, recover it in the followingprocedure. The sudo command, the options of the HNAS F command, and theCCI command are omitted when it is described in the following table. Specifythe appropriate options for the actual operation.

It is a prerequisite that S-VOL is reserved.


Table C-7 Recovery Procedure when the Cluster of the Main Site Went Down (when LVMis Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 cluster is down.



5 Executetakeover

horctakeover

6 PAIR PAIR – –

7 Begindeletingpair

pairsplit -S

8 – – SMPL SMPL – –

9 Confirm avolume pairstatus to beSMPL

pairvolchk




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




horcimport


Non-public

13 Mount fsmount


Mountcompleted




18 cluster is recovered.




21 Non-public



23 Non-public

Un-mount


24 horcexport

Separatefile system


26 horcmstart.sh

Start CCI


28 Confirm theconfiguration definitionfile beforecreatingpair

pairdisplay





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


30 Create apair

paircreate



pairdisplay



pairvolchk


36 When a pairstatuschange toPAIR, re-synchronizing pair iscompleted

pairvolchk








Non-public

43 Beginsplittingpair

pairsplit -rw


Non-public


pairvolchk


Non-public





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



pairvolchk


Non-public

49 horcimport


50 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

52 Non-public

Un-mount

SSUS PSUS – –



54 Non-public

Un-mount

COPY COPY – –

55 pairvolchk

Confirmpaircreation


Non-public

Un-mount

COPY COPY – –

57 pairvolchk

When a pairstatuschange toPAIR, paircreation iscompleted

58 Non-public

Un-mount

PAIR PAIR – –

59 fsmount Mount

60 Mountcompleted

Non-public



Share




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




63 Start the business operation at the mainsite. #2



When LVM is used:

Table C-8 Recovery Procedure when the Cluster of the Main Site Went Down (when LVMis Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 cluster is down.



5 Executetakeover

horctakeover

6 PAIR PAIR – –

7 Begindeletingpair

pairsplit -S





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


9 Confirmthat thevolume pairstatus isSMPL

pairvolchk



horcvmimport


Non-public

13 Mount fsmount


Mountcompleted





syncmount



20 cluster on main site is recovered.


22 syncumount





25 Non-public






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


27 Non-public

Un-mount


28 horcexport

Separatefile system


30 horcmstart.sh

Start CCI



pairdisplay


34 Create apair

paircreate



pairdisplay



pairvolchk


40 When thepair statuschanges toPAIR, paircreation iscompleted

pairvolchk




44 When usingthe file

syncumount




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


snapshotfunctionality







Non-public

50 Suppressoperations

horcfreeze


Non-public


pairsplit -rw


Non-public


pairvolchk


Non-public



pairvolchk


Non-public

58 Cancelsuppressionofoperations

horcunfreeze


Non-public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


60 horcvmimport


61 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

63 Non-public

Un-mount

SSUS PSUS – –



65 Non-public

Un-mount

COPY COPY – –

66 pairvolchk



Non-public

Un-mount

COPY COPY – –

68 pairvolchk

When thepair statuschanges toPAIR, paircreation iscompleted

69 Non-public

Un-mount

PAIR PAIR – –

70 fsmount Mount

71 Mountcompleted

Non-public






74 syncmount






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


76 Start business operations at the main site.#2



Operation when Multiple Failures Occurred in All theStorages on the Main Site

Assumed Scenarios

When multiple failures occurred in all the storages used in the user LU on themain site, the business operation cannot be performed on the main site, andthe business operation is continued on the remote site. Note thatShadowImage operation is not performed.

The failed storage on the main site is recovered by the replacement (servicepersonnel).


Commands to be Used

The following commands are used when multiple failures occurred in all thestorages used in the user LU on the main site.

Table C-9 Commands for Recovery when Multiple Failures Occurred in Allthe Storages Used in the User LU on the Main Site (HNAS F)



































19 sudo syncstop file-system-name Releases the device which stores thedifferential data.

20 sudo lumapctl -t m --on Change the mode of allocation of a userLU to the maintenance mode.

21 sudo lumapctl -t m --off Change the mode of allocation of a userLU to the normal management mode.

Table C-10 Commands for Recovery when Multiple Failures Occurred in Allthe Storages Used in the User LU on the Main Site (CCI)











Deletes a pair.











When multiple failures occurred in all the storages used in the user LU on themain site, recover it in the following procedure. The sudo command, theoptions of the HNAS F command, and the CCI command are omitted when itis described in the following table. Specify the appropriate options for theactual operation. It is a prerequisite that S-VOL is reserved.


Table C-11 Recovery Procedure when Multiple Failures Occurred in All the StoragesUsed in the User LU on the Main Site (when LVM is Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Multiple failures occurred in all user LU.

3 – – PSUE SSUS – –





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


5 Executetakeover

horctakeover



pairsplit -S



pairvolchk



horcimport


Non-public

13 Mount fsmount






19 lumapctl-t m --on

Maintenance mode



Deletes theshare

22 Non-public






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


24 Non-public

Un-mount


25 fsdelete Deletes afile system




29 Stop the node #1










39 Stop the node #2








47 Stop the cluster in the main site. #3


49 Shut down the OS in the main site.


51 Replace and format the storage with errorin the main site.


53 Start the OS in the main site.




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



55 Start the cluster in the main site.


57 horcmstart.sh

Start CCI


59 horcvmdefine



61 Confirm theconfiguration definitionfile beforecrating pair

pairdisplay



paircreate



pairdisplay



pairvolchk

68 – – COPY COPY Mount Public When a pair statuschanges to PAIR, paircreation is completed

69 When a pairstatuschanges toPAIR, paircreation iscompleted

pairvolchk




nfsdelete/




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


cifsdelete




Non-public


pairsplit -rw


Non-public


pairvolchk


Non-public



pairvolchk


Non-public

82 horcimport


83 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

85 Non-public

Un-mount

SSUS PSUS – –



87 pairvolchk

Non-public

Un-mount

COPY COPY – –




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


88 Confirmpaircreation


Non-public

Un-mount

COPY COPY – –

90 pairvolchk

When a pairstatuschange toPAIR,resynchronizationcompleted

91 Non-public

Un-mount

PAIR PAIR – –

92 fsmount Mount

93 Mountcompleted

Non-public



Share










#3:When the file system is blocked due to a drive failure etc., the OS needsto be restarted so that the blockade status recognized by the OS will bereleased.

When LVM is used:

Table C-12 Recovery Procedure when Multiple Failures Occurred in All the StoragesUsed in the User LU on the Main Site (When LVM is Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Failures occurred in all user LU.



5 Executetakeover

horctakeover


7 Deletes thevolume pair

pairsplit -S


9

When thepair statuschanges toSMPL,deletion iscompleted

pairvolchk



horcvmimport


Non-public

13 Mount fsmount


Mountcompleted







CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



syncmount



20 Non-public



Maintenance mode


23 syncumount






Deletes theshare#3

28 Non-public



30 Non-public

Un-mount


31 fsdelete Deletes filesystem




35 Stop the node #1





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/










45 Stop the node #2



48

– – SMPL SMPL Mount Public





53 Stop the cluster in the main site. #4


55 Shut down the OS in the main site.


57 Replace and format the faulty storage atthe main site.


59 Start the OS at the main site.


61 Start the cluster at the main site.


63 horcmstart.sh

Start CCI


65 horcvmdefine





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




pairdisplay


69 Create thevolume pair

paircreate



pairdisplay


73 Confirmthat thevolume pairhas beencreated

pairvolchk



pairvolchk





syncumount








CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




Non-public


horcfreeze


Non-public


pairsplit -rw


Non-public


pairvolchk


Non-public



pairvolchk


Non-public


horcunfreeze


Non-public

95 horcvmimport


96 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

98 Non-public

Un-mount

SSUS PSUS – –




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/




100 Non-public

Un-mount

COPY COPY – –

101 pairvolchk



Non-public

Un-mount

COPY COPY – –

103 pairvolchk


104 Non-public

Un-mount

PAIR PAIR – –

105 fsmount Mount

106 Mountcompleted

Non-public




108 Sharescreated


109 syncmount



111 lumapctl–t m --off








#4:When the file system is blocked due to a drive failure etc., the OS needsto be restarted to release the blocked status recognized by the OS.

Operation when Multiple Failures Occurred in a Part ofStorages on the Main Site

Assumed Scenarios

When multiple failures occurred in a part of storages on the main site, thebusiness operation cannot be performed on the main site, and the businessoperation is continued on the remote site. Note that ShadowImage operationis not performed.

The failed storage on the main site is recovered by the replacement (servicepersonnel).


Commands to be Used

The following commands are used when multiple failures occurred in a part ofstorages used in the user LU on the main site.

Table C-13 Commands for Recovery when Multiple Failures Occurred in aPart of Storages Used in the User LU on the Main Site (HNAS F)


































18 sudo syncstop file-system-name Releases the device which stores thedifferential data.

19 sudo lumapctl -t m --on Change the mode of allocation of a userLU to the maintenance mode.

20 sudo lumapctl -t m --off Change the mode of allocation of a userLU to the normal management mode.

Table C-14 Commands for Recovery when Multiple Failures Occurred in aPart of Storages Used in the User LU on the Main Site (CCI)











Deletes a pair.









9 sudo horcmstart.sh Start CCI


When multiple failures occurred in a part of storages used in the user LU onthe main site, recover it in the following procedure. The sudo command, theoptions of the HNAS F command, and the CCI command are omitted when itis described in the following table. Specify the appropriate options for theactual operation. It is a prerequisite that S-VOL is reserved.

When using Universal Replicator, you cannot take over individual volumes.This means that when a fault is localized to a single LU, you will need to takeover the entire group.


Table C-15 Recovery Procedure when Multiple Failures Occurred in a Part of StoragesUsed in the User LU on the Main Site (When LVM is Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Multiple failures occurred in a part of userLU.






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


5 Executetakeover

horctakeover



pairsplit -S



pairvolchk



horcimport


Non-public

13 Mount fsmount


Mountcompleted






Maintenance mode




22 Non-public






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


24 Non-public

Un-mount




27 Replace and format the storage with errorin the main site.


29 Failover from the node1 (blocked node) tothe node2 (non-blocked node) in the mainsite. #3


31 Stop the node1 in the main site.


33 Reboot the OS1 in the main site.


35 Failback in the main site1.




39 Failover from the node2 to the node1 in themain site.




43 Reboot the OS2 in the main site.




47 Start the node2 in the main site.


49 horcmstart.sh

Start CCI





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


51 horcvmdefine



53 Confirm theconfiguration definitionfile beforecreatingpair

pairdisplay



paircreate



pairdisplay



pairvolchk


61 When a pairstatuschange toPAIR, paircreation iscompleted

pairvolchk








Non-public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



pairsplit -rw


Non-public


pairvolchk


Non-public



pairvolchk


Non-public

74 horcimport


75 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

77 Non-public

Un-mount

SSUS PSUS – –



79 Non-public

Un-mount

COPY COPY – –

80 pairvolchk

At this pointa pairstatus isCOPY


Non-public

Un-mount

COPY COPY – –

82 pairvolchk

When a pairstatuschange toPAIR,resynchroni




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


zation iscompleted

83 Non-public

Un-mount

PAIR PAIR – –

84 fsmount Mount

85 Mountcompleted

Non-public



Share









#3:When the file system is blocked due to a drive failure etc., the OS needsto be restarted so that the blockade status recognized by the OS will bereleased.


When LVM is used:

Table C-16 Recovery Procedure when Multiple Failures Occurred in Part of the StorageUsed in the User LU on the Main Site (When LVM is Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Failures occurred in a part of user LU.



5 Executetakeover

horctakeover


7 Beginvolume pairdeletion

pairsplit -S


9 When thepair statuschanges toSMPL,deletion iscompleted

pairvolchk



horcvmimport


Non-public

13 Mount fsmount


Mountcompleted





syncmount





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/





Maintenance mode


23 syncumount






Delete NFS/CIFSshares.#3

28 Non-public



30 Non-public

Un-mount






35 Failover from node1 (blocked node) tonode2 (non-blocked node) at the main site.#4


37 Stop node1 at the main site.


39 Reboot OS1 at the main site.




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/





43 Start node1 at the main site.


45 Failover from node2 to the node1 at themain site.


47 Stop node2 at the main site.


49 Reboot OS2 at the main site.




53 Start node2 at the main site.


55 horcmstart.sh

Start CCI


57 horcvmdefine




pairdisplay



paircreate



pairdisplay




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



65 Confirmthat thepair hasbeencreated

pairvolchk



pairvolchk





syncumount







Non-public


horcfreeze


Non-public


pairsplit -rw


Non-public

81 At this pointthe pair

pairvolchk




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


status isCOPY


Non-public



pairvolchk


Non-public


horcunfreeze


Non-public

87 horcvmimport


88 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

90 Non-public

Un-mount

SSUS PSUS – –



92 Non-public

Un-mount

COPY COPY – –

93 pairvolchk



Non-public

Un-mount

COPY COPY – –

95 pairvolchk





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


96 Non-public

Un-mount

PAIR PAIR – –

97 fsmount Mount

98 Mountcompleted

Non-public




100 Sharescreated


101 syncmount










#4:


When the file system is blocked due to a drive failure etc., the OS needsto be restarted to release the blocked status recognized by the OS.

Operation when Network Failures Occurred Between theMain Site and the Remote Site Causing Journal Overflow

Assumed Scenarios

When using Universal Replicator, communication was lost between the mainsite and the remote site because a cable was disconnected. This led to thejournal overflowing.

Commands to be Used

Table C-17 Commands for Recovery when Network Failures OccurredBetween the Main Site and the Remote Site Causing Journal Overflow

(CCI)



Resynchronizes split pairs.






Use the following procedure to recover from a situation in which a networkfailure has occurred between the main site and the remote site, causing thejournal to overflow. The options of the sudo command, the HNAS Fcommand, and the CCI commands are omitted in the following table. Specifythe appropriate options when using the commands.

Table C-18 Recovery Procedure when Network Failures Occurred Between the Main Siteand the Remote Site Causing Journal Overflow



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Journal overflows.

3 Public Mount SSUS PSUS – –

4 Reconnect the cable between the main siteand the remote site.

5 Public Mount SSUS PSUS – –

6 pairresync

Resynchronize the pair


8 pairvolchk




10 pairvolchk




Operation when Network Failures Occurred Between theMain Site and the Remote Site Without Causing JournalOverflow

Assumed Scenarios

When using Universal Replicator, communication was lost between the mainsite and the remote site because a cable was disconnected, but the journaldid not overflow.


Use the following procedure to recover from a situation in which a networkfailure occurred between the main site and the remote site, but the journaldid not overflow.

Table C-19 Recovery Procedure when Network Failures Occurred Between the Main Siteand the Remote Site Without Causing Journal Overflow



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


2 Cable is disconnected.


4 Reconnect the cable between the main siteand the remote site.

Operation when Network Failures Occurred in the Main Site

Assumed Scenarios

The access from the client cannot be performed due to the switch failures inthe main site.


Commands to be Used

Table C-20 Commands for Recovery when a Network Failure Occurredwithin the Main Site (HNAS F)



















10 Un-mount a file system. Mount a file system.










Table C-21 Commands for Recovery when a Network Failure Occurredwithin the Main Site (CCI)









Deletes a pair.









9 sudo horcmstart.sh Start CCI


The switch failure is usually removed by replacing the switch. However, whenyou want to start the operation on the remote site immediately, remove it inthe following procedure. The sudo command, the options of the HNAS Fcommand, and the CCI command are omitted when it is described in thefollowing table. Specify the appropriate options for the actual operation.


Table C-22 Recovery Procedure when Switch Failures Occurred (when LVM is Not Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


2 Switch obstacle occurs.



5 Executetakeover

horctakeover


7 Begin pairdeletion

pairsplit -S


9 Confirm avolume pairstatus to beSMPL

pairvolchk



horcimport


Non-public

13 Mount fsmount


Mountcompleted




18 Replace the switch.



20 Non-public



22 Non-public

Un-mount





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


23 horcexport

Separatefile system



pairdisplay


27 Create pair paircreate



pairdisplay


31 Confirm thevolume pairstatus

pairvolchk



pairvolchk








Non-public


pairsplit -rw




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



Non-public


pairvolchk


Non-public



pairvolchk


Non-public

46 horcimport


47 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

49 Non-public

Un-mount

SSUS PSUS – –


Resynchronize

51 Non-public

Un-mount

COPY COPY – –

52 pairvolchk



Non-public

Un-mount

COPY COPY – –

54 pairvolchk





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


55 Non-public

Un-mount

PAIR PAIR – –

56 fsmount Mount

57 Mountcompleted

Non-public



Share






When LVM is used:

Table C-23 Recovery Procedure when Switch Failures Occurred (when LVM is Used)



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Switch fault occurs.






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


5 Executetakeover

horctakeover


7 Begin pairdeletion

pairsplit -S


9 Confirmthat thevolume pairstatus isSMPL

pairvolchk



horcvmimport


Non-public

13 Mount fsmount


Mountcompleted





syncmount



20 Replace the switch.

21 syncumount



23 nfsdelete/

Delete NFS/CIFSshares.#3




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


cifsdelete

24 Non-public



26 Non-public

Un-mount


27 horcexport

Separatefile system



pairdisplay


31 Create pair paircreate


33 Checkwhether thevolume pairhas beencreated

pairdisplay


35 Checkwhether thevolume pairhas beencreated

pairvolchk



pairvolchk







CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



syncumount







Non-public


horcfreeze


Non-public


pairsplit -rw


Non-public


pairvolchk


Non-public



pairvolchk


Non-public


horcunfreeze


Non-public




CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


57 horcvmimport


58 Non-public

Un-mount

SSUS PSUS Un-mount

Non-public


horcexport

60 Non-public

Un-mount

SSUS PSUS – –


Resynchronize

62 Non-public

Un-mount

COPY COPY – –

63 pairvolchk



Non-public

Un-mount

COPY COPY – –

65 pairvolchk


66 Non-public

Un-mount

PAIR PAIR – –

67 fsmount Mount

68 Mountcompleted

Non-public




70 Sharescreated


71 syncmount






CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


73 Resume business operations at the mainsite.#2




Operation when Multiple Failures Occurred in Some or AllJournals on the Main Site

Assumed Scenarios

Multiple failures occurred on all or some of the journals on the main site.


Commands to be Used

Table C-24 Commands for Recovery when Multiple Failures Occurred in Allor Some Journals on the Main Site












Use the following procedure to recover from multiple failures of all or some ofthe journals on the main site. The options of the sudo command, the HNAS Fcommands, and the CCI commands are omitted in the following table. Specifythe appropriate options when using the commands.

Table C-25 Recovery Procedure when Multiple Failures Occurred in Some or All Journalson the Main Site



CIFS FS P-VOLStatus

S-VOLStatus FS NFS/



2 Journal failure occurs.

3 Public Mount PSUE SSUS – –

4 pairsplit -S

Begindeletingpair


6 At the main site, clear the definition for thejournal(s) where the fault occurred.




10 Define the journal(s).


12 pairdisplay

Confirm thecopy targetbeforecreatingpair


14 paircreate

Create apair


16 pairdisplay

Checkwhethervolume pairhas beencreated





CIFS FS P-VOLStatus

S-VOLStatus FS NFS/


18 pairvolchk



20 pairvolchk

When thepair statuschanges toPAIR, paircreation iscompleted



DSystem Option Settings

This appendix describes the system option for storage systems.

□ mode495

System Option Settings D-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

mode495If a replication function (ShadowImage) is used by a script which does notuse the horcexport command (supported by Hitachi NAS Base Suite06-03-00-00), set the system option mode to 495 for the storage system.

D-2 System Option SettingsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

EAcronyms

This section lists the acronyms used in the HNAS F manuals.

□ Acronyms used in the HNAS F manuals.

Acronyms E-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Acronyms used in the HNAS F manuals.Following acronyms used in the HNAS F manuals.

ABE Access Based Enumeration

ACE access control entry

ACL access control list

AJP Apache JServ Protocol

API application programming interface

ARP Address Resolution Protocol

ASCII American Standard Code for Information Interchange

ASN Abstract Syntax Notation

BDC Backup Domain Controller

BMC baseboard management controller

CA certificate authority

CHA channel adapter

CHAP Challenge-Handshake Authentication Protocol

CIFS Common Internet File System

CIM Common Information Model

CLI command line interface

CPU central processing unit

CSR certificate signing request

CSV comma-separated values

CTL controller

CU control unit

CV custom volume

DACL discretionary access control list

DAR Direct Access Recovery

DB database

DBMS database management system

DC domain controller

DEP data execution prevention

DES Data Encryption Standard

DFS distributed file system

DIMM dual in-line memory module

DLL dynamic-link library

DN distinguished name

E-2 AcronymsHitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

DNS Domain Name System

DOM Document Object Model

DOS Disk Operating System

DRAM dynamic random access memory

DSA digital signal algorithm

DTD Document Type Definition

ECC error-correcting code

EUC Extended UNIX Code

FC Fibre Channel

FIB forwarding information base

FIFO First In, First Out

FQDN fully qualified domain name

FTP File Transfer Protocol

FV Fixed Volume

FXP File Exchange Protocol

GbE Gigabit Ethernet

GID group identifier

GMT Greenwich Mean Time

GPL GNU General Public License

GUI graphical user interface

HBA host bus adapter

H-LUN host logical unit number

HPFS High Performance File System

HSSO HiCommand single sign-on

HTML HyperText Markup Language

HTTP Hypertext Transfer Protocol

HTTPS Hypertext Transfer Protocol Secure

I/O input/output

ICAP Internet Content Adaptation Protocol

ICMP Internet Control Message Protocol

ID identifier

IP Internet Protocol

IP-SW IP switch

JDK Java Development Kit

JIS Japanese Industrial Standards


JSP JavaServer Pages

KDC Key Distribution Center

LACP Link Aggregation Control Protocol

LAN local area network

LBA logical block addressing

LCD Local Configuration Datastore

LDAP Lightweight Directory Access Protocol

LDEV logical device

LDIF LDAP Data Interchange Format

LDKC logical disk controller

LED light-emitting diode

LF Line Feed

LTS long term support

LU logical unit

LUN logical unit number

LUSE logical unit size expansion

LVI Logical Volume Image

LVM Logical Volume Manager

MAC Media Access Control

MD5 Message-Digest algorithm 5

MIB management information base

MMC Microsoft Management Console

MP microprocessor

MSS maximum segment size

MTU maximum transmission unit

NAS Network-Attached Storage

NAT network address translation

NDMP Network Data Management Protocol

NetBIOS Network Basic Input/Output System

NFS Network File System

NIC network interface card

NIS Network Information Service

NTFS New Technology File System

NTP Network Time Protocol

OID object identifier


ORB object request broker

OS operating system

PAP Password Authentication Protocol

PC personal computer

PCI Peripheral Component Interconnect

PDC Primary Domain Controller

PDU protocol data unit

PID process identifier

POSIX Portable Operating System Interface for UNIX

PP program product

RADIUS Remote Authentication Dial In User Service

RAID Redundant Array of Independent Disks

RAM random access memory

RAS Reliability Availability Serviceability

RCS Revision Control System

RD relational database

RFC Request for Comments

RID relative identifier

RPC remote procedure call

RSA Rivest, Shamir, and Adleman

SACL system access control list

SAN storage area network

SAS Serial Attached SCSI

SATA serial ATA

SAX Simple API for XML

SCSI Small Computer System Interface

SFTP SSH File Transfer Protocol

SHA secure hash algorithm

SID security identifier

SJIS Shift JIS

SLPR Storage Logical Partition

SMB Server Message Block

SMD5 Salted Message Digest 5

SMTP Simple Mail Transfer Protocol

SNMP Simple Network Management Protocol


SOAP Simple Object Access Protocol

SP service pack

SSD solid-state drive

SSH Secure Shell

SSHA Salted Secure Hash Algorithm

SSL Secure Sockets Layer

SSO single sign-on

SVGA Super Video Graphics Array

TCP Transmission Control Protocol

TFTP Trivial File Transfer Protocol

TOS type of service

TTL time to live

UAC User Account Control

UDP User Datagram Protocol

UID user identifier

UNC Universal Naming Convention

URI Uniform Resource Identifier

URL Uniform Resource Locator

UTC Coordinated Universal Time

UTF UCS Transformation Format

VDEV Virtual Device

VLAN virtual LAN

VLL Virtual LVI/LUN

WADL Web Application Description Language

WAN wide area network

WINS Windows Internet Name Service

WORM Write Once, Read Many

WS workstation

WWN World Wide Name

WWW World Wide Web

XDR External Data Representation

XFS extended file system

XML Extensible Markup Language


Index

C

Command Control Interface 2-2Command device 2-3

D

Dynamic Provisioning 3-2Dynamic Tiering 3-2

E

Encryption License Key 3-14

J

journal volume 2-2

L

LU 2-2

P

P-VOL 2-2performance management function 1-2primary volume 2-2program product 1-2

R

replication functions 1-2, 2-2resource management function 1-2

S

S-VOL 2-2secondary volume 2-2secondary volumes 2-2ShadowImage 2-2

T

TrueCopy 2-2

U

Universal Replicator 2-2Universal Volume Manager 3-3

V

volume management function 1-2volume management functions 3-1Volume Migration 3-14Volume Shredder 3-14

Index-1Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Index-2Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Hitachi NAS Platform F1000 Series Enterprise Array Features Administrator's Guide

Hitachi Data Systems

Corporate Headquarters2845 Lafayette StreetSanta Clara, California 95050-2639U.S.A.www.hds.com

Regional Contact Information

Americas+1 408 970 [email protected]

Europe, Middle East, and Africa+44 (0)1753 [email protected]

Asia Pacific+852 3189 [email protected]

MK-92NAS094-02