Hitachi AMS 2000 User Manual

Download

Page 1

Hitachi AMS 2000 Family TrueCopy Extended

Distance User Guide

AST

IND

INKS

Document organization

Release notes and readme

Getting help Table of Contents

MK-97DF8054-23

Page 2

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 retrieval system for any purpose without the express written permission of Hitachi, Ltd. and Hitachi Data Systems Corporation (hereinafter referred to as “Hitachi”).

Hitachi, Ltd. and Hitachi Data Systems reserve the right to make changes to this document at any time without notice and assume no responsibility for its use. Hitachi, Ltd. and Hitachi Data Systems products and services can only be ordered under the terms and conditions of Hitachi Data Systems' applicable agreements.

All of the features described in this document may not be currently available. Refer to the most recent product announcement or contact your local Hitachi Data Systems sales office for information on feature and product availability.

Notice: Hitachi Data Systems products and services can be ordered only under the terms and conditions of Hitachi Data Systems’ applicable agreements. The use of Hitachi Data Systems products is governed by the terms of your agreements with Hitachi Data Systems.

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

All other trademarks, service marks, and company names are properties of their respective owners.

Export authorization is required for the AMS 2000 Data At Rest Encryption

• Import/Use regulations may restrict export of the AMS2000 SED to certain countries

• China – AMS2000 is eligible for import but the License Key and SED may not be sent to China

• France – Import pending completion of registration formalities

• Hong Kong – Import pending completion of registration formalities

• Israel – Import pending completion of registration formalities

• Russia – Import pending completion of notification formalities

• Distribution Centers – IDC, EDC and ADC cleared for exports

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 3

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Intended audience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .x

Product version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .x

Release notes and readme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .x

Product Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .x

Document revision level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Changes in this release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii

Document organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii

Document conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiv

Convention for storage capacity values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Related documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Getting help. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx

Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx

1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1

How TCE works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2

Typical environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2

Volume pairs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3

Data pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4

Guaranteed write order and the update cycle. . . . . . . . . . . . . . . . . . . . . . 1-4

Extended update cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5

Consistency groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6

Differential Management LUs (DMLU) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7

TCE interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7

2 Plan and design — sizing data pools and bandwidth . . . . . . . . . 2-1

Plan and design workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2

Assessing business needs — RPO and the update cycle . . . . . . . . . . . . . . . . 2-2

Measuring write-workload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3

Collecting write-workload data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3

Calculating data pool size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4

Contents iii

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 4

Data pool key points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-7

Determining bandwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-8

3 Plan and design — remote path . . . . . . . . . . . . . . . . . . . . . . . . . . .3-1

Remote path requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-2

Management LAN requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-3

Remote data path requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-3

WAN optimization controller (WOC) requirements . . . . . . . . . . . . . . . . . . .3-4

Remote path configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-5

Fibre channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-5

Direct connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-6

Single FC switch, network connection . . . . . . . . . . . . . . . . . . . . . . . . . .3-7

Double FC switch, network connection . . . . . . . . . . . . . . . . . . . . . . . . .3-8

Fibre channel extender connection . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-9

Port transfer rate for Fibre channel. . . . . . . . . . . . . . . . . . . . . . . . . . .3-10

iSCSI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-11

Direct connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-11

Single LAN switch, WAN connection . . . . . . . . . . . . . . . . . . . . . . . . . .3-12

Multiple LAN switch, WAN connection. . . . . . . . . . . . . . . . . . . . . . . . .3-13

Single LAN switch, WOC, WAN connection. . . . . . . . . . . . . . . . . . . . . .3-14

Multiple LAN switch, WOC, WAN connection . . . . . . . . . . . . . . . . . . . .3-15

Multiple array, LAN switch, WOC connection with single WAN . . . . . . . .3-16

Multiple array, LAN switch, WOC connection with two WANs. . . . . . . . .3-17

Supported connections between various types of arrays . . . . . . . . . . . . . . .3-18

Notes when Connecting Hitachi AMS2000 Series to other arrays . . . . . . . .3-18

Using the remote path — best practices. . . . . . . . . . . . . . . . . . . . . . . . . . .3-20

4 Plan and design—arrays, volumes and operating systems. . . . .4-1

Planning workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-2

Planning arrays—moving data from earlier AMS models. . . . . . . . . . . . . . . . .4-2

Planning logical units for TCE volumes. . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-3

Volume pair and data pool recommendations . . . . . . . . . . . . . . . . . . . . . .4-3

Operating system recommendations and restrictions. . . . . . . . . . . . . . . . . . .4-4

Host time-out. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-4

P-VOL, S-VOL recognition by same host on VxVM, AIX®, LVM . . . . . . . . . .4-4

HP server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-4

Windows Server 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-7

Windows Server 2003/2008. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-7

Identifying P-VOL and S-VOL LUs on Windows. . . . . . . . . . . . . . . . . . . .4-8

Windows 2000 or Windows Server and TCE Configuration . . . . . . . . . . .4-9

Dynamic Disk in Windows 2000/Windows Server . . . . . . . . . . . . . . . . .4-10

VMWare and TCE Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-10

Concurrent Use of Dynamic Provisioning. . . . . . . . . . . . . . . . . . . . . . .4-11

iv Contents

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 5

User Data Area of Cache Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-16

Formatting the DMLU in the Event of a Drive Failure. . . . . . . . . . . . . . . . .4-22

Maximum supported capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-23

TCE and SnapShot capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-24

TCE, SnapShot, ShadowImage concurrent capacity . . . . . . . . . . . . . . . . .4-26

. . . . . . . . . . . . . . .Maximum Supported Capacity of P-VOL and Data Pool4-28

No SnapShot-TCE cascade configuration . . . . . . . . . . . . . . . . . . . . . . .4-29

SnapShot-TCE cascade configuration. . . . . . . . . . . . . . . . . . . . . . . . . .4-30

Cache limitations on data and data pool volumes . . . . . . . . . . . . . . . . . . .4-31

Cautions for Reconfiguring the Cache Memory. . . . . . . . . . . . . . . . . . . . .4-32

5 Requirements and specifications . . . . . . . . . . . . . . . . . . . . . . . . . .5-1

TCE system requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2

Displaying the hardware revision number . . . . . . . . . . . . . . . . . . . . . . . . 5-3

TCE system specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3

6 Installation and setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-1

Installation procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2

Installing TCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3

Enabling or disabling TCE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-6

Un-installing TCE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-8

Setup procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-11

Setting up DMLUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-11

Setting up data pools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-12

Adding or changing the remote port CHAP secret. . . . . . . . . . . . . . . . . . .6-14

Setting up the remote path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-15

7 Pair operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-1

Operations work flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2

TCE operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3

Checking pair status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3

Creating the initial copy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3

Prerequisites and best practices for pair creation . . . . . . . . . . . . . . . . . . . 7-4

TCE setup wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4

Create pair procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5

Splitting a pair. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6

Resynchronizing a pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-7

Swapping pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9

8 Example scenarios and procedures . . . . . . . . . . . . . . . . . . . . . . . .8-1

CLI scripting procedure for S-VOL backup. . . . . . . . . . . . . . . . . . . . . . . . . . 8-2

Scripted TCE, SnapShot procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4

Procedure for swapping I/O to S-VOL when maintaining local array . . . . . . . 8-9

Contents v

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 6

Procedure for moving data to a remote array. . . . . . . . . . . . . . . . . . . . . . .8-10

Example procedure for moving data . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-11

Process for disaster recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-11

Takeover processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-11

Swapping P-VOL and S-VOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-12

Failback to the local array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-12

9 Monitoring and maintenance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-1

Monitoring pair status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-2

Monitoring data pool capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-5

Monitoring data pool usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-5

Expanding data pool capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-6

Changing data pool threshold value . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-6

Monitoring the remote path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-7

Changing remote path bandwidth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-7

Monitoring cycle time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-7

Changing cycle time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-8

Changing copy pace. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-8

Checking RPO — Monitoring P-VOL/S-VOL time difference. . . . . . . . . . . . . . .9-9

Routine maintenance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-9

Deleting a volume pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-9

Deleting data pools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-10

Deleting a DMLU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-10

Deleting the remote path. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-11

TCE tasks before a planned remote array shutdown. . . . . . . . . . . . . . . . .9-11

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-11

10 Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-1

Troubleshooting overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-2

Correcting data pool shortage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-2

Correcting array problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-3

Delays in settling of S-VOL Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-4

Correcting resynchronization errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-4

Using the event log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-7

Recovering a TrueCopy path. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-8

Miscellaneous troubleshooting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10-10

A Operations using CLI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1

Installation and setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A-2

Installing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A-3

Enabling and disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A-5

Un-installing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A-8

Setting the Differential Management Logical Unit. . . . . . . . . . . . . . . . . . .A-10

vi Contents

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 7

Release a DMLU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-10

Setting the data pool. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-10

Setting the LU ownership. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-12

Setting the cycle time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-13

Setting mapping information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-13

Setting the remote port CHAP secret. . . . . . . . . . . . . . . . . . . . . . . . . . . A-14

Setting the remote path. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-15

Deleting the remote path. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-17

Pair operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-20

Displaying status for all pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-20

Displaying detail for a specific pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-20

Creating a pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-21

Splitting a pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-21

Resynchronizing a pair. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-22

Swapping a pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-22

Deleting a pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-22

Changing pair information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-23

Monitoring pair status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-23

Confirming Consistency Group (CTG) status. . . . . . . . . . . . . . . . . . . . . . A-24

Procedures for failure recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-25

Displaying the event log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-25

Reconstructing the remote path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-25

Sample script. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-26

B Operations using CCI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-1

Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-2

Setting the command device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-2

Setting LU mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-3

Defining the configuration definition file . . . . . . . . . . . . . . . . . . . . . . . . . B-4

Setting the environment variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-6

Pair operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-7

Checking pair status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-7

Creating a pair (paircreate) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-8

Splitting a pair (pairsplit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-8

Resynchronizing a pair (pairresync). . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-9

Suspending pairs (pairsplit -R) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-9

Releasing pairs (pairsplit -S). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-10

Splitting TCE S-VOL/SnapShot V-VOL pair (pairsplit -mscas) . . . . . . . . . . B-10

Confirming data transfer when status is PAIR . . . . . . . . . . . . . . . . . . . . B-13

Pair creation/resynchronization for each CTG. . . . . . . . . . . . . . . . . . . . . B-13

Response time of pairsplit command. . . . . . . . . . . . . . . . . . . . . . . . . . . B-15

Pair, group name differences in CCI and Navigator 2. . . . . . . . . . . . . . . . . B-18

C Cascading with SnapShot. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-1

Contents vii

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 8

Cascade configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-2

Replication operations supported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-2

TCE operations supported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-3

SnapShot operations supported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-4

Status combinations, Read/write supported . . . . . . . . . . . . . . . . . . . . . . . . .C-4

Guidelines and restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-6

Cascading with SnapShot on the remote side . . . . . . . . . . . . . . . . . . . . . .C-6

TCE, SnapShot behaviors compared . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-7

D Installing TCE when Cache Partition Manager is in use . . . . . . D-1

Initializing Cache Partition when TCE and SnapShot are installed . . . . . . . . . D-2

E Wavelength Division Multiplexing (WDM) and dark fibre. . . . . . .E-1

WDM and dark fibre. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-2

Glossary

Index

viii Contents

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 9

Preface

This document provides instructions for planning, setting up, and operating TrueCopy Extended Distance.

This preface includes the following information:

 Intended audience

 Product version

 Release notes and readme

 Changes in this release

 Document organization

 Document conventions

 Convention for storage capacity values

 Related documents

 Getting help

Preface ix

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 10

Intended audience

This document is intended for system administrators, Hitachi Data Systems representatives, and Authorized Service Providers who install, configure, and operate Hitachi Adaptable Modular System (AMS) 2000 family storage systems.

Product version

This document applies to Hitachi AMS 2000 Family firmware version 08D1/B or later.

Release notes and readme

Read the release notes and readme file before installing and using this product. They may contain requirements or restrictions that are not fully described in this document and/or updates or corrections to this document

Product Abbreviations

Product Abbreviation

ShadowImage ShadowImage In-system Replication

Snapshot Copy-on-Write Snapshot

TrueCopy Remote TrueCopy Remote Replication

TCE TrueCopy Extended Distance

TCMD TrueCopy Modular Distributed

Windows Server Windows Server 2003, Windows Server 2008,

and Windows Server 2012.

Product Full Name

x Preface

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 11

Changes in this release

•In Table 5-2 (page 5-3), added the parameter Remote Copy over iSCSI in the WAN environment.

Document organization

Thumbnail descriptions of the chapters are provided in the following table. Click the chapter title in the first column to go to that chapter. The first page of every chapter or appendix contains links to the contents.

Chapter/Appendix

Title

Chapter 1, Overview Provides descriptions of TrueCopy Extended Distance

components and how they work together.

Chapter 2, Plan and design — sizing data pools and bandwidth

Chapter 3, Plan and design — remote path

Chapter 4, Plan and design—arrays, volumes and operating systems

Chapter 5, Requirements and specifications

Chapter 6, Installation and setup

Chapter 7, Pair operations

Chapter 8, Example scenarios and procedures

Chapter 9, Monitoring and maintenance

Chapter 10, Troubleshooting

Appendix A, Operations using CLI

Appendix B, Operations using CCI

Appendix C, Cascading with SnapShot

Appendix D, Installing TCE when Cache Partition Manager is in use

Provides instructions for measuring write-workload, calculating data pool size and bandwidth.

Provides supported iSCSI and Fibre Channel configurations, with information on WDM and dark fibre.

Discusses the arrays and volumes you can use for TCE.

Provides TCE system requirements and specifications.

Provides procedures for installing and setting up the TCE system and creating the initial copy.

Provides information and procedures for TCE operations.

Provides backup, data moving, and disaster recovery scenarios and procedures.

Provides monitoring and maintenance information.

Provides troubleshooting information.

Provides detailed Command Line Interface instructions for configuring and using TCE.

Provides supported configurations, operations, etc. for cascading TCE with SnapShot.

Provides required information when using Cache Partition Manager.

Description

Preface xi

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 12

Chapter/Appendix

Title

Description

Appendix E, Wavelength

Provides a discussion of WDM and dark fibre for channel extender.

Division Multiplexing (WDM) and dark fibre

Glossary Provides definitions for terms and acronyms found in this

document.

Index Provides links and locations to specific information in this

document.

xii Preface

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 13

Document conventions

This document uses the following symbols to draw attention to important safety and operational information.

Symbol Meaning Description

Tip Tips provide helpful information, guidelines, or suggestions for

Note Notes emphasize or supplement important points of the main

Caution Cautions indicate that failure to take a specified action could

The following typographic conventions are used in this document.

Convention 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 text provided by

the user or system. Example: copy source-file target-file

Angled brackets (< >) are also used to indicate variables.

screen/code Indicates text that is displayed on screen or entered by the user.

Example: # pairdisplay -g oradb

< > angled brackets

Indicates a variable, which is a placeholder for actual text provided by

the user or system. Example: # pairdisplay -g <group>

performing tasks more effectively.

text.

result in damage to the software or hardware.

Italic font is also used to indicate variables.

[ ] square brackets

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

| vertical bar Indicates that you have a choice between two or more options or

underline Indicates the default value. Example: [ a | b ]

Indicates optional values. Example: [ a | b ] indicates that you can choose a, b, or nothing.

you must choose either a or b.

arguments. Examples: [ a | b ] indicates that you can choose a, b, or nothing. { a | b } indicates that you must choose either a or b.

Preface xiii

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 14

Convention for storage capacity values

Physical storage capacity values (e.g., disk drive capacity) are calculated based on the following values:

Physical capacity unit Value

1 KB 1,000 bytes

1 MB 1,000 KB or 1,000

1 GB 1,000 MB or 1,0003 bytes

1 TB 1,000 GB or 1,0004 bytes

1 PB 1,000 TB or 1,0005 bytes

1 EB 1,000 PB or 1,000

Logical storage capacity values (e.g., logical device capacity) are calculated based on the following values:

Logical capacity unit Value

1 block 512 bytes

1 KB 1,024 (210) bytes

1 MB 1,024 KB or 1024

1 GB 1,024 MB or 10243 bytes

1 TB 1,024 GB or 10244 bytes

1 PB 1,024 TB or 1024

1 EB 1,024 PB or 10246 bytes

bytes

Getting help

If you need to contact the Hitachi Data Systems support center, please provide as much information about the problem as possible, including:

Comments

• The circumstances surrounding the error or failure.

• The exact content of any messages displayed on the host systems.

• The exact content of any messages displayed on Storage Navigator Modular 2.

• The Storage Navigator Modular 2 configuration information. This information is used by service personnel for troubleshooting purposes.

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

Ple as e se nd u s y o ur co m me nt s on th i s d oc u me nt : doc.comments@hds.com. Include the document title, number, and revision, and refer to specific sections and paragraphs whenever possible.

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

Preface xix

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 20

xx Preface

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 21

Overview

This manual provides instructions for designing, planning, implementing, using, monitoring, and troubleshooting TrueCopy Extended Distance (TCE). This chapter consists of:

 How TCE works

 Typical environment

 TCE interfaces

Overview 1–1

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 22

How TCE works

With TrueCopy Extended Distance (TCE), you create a copy of your data at a remote location. After the initial copy is created, only changed data transfers to the remote location.

You create a TCE copy when you:

• Select a volume on the production array that you want to replicate

• Create a volume on the remote array that will contain the copy

• Establish a Fibre Channel or iSCSI link between the local and remote arrays

• Make the initial copy across the link on the remote array.

During and after the initial copy, the primary volume on the local side continues to be updated with data from the host application. When the host writes data to the P-VOL, the local array immediately returns a response to the host. This completes the I/O processing. The array performs the subsequent processing independently from I/O processing.

Updates are periodically sent to the secondary volume on the remote side at the end of the “update cycle”. This is a time period established by the user. The cycle time is based on the recovery point objective (RPO), which is the amount of data in time (2-hours’ worth, 4 hour’s worth) that can be lost after a disaster, until the operation is irreparably damaged. If the RPO is two hours, the business must be able to recover all data up to two hours before the disaster occurred.

When a disaster occurs, storage operations are transferred to the remote site and the secondary volume becomes the production volume. All the original data is available in the S-VOL, from the last completed update. The update cycle is determined by your RPO and by measuring write-workload during the TCE planning and design process.

For a detailed discussion of the disaster recovery process using TCE, please refer to Process for disaster recovery on page 8-11.

Typical environment

A typical configuration consists of the following elements. Many but not all require user set up.

• Two AMS arrays—one on the local side connected to a host, and one on the remote side connected to the local array. Connections are made via Fibre Channel or iSCSI.

• A primary volume on the local array that is to be copied to the secondary volume on the remote side.

• A differential management LU on local and remote arrays, which hold TCE information when the array is powered down

1–2 Overview

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 23

• Interface and command software, used to perform TCE operations. Command software uses a command device (volume) to communicate with the arrays.

Figure 1-1 shows a typical TCE environment.

•

Volume pairs

When the initial TCE copy is completed, the production and backup volumes are said to be “Paired”. The two paired volumes are referred to as the primary volume (P-VOL) and secondary volume (S-VOL). Each TCE pair consists of one P-VOL and one S-VOL. When the pair relationship is established, data flows from the P-VOL to the S-VOL.

While in the Paired status, new data is written to the P-VOL and then periodically transferred to the S-VOL, according to the user-defined update cycle.

When a pair is “split”, the data flow between the volumes stops. At this time, all the differential data that has accumulated in the local array since the last update is copied to the S-VOL. This insures that its data is the same as the P-VOL’s and is consistent and usable data.

During normal TCE operations, the P-VOL remains available for read/write from the host. When the pair is split, the S-VOL also is available for read/ write operations from a host.

Figure 1-1: Typical TCE Environment

Overview 1–3

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 24

Data pools

Data from the host is continually updated to the P-VOL, as it occurs. The data pool on the local side stores the changed data that accumulates before the next the update cycle. The local data pool is used to update the S-VOL.

Data that accumulates in the data pool is referred to as differential data because it contains the difference data between the P-VOL and S-VOL.

The data in the S-VOL following an update is complete, consistent, and usable data. When the next update is to begin, this consistent data is copied to the remote data pool. This data pool is used to maintain previous pointin-time copies of the S-VOL, which are used in the event of failback.

Guaranteed write order and the update cycle

S-VOL data must have the same order in which the host updates the P-VOL. When write order is guaranteed, the S-VOL has data consistency with the P-VOL.

As explained in the previous section, data is copied from the P-VOL and local data pool to the S-VOL following the update cycle. When the update is

complete, S-VOL data is identical to P-VOL data at the end of the cycle.

Since the P-VOL continues to be updated while and after the S-VOL is being updated, S-VOL data and P-VOL data are not identical.

However, the S-VOL and P-VOL can be made identical when the pair is split. During this operation, all differential data in the local data pool is transferred to the S-VOL, as well as all cached data in host memory. This cached data is flushed to the P-VOL, then transferred to the S-VOL as part of the split operation, thus ensuring that the two are identical.

If a failure occurs during an update cycle, the data in the update is inconsistent. Write order in the S-VOL is nevertheless guaranteed — at the point-in-time of the previous update cycle, which is stored in the remote data pool.

Figure 1-2 shows how S-VOL data is maintained at one update cycle back

of P-VOL data.

1–4 Overview

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 25

•

Extended update cycles

If inflow to the P-VOL increases, all of the update data may not be sent within the cycle time. This causes the cycle to extend beyond the userspecified cycle time.

As a result, more update data in the P-VOL accumulates to be copied at the next update. Also, the time difference between the P-VOL data and S-VOL data increases, which degrades the recovery point value. In Figure 1-2, if a failure occurs at the primary site immediately before time T3, for example, data consistency in the S-VOL during takeover is P-VOL data at time T1.

When inflow decreases, updates again complete within the cycle time. Cycle time should be determined according to a realistic assessment of write workload, as discussed in Chapter 2, Plan and design — sizing data pools

and bandwidth.

Figure 1-2: Update Cycles and Differential Data

Overview 1–5

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 26

Consistency groups

Application data often spans more than one volume. With TCE, it is possible to manage operations spanning multiple volumes as a single group. In a consistency group (CTG), all primary logical volumes are treated as a single entity.

Managing primary volumes as a consistency group allows TCE operations to be performed on all volumes in the group concurrently. Write order in secondary volumes is guaranteed across application logical volumes.

Figure 1-3 shows TCE operations with a consistency group.

•

Figure 1-3: TCE Operations with Consistency Groups

In this illustration, observe the following:

• The P-VOLs belong to the same consistency group. The host updates the P-VOLs as required (1).

• The local array identifies the differential data in the P-VOLs when the cycle is started (2) in an atomic manner. The differential data of the group of the P-VOLs are determined at time T2.

• The local array transfers the differential data to the corresponding SVOLs (3). When all differential data is transferred, each S-VOL is identical to its P-VOL at time T2 (4).

• If pairs are split or deleted, the local array stops the cycle update for the consistency group. Differential data between P-VOLs and S-VOLs is determined at that time. All differential data is sent to the S-VOLs, and the split or delete operations on the pairs completes. S-VOLs maintain

1–6 Overview

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 27

data consistency across pairs in the consistency group. The pair that is using different data pool can be belongs to the same consistency group.

Differential Management LUs (DMLU)

The DMLU is an exclusive volume used for storing TrueCopy information when the local or remote array is powered down. The DMLU is hidden from a host. User setup is required on the local and remote arrays.

TCE interfaces

TCE can be setup, used and monitored using of the following interfaces:

• The GUI (Hitachi Storage Navigator Modular 2 Graphical User Interface), which is a browser-based interface from which TCE can be setup, operated, and monitored. The GUI provides the simplest method for performing operations, requiring no previous experience. Scripting is not available.

• CLI (Hitachi Storage Navigator Modular 2 Command Line Interface), from which TCE can be setup and all basic pair operations can be performed—create, split, resynchronize, restore, swap, and delete. The GUI also provides these functionalities. CLI also has scripting capability.

• CCI (Hitachi Command Control Interface (CCI), which is used to display volume information and perform all copying and pair-managing operations. CCI provides a full scripting capability which can be used to automate replication operations. CCI requires more experience than the GUI or CLI. CCI is required for performing failover and fall back operations, and, on Windows 2000 Server, mount/unmount operations.

HDS recommends using the GUI to begin operations for new users with no experience with CLI or CCI. Users who are new to replication software but have CLI experience in managing arrays may want to continue using CLI, though the GUI is an option. The same recommendation applies to CCI users.

Overview 1–7

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 28

1–8 Overview

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 29

Plan and design — sizing

data pools and bandwidth

This chapter provides instructions for measuring write-workload and sizing data pools and bandwidth.

 Plan and design workflow

 Assessing business needs — RPO and the update cycle

 Measuring write-workload

 Calculating data pool size

 Determining bandwidth

Plan and design — sizing data pools and bandwidth 2–1

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 30

Plan and design workflow

You design your TCE system around the write-workload generated by your host application. Data pools and bandwidth must be sized to accommodate write-workload. This chapter helps you perform these tasks as follows:

• Assess business requirements regarding how much data your operation must recover in the event of a disaster.

• Measure write-workload. This metric is used to ensure that data pool size and bandwidth are sufficient to hold and pass all levels of I/O.

• Calculate data pool size. Instructions are included for matching data pool capacity to the production environment.

• Calculate remote path bandwidth: This will make certain that you can copy your data to the remote site within your update cycle.

Assessing business needs — RPO and the update cycle

In a TCE system, the S-VOL will contain nearly all of the data that is in the P-VOL. The difference between them at any time will be the differential data that accumulates during the TCE update cycle.

This differential data accumulates in the local data pool until the update cycle starts, then it is transferred over the remote data path.

Update cycle time is a uniform interval of time during which differential data copies to the S-VOL. You will define the update cycle time when creating the TCE pair.

The update cycle time is based on:

• the amount of data written to your P-VOL

• the maximum amount of data loss your operation could survive during a disaster.

The data loss that your operation can survive and remain viable determines to what point in the past you must recover.

An ho ur ’s wo rth of da ta l os s me an s t hat yo ur re cov er y p oi nt i s o ne hou r a go . If disaster occurs at 10:00 am, upon recovery your restart will resume operations with data from 9:00 am.

Fifteen minutes worth of data loss means that your recovery point is 15 minutes prior to the disaster.

You must determine your recovery point objective (RPO). You can do this by measuring your host application’s write-workload. This shows the amount of data written to the P-VOL over time. You or your organization’s decisionmakers can use this information to decide the number of business transactions that can be lost, the number of hours required to key in lost data and so on. The result is the RPO.

2–2 Plan and design — sizing data pools and bandwidth

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 31

Measuring write-workload

Bandwidth and data pool size are determined by understanding the writeworkload placed on the primary volume from the host application.

• After the initial copy, TCE only copies changed data to the S-VOL.

• Data is changed when the host application writes to storage.

• Write-workload is a measure of changed data over a period of time.

When you know how much data is changing, you can plan the size of your data pools and bandwidth to support your environment.

Collecting write-workload data

Workload data is collected using your operating system’s performance monitoring feature. Collection should be performed during the busiest time of month, quarter, and year so you can be sure your TCE implementation will support your environment when demand is greatest. The following procedure is provided to help you collect write-workload data.

To collect workload data

1. Using your operating system’s performance monitoring software, collect the following:

- Disk-write bytes-per-second for every physical volume that will be

replicated.

- Collect this data at 10 minute intervals and over as long a period

as possible. Hitachi recommends a 4-6 week period in order to accumulate data over all workload conditions including times when the demands on the system are greatest.

•

2. At the end of the collection period, convert the data to MB/second and import into a spreadsheet tool. In Figure 2-1, Write-Workload

Spreadsheet, column C shows an example of collected raw data over 10-

minute segments.

Plan and design — sizing data pools and bandwidth 2–3

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 32

•

Figure 2-1: Write-Workload Spreadsheet

Fluctuations in write-workload can be seen from interval to interval. To calculate data pool size, the interval data will first be averaged, then used in an equation. (Your spreadsheet at this point would have only rows B and C populated.)

Calculating data pool size

In addition to write-workload data, cycle time must be known. Cycle time is the frequency that updates are sent to the remote array. This is a userdefined value that can range from 30 seconds to 1 hour. The default cycle time is 5-minutes (300 seconds). If consistency groups are used, the minimum must be 30 seconds for one CTG, increasing 30 seconds for each additional CTG, up to 16. Since the data pool stores all updated data that accumulates during the cycle time, the longer the cycle time, the larger the data pool must be. For more information on cycle time, see the discussion in Assessing business needs — RPO and the update cycle on page 2-2, and also Changing cycle time on page 9-8.

To calculate TCE data pool capacity

1. Using write-workload data imported into a spreadsheet tool and your cycle time, calculate write rolling-averages, as follows. (Most spreadsheet tools have an average function.)

- If cycle time is 1 hour, then calculate 60 minute rolling averages.

Do this by arranging the values in six 10-minute intervals.

- If cycle time is 30 minutes, then calculate 30 minute rolling

averages, arranging the values in three 10-minute intervals.

2–4 Plan and design — sizing data pools and bandwidth

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 33

Example rolling-average procedure for cycle time in Microsoft Excel

Cycle time in the following example is 1 hour; rolling averages are calculated using six 10-minute intervals.

a. After converting workload data into the spreadsheet (Figure 2-1,

Write-Workload Spreadsheet), in cell E4 type, =average(b2:b7),

and press Enter.

This instructs the tool to calculate the average value in cells B2 through B7 (six 10-minute intervals) and populate cell E4 with that

data. (The calculations used here are for example purposes only. Base your calculations on your cycle time.)

b. Copy the value that displays in E4.

c. Highlight cells E5 to the E cell in the last row of workload data in the

spreadsheet.

d. Right-click the highlighted cells and select the Paste option.

Excel maintains the logic and increments the formula values initially entered in E4. It then calculates all the 60-minute averages for every 10-minute increment, and populates the E cells, as shown in

Figure 2-2.

•

Figure 2-2: Rolling Averages Calculated Using 60 Minute Cycle Time

Plan and design — sizing data pools and bandwidth 2–5

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 34

For another perspective, you can graph the data, as shown in

Figure 2-3.

•

Figure 2-3: 60-Minute Rolling Averages Graphed Over Raw Data

2. From the spreadsheet or graph, locate the largest value in the E column. This is your Peak Rolling Average (PRA) value. Use the PRA to calculate

the cumulative peak data change over cycle time. The following formula

calculates the largest expected data change over the cycle time. This will ensure that you do not overflow your data pool.

(PRA in MB/sec) x (cycle time seconds) = (Cumulative peak data change)

For example, if the PRA is 3 MB/sec, and the cycle time is 3600 seconds (1 hour), then:

3MB/sec x 3600 seconds = 10,800 MB

This shows the maximum amount of changed data (pool data) that you can expect in a 60 minute time period. This is the base data pool size required for TCE.

3. Hitachi recommends a 20-percent safety factor for data pools. Calculate a safety factor with the following formula:

(Combined base data pool size) x 1.2. For example: 529,200 MB x 1.2 = 635,040 MB

2–6 Plan and design — sizing data pools and bandwidth

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 35

4. It is also recommended that annual increases in data transactions be factored into data pool sizing. This is done to minimize reconfiguration in the future. Do this by multiplying the pool size with safety factor by the percentage of expected annual growth. For example:

635,040 MB x 1.2 (20 percent growth rate for per year) = 762,048 MB

Repeat this step for each year the solution will be in place.

5. Convert to gigabytes, dividing by 1,000. For example:

762,048 MB / 1,000 = 762 GB

This is the size of the example data pool with safety and growth (2nd year) factored in.

Data pool key points

• Data pools must be set up on the local array and the remote array.

• The data pool must be on the same controller as the P-VOL and VVOL(s).

• Up to 64 LUs can be assigned to a data pool.

• Plan for highest workload and multi-year growth.

• For set up information, see Setting up data pools on page 6-12.

Plan and design — sizing data pools and bandwidth 2–7

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 36

Determining bandwidth

The purpose of this section is to ensure that you have sufficient bandwidth between the local and remote arrays to copy all your write data in the timeframe you prescribe. The goal is to size the network so that it is capable of transferring estimated future write workloads.

TCE requires two remote paths, each with a minimum bandwidth of 1.5 Mbs.

To determine the bandwidth

1. Graph the data in column “C” in the Write-Workload Spreadsheet on

page 2-4.

2. Locate the highest peak. Based on your write-workload measurements, this is the greatest amount of data that will need to be transferred to the remote array. Bandwidth must accommodate maximum possible workload to insure that the system does not become subject to its capacity being exceeded. This would cause further problems, such as the new write data backing up in the data pool, update cycles becoming extended, and so on.

3. Though the highest peak in your workload data should be used for determining bandwidth, you should also take notice of extremely high peaks. In some cases a batch job, defragmentation, or other process could be driving workload to abnormally high levels. It is sometimes worthwhile to review the processes that are running. After careful analysis, it may be possible to lower or even eliminate some spikes by optimizing or streamlining high-workload processes. Changing the timing of a process may lower workload.

4. Although bandwidth can be increased, Hitachi recommends that projected growth rate be factored over a 1, 2, or 3 year period.

Table 2-1 shows TCE bandwidth requirements.

Table 2-1: Bandwidth Requirements

Average Inflow Bandwidth Requirements WAN Types

.08 - .149 MB/s 1.5 Mb/s or more T1

.15 - .299 MB/s 3 Mb/s or more T1 x two lines

.3 - .599 MB/s 6 Mb/s or more T2

.6 - 1.199 MB/s 12 Mb/s or more T2 x two lines

1.2 - 4.499 MB/s 45 Mb/s or more T3

4.500 - 9.999 MB/s 100 Mb/s or more Fast Ethernet

2–8 Plan and design — sizing data pools and bandwidth

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 37

Plan and design — remote

path

A remote path is required for transferring data from the local array to the remote array. This chapter provides network and bandwidth requirements, and supported remote path configurations.

 Remote path requirements

 Remote path configurations

 Using the remote path — best practices

Plan and design — remote path 3–1

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 38

Remote path requirements

The remote path is the connection used to transfer data between the local array and remote array. TCE supports Fibre Channel and iSCSI port connectors and connections. The connections you use must be either one or the other: they cannot be mixed.

The following kinds of networks are used with TCE:

• Local Area Network (LAN), for system management. Fast Ethernet is required for the LAN.

• Wide Area Network (WAN) for the remote path. For best performance:

- A Fibre Channel extender is required.

- iSCSI connections may require a WAN Optimization Controller

(WOC).

Figure 3-1 shows the basic TCE configuration with a LAN and WAN.

•

Figure 3-1: Remote Path Configuration

Requirements are provided in the following:

• Management LAN requirements on page 3-3

• Remote data path requirements on page 3-3

• WAN optimization controller (WOC) requirements on page 3-4

• Fibre channel extender connection on page 3-9.

3–2 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 39

Management LAN requirements

Fast Ethernet is required for an IP LAN.

Remote data path requirements

This section discusses the TCE remote path requirements for a WAN connection. This includes the following:

•Types of lines

•Bandwidth

• Distance between local and remote sites

• WAN Optimization Controllers (WOC) (optional)

For instructions on assessing your system’s I/O and bandwidth requirements, see:

• Measuring write-workload on page 2-3

• Determining bandwidth on page 2-8

Table 3-1 provides remote path requirements for TCE. A WOC may also be

required, depending on the distance between the local and remote sites and other factors listed in Table 3-3.

•

Table 3-1: Remote Data Path Requirements

Item Requirements

Bandwidth • Bandwidth must be guaranteed.

• Bandwidth must be 1.5 Mb/s or more for each pair. 100 Mb/s recommended.

• Requirements for bandwidth depend on an average inflow from the host into the array.

•See Table 2-1 on page 2-8 for bandwidth requirements.

Remote Path Sharing • The remote path must be dedicated for TCE pairs.

• When two or more pairs share the same path, a WOC is recommended for each pair

•

Table 3-2 shows types of WAN cabling and protocols supported by TCE and

those not supported.

•

Table 3-2: Supported, Not Supported WAN Types

WAN Types

Supported • Dedicated Line (T1, T2, T3 etc)

Not-supported • ADSL, CATV, FTTH, ISDN

Plan and design — remote path 3–3

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 40

WAN optimization controller (WOC) requirements

WAN Optimization Controller (WOC) is a network appliance that enhances WAN performance by accelerating long-distance TCP/IP communications. TCE copy performance over longer distances is significantly increased when WOC is used. A WOC guarantees bandwidth for each line.

•Use Table 3-3 to determine whether your TCE system requires the addition of a WOC.

• Table 3-4 shows the requirements for WOCs.

•

Table 3-3: Conditions Requiring a WOC

Item Condition

Latency, Distance • If round trip time is 5 ms or more, or distance

between the local site and the remote site is 100 miles (160 km) or further, WOC is highly recommended.

WAN Sharing • If two or more pairs share the same WAN, A WOC

is recommended for each pair.

•

Table 3-4: WOC Requirements

Item Requirements

LAN Interface • Gigabit Ethernet or fast Ethernet must be

supported.

Performance • Data transfer capability must be equal to or more

than bandwidth of WAN.

Functions • Traffic shaping, bandwidth throttling, or rate

limiting must be supported. These functions reduce data transfer rates to a value input by the user.

• Data compression must be supported.

• TCP acceleration must be supported.

3–4 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 41

Remote path configurations

TCE supports both Fibre Channel and iSCSI connections for the remote path.

• Two remote paths must be set up between, one per controller. This ensures that an alternate path is available in the event of link failure during copy operations.

• Paths can be configured from:

- Local controller 0 to remote controller 0 or 1

- Local controller 1 to remote controller 0 or 1

• Paths can connect a port A with a port B, and so on. Hitachi recommends making connections between the same controller/port, such as port 0B to 0B, and 1 B to 1 B, for simplicity. Ports can be used for both host I/O and replication data.

The following sections describe supported Fibre Channel and iSCSI path configurations. Recommendations and restrictions are included.

Fibre channel

The Fibre Channel remote data path can be set up in the following configurations:

• Direct connection

• Single Fibre Channel switch and network connection

• Double FC switch and network connection

• Wavelength Division Multiplexing (WDM) and dark fibre extender

The array supports direct or switch connection only. Hub connections are not supported.

General recommendations

The following is recommended for all supported configurations:

• TCE requires one path between the host and local array. However, two paths are recommended; the second path can be used in the event of a path failure.

Plan and design — remote path 3–5

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 42

Direct connection

Figure 3-2 illustrates two remote paths directly connecting the local and

remote arrays. This configuration can be used when distance is very short, as when creating the initial copy or performing data recovery while both arrays are installed at the local site.

•

Figure 3-2: Direct FC Connection

3–6 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 43

Single FC switch, network connection

Switch connections increase throughput between the arrays. Figure 3-3 illustrates two remote paths routed through one FC switch and one FC network to make the connection to the remote site.

•

Figure 3-3: Single FC Switch, Network Connection

Recommendations

• While this configuration may be used, it is not recommended since failure in an FC switch or the network would halt copy operations.

• Separate switches should be set up for host I/O to the local array and for data transfer between arrays. Using one switch for both functions results in deteriorated performance.

Plan and design — remote path 3–7

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 44

Double FC switch, network connection

Figure 3-4 illustrates two remote paths using two FC switches and two FC

networks to make the connection to the remote site.

•

Figure 3-4: Double FC Switches, Networks Connection

Recommendations

• Separate switches should be set up for host I/O to the local array and for data transfer between arrays. Using one switch for both functions results in deteriorated performance.

3–8 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 45

Fibre channel extender connection

Channel extenders convert Fibre Channel to FCIP or iFCP, which allows you to use IP networks and significantly improve performance over longer distances.

Figure 3-5 illustrates two remote paths using two FC switches, Wavelength

Division Multiplexor (WDM) extender, and dark fibre to make the connection to the remote site.

•

Figure 3-5: Fibre Channel Switches, WDM, Dark Fibre Connection

Recommendations

• Only qualified components are supported.

For more information on WDM, see Appendix E, Wavelength Division

Multiplexing (WDM) and dark fibre.

Plan and design — remote path 3–9

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 46

Port transfer rate for Fibre channel

The communication speed of the Fibre Channel port on the array must match the speed specified on the host port. These two ports—Fibre Channel port on the array and host port—are connected via the Fibre Channel cable. Each port on the array must be set separately.

•

Table 3-5: Setting Port Transfer Rates

Set the remote array

port to

Manual mode

Auto mode

If the host port is set to

1 Gbps 1 Gbps

2 Gbps 2 Gbps

4 Gbps 4 Gbps

8 Gbps 8 Gbps

2 Gbps Auto, with max of 2 Gbps

4 Gbps Auto, with max of 4 Gbps

8 Gbps Auto, with max of 8 Gbps

Maximum speed is ensured using the manual settings.

You can specify the port transfer rate using the Navigator 2 GUI, on the Edit FC Port screen (Settings/FC Settings/port/Edit Port button).

•

NOTE: If your remote path is a direct connection, make sure that the

array power is off when modifying the transfer rate to prevent remote path blockage.

Find details on communication settings in the Hitachi AMS 2100/2300 Storage System Hardware Guide.

3–10 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 47

iSCSI

The iSCSI remote data path can be set up in the following configurations:

• Direct connection

• Local Area Network (LAN) switch connections

• Wide Area Network (WAN) connections

• WAN Optimization Controller (WOC) connections

Recommendations

The following is recommended for all supported configurations:

• Two paths should be configured from the host to the array. This provides a backup path in the event of path failure.

Direct connection

Figure 3-6, illustrates two remote paths directly connecting the local and

remote arrays. Direct connections are used when the local and remote arrays are set up at the same site. In this case, category 5e or 6 copper LAN cable is recommended.

•

Figure 3-6: Direct iSCSI Connection

Recommendations

• When a large amount of data is to be copied to the remote site, the initial copy between local side and remote systems may be performed at the same location. In this case, category 5e or 6 copper LAN cable is recommended.

Plan and design — remote path 3–11

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 48

Single LAN switch, WAN connection

Figure 3-7, illustrates two remote paths using one LAN switch and network

to the remote array.

•

Figure 3-7: Single-Switch Connection

Recommendations

• This configuration is not recommended because a failure in a LAN switch or WAN would halt operations.

• Separate LAN switches and paths should be used for host-to-array and array-to-array, for improved performance.

3–12 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 49

Multiple LAN switch, WAN connection

Figure 3-8, illustrates two remote paths using multiple LAN switches and

WANs to make the connection to the remote site.

•

Figure 3-8: Multiple-Switch and WAN Connection

Recommendations

• Separate LAN switches and paths should be used for the host-to-array and the array-to-array paths for better performance and to provide a backup.

Plan and design — remote path 3–13

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 50

Single LAN switch, WOC, WAN connection

WOCs may be required for TCE, depending on your system’s bandwidth, latency, and so on. Use of a WOC improves performance. See WAN

optimization controller (WOC) requirements on page 3-4 for more

information.

Figure 3-9, illustrates two remote paths using a single LAN switch, WOC,

and WAN to make the connection to the remote site.

•

Figure 3-9: Single Switch, WOC, and WAN Connection

3–14 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 51

Multiple LAN switch, WOC, WAN connection

Figure 3-10, illustrates two remote connections using multiple LAN

switches, WOCs, and WANs to make the connection to the remote site.

•

Figure 3-10: Connection Using Multiple Switch, WOC, WAN

Recommendations

• If a Gigabit Ethernet port (1000BASE-T) is provided on the WOC, the LAN switch to the WOC is not required. Connect array ports 0B and 1B to the WOC directly. If your WOC does not have 1Gbps ports, the LAN switch is required.

• Using separate LAN switch, WOC and WAN for each remote path ensures that data copy automatically continues on the second path in the event of a path failure.

Plan and design — remote path 3–15

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 52

Multiple array, LAN switch, WOC connection with single WAN

Figure 3-11, shows two local arrays connected to two remote arrays, each

via a LAN switch and WOC.

Figure 3-11: Multiple Array Connection Using Single WAN

Recommendations

• You can reduce the number of switches by using a switch with VLAN capability. If a VLAN switch is used, port 0B of local array 1 and the WOC1 should be in one LAN (VLAN1); port 0B of local array 2 and WOC3 should be in another LAN (VLAN2). Connect the VLAN2 port directly to Port 0B of the local array 2 and WOC3.

3–16 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 53

Multiple array, LAN switch, WOC connection with two WANs

Figure 3-12, shows two local arrays connected to two remote arrays, each

via two LAN switches, WANs, and WOCs.

•

Figure 3-12: Multiple Array Connection Using Two WANs

Recommendations

• You can reduce the number of switches by using a switch with VLAN capability. If a VLAN switch is used, port 0B of local array 1 and WOC1 should be in one LAN (VLAN1); port 0B of local array 2 and WOC3 should be in another LAN (VLAN2). Connect the VLAN2 port directly to Port 0B of the local array 2 and WOC3.

Plan and design — remote path 3–17

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 54

Supported connections between various types of arrays

Hitachi AMS2100, AMS2300, or AMS2500 can be connected with Hitachi WMS100, AMS200, AMS500, or AMS1000. The following table shows the supported connections between various types of arrays.

Table 3-6: Supported Connections between Various Types of Arrays

Remote Side Array Type

Local Side

Array Type

WMS100

AMS200

AMS200 AMS500 AMS1000 AMS2100 AMS2300 AMS2500 HUS100

WMS100 AMS200

AMS500 Not

AMS1000 Not

AMS2100 Not

AMS2300 Not

AMS2500 Not

HUS100 Not

Not supported

Supported

supported

Not supported

Available Available Available Available Available Available Available

Not supported

Notes when Connecting Hitachi AMS2000 Series to other arrays

• The maximum number of pairs that can be created is limited to the maximum number of pairs supported by the arrays, whichever is fewer.

• The firmware version of AMS500/1000 must be 0780/A or later when connecting with AMS2100, AMS2300, or AMS2500 of the H/W Rev. is

0100.

Not supported

• The firmware version of AMS500/1000 must be 0786/A or later when connecting with AMS2100, AMS2300, or AMS2500 of the H/W Rev. is

0200.

• The firmware version of AMS2100, AMS2300, or AMS2500 must be 08B7/B or later when connecting with HUS100.

• If a Hitachi Unified Storage as the local array connects to an AMS2010, AMS2100, AMS2300, or AMS2500 with under 08B7/B as the remote array, the remote path will be blocked along with the following message:

• For Fibre Channel connection:

The target of remote path cannot be connected(Port-xy) Path alarm(Remote-X,Path-Y)

• For iSCSI connection:

• Path Login failed

3–18 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 55

• The bandwidth of the remote path to AMS500/1000 must be 20 Mbps or more.

• The pair operation of AMS500/1000 cannot be done from Navigator 2.

• Because AMS500 or AMS1000 has only one data pool per controller, the user cannot specify which data pool to use. For that reason, when connecting AMS500 or AMS1000 with AMS2100, AMS2300, or AMS2500, the specifications about the data pools become the followings:

- When AMS500 or AMS1000 is the local array, the data pool 0 is

selected if the LUN of the S-VOL is even, and the data pool 1 is selected if it is odd. In the configuration that the LU numbers of the S-VOL include odd pairs and even pairs, both data pool 0 and data pool 1 are required.

- When AMS2100, AMS2300, or AMS2500 is the local array, the data

pool number is ignored even if specified. The data pool 0 is selected the owner controller of the S-VOL is 0, and data pool 1 is selected if it is 1.

• AMS500 or AMS1000 cannot use the functions that are newly supported by AMS2100, AMS2300, or AMS2500.

• AMS2100, AMS2300, or AMS2500 cannot use the functions that are newly supported by HUS100.

Plan and design — remote path 3–19

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 56

Using the remote path — best practices

The following best practices are provided to reduce and eliminate path failure.

• If both arrays are powered off, power-on the remote array first.

• When powering down both arrays, turn off the local array first.

• Before powering off the remote array, change pair status to Split. In Paired or Synchronizing status, a power-off results in Failure status on the remote array.

• If the remote array is not available during normal operations, a blockage error results with a notice regarding SNMP Agent Support Function and TRAP. In this case, follow instructions in the notice.

Path blockage automatically recovers after restarting. If the path blockage is not recovered when the array is READY, contact Hitachi Customer Support.

• Power off the arrays before performing the following operation:

- Setting or changing the fibre transfer rate

3–20 Plan and design — remote path

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 57

Plan and design—arrays,

volumes and operating

systems

This chapter provides the information you need to prepare your arrays and volumes for TCE operations

 Planning arrays—moving data from earlier AMS models

 Planning logical units for TCE volumes

 Operating system recommendations and restrictions

 Maximum supported capacity

Plan and design—arrays, volumes and operating systems 4–1

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 58

Planning workflow

Planning a TCE system consists of determining business requirements for recovering data, measuring production write-workload and sizing data pools and bandwidth, designing the remote path, and planning your arrays and volumes. This chapter discusses arrays and volumes as follows:

• Requirements and recommendations for using previous versions of AMS with the AMS 2000 Family.

• Logical unit set up: LUs must be set up on the arrays before TCE is implemented. Volume requirements and specifications are provided.

• Operating system considerations: Operating systems have specific restrictions for replication volumes pairs. These restrictions plus recommendations are provided.

• Maximum Capacity Calculations: Required to make certain that your array has enough capacity to support TCE. Instructions are provided for calculating your volumes’ maximum capacity.

Planning arrays—moving data from earlier AMS models

Logical units on AMS 2100, 2300, and 2500 systems can be paired with logical units on AMS 500 and AMS 1000 systems. Any combination of these arrays may be used on the local and remote sides.

TCE pairs with WMS 100 and AMS 200 are not supported with AMS2100, 2300, or 2500.

When using the earlier model arrays, please observe the following:

• The bandwidth of the remote path to AMS 500 or AMS 1000 must be 20 Mbps or more.

• The maximum number of pairs between different model arrays is limited to the maximum number of pairs supported by the smallest array.

• The firmware version of AMS 500 or AMS 1000 must be 0780/A or later when pairing with an AMS 2100, 2300, or 2500 where the hardware Rev is 0100.

• The firmware version of AMS 500 or AMS 1000 must be 0786/A or later when pairing with an AMS 2010, 2100, 2300, or 2500 where the hardware Rev is 0200.

• Pair operations for AMS 500 and AMS 1000 cannot be performed using the Navigator 2 GUI.

• AMS500 and AMS1000 cannot use functions that are newly supported by AMS2100 or AMS2300.

• Because AMS 500 or AMS 1000 can have only one data pool per controller, you are not able to specify which data pool to use. Because of this, the data pool that is used is determined as follows:

- When AMS 500 or AMS 1000 is the local array, data pool 0 is used

if the S-VOL LUN is even; data pool 1 is used if the S-VOL LUN is odd.

4–2 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 59

- When an AMS 2100, 2300, or 2500 is the local array, the data pool number is ignored even if specified. Data pool 0 is used if the SVOL owner-controller is 0, and data pool 1 is selected if the S-VOL owner-controller is 1.

• The AMS 500 or AMS 1000 cannot use the functions that are newly supported by AMS 2010, 2100, 2300, or 2500.

Planning logical units for TCE volumes

Please review the recommendations in the following sections before setting up TrueCopy volumes. Also, review Requirements and specifications on

page 5-1.

Volume pair and data pool recommendations

• The P-VOL and S-VOL must be identical in size, with matching block count. To check block size, in the Navigator 2 GUI, navigate to Groups/ RAID Groups/Logical Units tab. Click the desired LUN. On the popup window that appears, review the Capacity field. This shows block size.

• The number of volumes within the same RAID group should be limited. Pair creation or resynchronization for one of the volumes may impact I/O performance for the others because of contention between drives.

When creating two or more pairs within the same RAID group, standardize the controllers for the LUs in the RAID group. Also, perform pair creation and resynchronization when I/O to other volumes in the RAID group is low.

• Assign primary and secondary volumes and data pools to a RAID group consisting of SAS drives, SAS7.2K drives, SSD drives, or SAS (SED) drives to achieve best possible performance. SATA drives can be used, however.

• When cascading TrueCopy and SnapShot pairs, assign a volume of the SAS drives, the SAS7.2K drives, the SSD drives, or the SAS (SED) drives and assign four or more disks to a data pool.

• Assign an LU consisting of four or more data disks, otherwise host and copying performance may be lowered.

• Limit the I/O load on both local and remote arrays to maximize performance. Performance on each array also affects performance on the other array, as well as data pool capacity and the synchronization of volumes. Therefore, it is best to assign a volume of SAS drives, SAS7.2K drives, SSD drives, or SAS (SED) drives and assign four or more disks (which have higher performance) than SATA drives, to a data pool.

Plan and design—arrays, volumes and operating systems 4–3

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 60

Operating system recommendations and restrictions

The following sections provide operating system recommendations and restrictions.

Host time-out

I/O time-out from the host to the array should be more than 60 seconds. You can figure I/O time-out by increasing the remote path time limit times

6. For example, if the remote path time-out value is 27 seconds, set host I/ O time-out to 162 seconds (27 x 6) or more.

P-VOL, S-VOL recognition by same host on VxVM, AIX®, LVM

VxVM, AIX®, and LVM do not operate properly when both the P-VOL and SVOL are set up to be recognized by the same host. The P-VOL should be recognized one host on these platforms, and the S-VOL recognized by a different host.

HP server

When MC/Service Guard is used on a HP server, connect the host group (Fibre Channel) or the iSCSI Target to HP server as follows:

For Fibre Channel interfaces

1. In the Navigator 2 GUI, access the array and click Host Groups in the

Groups tree view. The Host Groups screen displays.

2. Click the check box for the Host Group that you want to connect to the

HP server.

WARNING! Your host group changes will be applied to multiple ports. This

change will delete existing host group mappings and corresponding Host Group IDs, corrupting or removing data associated with the host groups. To keep specified host groups you do not want to remove, please cancel this operation and make changes to only one host group at a time.

4–4 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 61

3. Click Edit Host Group.

•

Plan and design—arrays, volumes and operating systems 4–5

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 62

The Edit Host Group screen appears.

4. Select the Options tab.

5. From the Platform drop-down list, select HP-UX. Doing this causes

Enable HP-UX Mode, Enable PSUE Read Reject Mode, and Enable PSUE Read Reject Mode to be selected in the Additional Setting box.

6. Click OK. A message appears, click Close.

4–6 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 63

For iSCSI interfaces

1. In the Navigator 2 GUI, access the array and click iSCSI Targets in the

Groups tree view. The iSCSI Targets screen displays.

2. Click the check box for the iSCSI Targets that you want to connect to the HP server.

3. Click Edit Target. The Edit iSCSI Target screen appears.

4. Select the Options tab.

5. From the Platform drop-down list, select HP-UX. Doing this causes

“Enable HP-UX Mode” and “Enable PSUE Read Reject Mode” to be selected in the Additional Setting box.

6. Click OK. A message appears, click Close.

Windows Server 2000

• A P-VOL and S-VOL cannot be made into a dynamic disk on Windows

Server 2000 and Windows Server™ 2008.

• Native OS mount/dismount commands can be used for all platforms, except Windows Server 2000. The native commands on this environment do not guarantee that all data buffers are completely flushed to the volume when dismounting. In these instances, you must use CCI to perform volume mount/unmount operations. For more

information on the CCI mount/unmount commands, see the Hitachi AMS Command Control Interface (CCI) Reference Guide.

Windows Server 2003/2008

• A P-VOL and S-VOL can be made into a dynamic disk on Windows

Server 2003.

• In Windows Server™ 2008, refer to the Hitachi Adaptable Modular Storage Command Control Interface (CCI) Reference Guide for the restrictions when the mount/unmount command is used.

• Windows resynchronized while retaining data on the S-VOL on the server memory, the compatible backup cannot be collected. Therefore, execute the CCI sync command immediately before re-synchronizing the pair for the un-mounted S-VOL.

• In Windows Server™ 2008, set only the P-VOL of TCE to be recognized by the host and let another host recognize the S-VOL.

• (CCI only) If a path detachment is caused by controller detachment or Fibre Channel failure, and the detachment continues for longer than one minute, the command device may not be recognized when recovery occurs. In this case, execute the “re-scanning of the disks” in Windows. If Windows cannot access the command device, though CCI recognizes the command device, restart CCI.

• Volumes to be recognized by the same host: If you recognize the P-VOL and S-VOL on Windows Server 2008 at the same time, it may cause an error because the P-VOL and S-VOL have the same disk signature. When the P-VOL and S-VOL have the same data, split the pair and then

may write for the un-mounted volume. If a pair is

Plan and design—arrays, volumes and operating systems 4–7

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 64

rewrite the disk signature so that they can retain different disk signatures. You can use the uniqueid command to rewrite a disk

signature. See the Hitachi Adaptable Modular Storage Command Control Interface (CCI) User's Guide for details.

Identifying P-VOL and S-VOL LUs on Windows

In Navigator 2, the P-VOL and S-VOL are identified by their LU number. In Windows 2003 Server, LUs are identified by HLUN. To map LUN to HLUN on Windows, proceed as follows. These instructions provide procedures for iSCSI and Fibre Channel interfaces.

1. Identify the HLUN of your Windows disk. a. From the Windows Server 2003 Control Panel, select Computer

Management>Disk Administrator.

b. Right-click the disk whose HLUN you want to know, then select

Properties. The number displayed to the right of “LUN” in the dialog

window is the HLUN.

2. Identify HLUN-to-LUN Mapping for the iSCSI interface as follows. (If using Fibre Channel, skip to Step 3.)

a. In the Navigator 2 GUI, select the desired array.

b. In the array tree that displays, click the Group icon, then click the

iSCSI Targets icon in the Groups tree.

c. On the iSCSI Target screen, select an iSCSI target.

d. On the target screen, select the Logical Units tab. Find the

identified HLUN. The LUN displays in the next column.

e. If the HLUN is not present on a target screen, on the iSCSI Target

screen, select another iSCSI target and repeat Step 2d.

3. Identify HLUN-to-LUN Mapping for the Fibre Channel interface, as follows:

a. In Navigator 2, select the desired array.

b. In the array tree that displays, click the Groups icon, then click the

Host Groups icon in the Groups tree.

WARNING! Your host group changes will be applied to multiple ports. This

c. On the Host Groups screen, select a Host group.

d. On the host group screen, select the Logical Units tab. Find the

identified H-LUN. The LUN displays in the next column.

e. If the HLUN is not present on a host group target screen, on the Host

Groups screen, select another Host group and repeat Step 3d.

4–8 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 65

Windows 2000 or Windows Server and TCE Configuration

• Volume mount:

In order to make a consistent backup using a storage-based replication such as TCE, you must have a way to flush the data residing on the server memory to the array, so that the source volume of the replication has the complete data.

You can flush the date on the server memory using the umount command of CCI to unmount the volume. When using the umount command of CCI for unmount, use the mount command of CCI for mount.

When using Window® 2000, do not use the mountvol command attached to Windows® 2000 as standard and use the mount/umount command of CCI even if you are using GUI or CLI of Navigator 2 for the pair operation.

If you are using Windows Server™ 2003, mountvol /P to flush data on the server memory when un-mounting the volume is supported. Please understand the specification of the command and run sufficient test before you use it for your operation.

In Windows Server™ 2008, refer to the Hitachi Adaptable Modular Storage Command Control Interface (CCI) Reference Guide for the restrictions when mount/unmount command is used.

Windows® may write for the un-mounted volume. If a pair is resynchronized while remaining the data to the S-VOL on the memory of the server, the compatible backup cannot be collected. Therefore, execute the sync command of CCI immediately before re-synchronizing the pair for the un-mounted S-VOL.

For more detail about the CCI commands, see the Hitachi Adaptable Modular Storage Command Control Interface (CCI) Reference Guide.

• Volumes to be recognized by the same host

If you recognize the P-VOL and S-VOL on Windows Server™ 2008 at the same time, it may cause an error because the P-VOL and S-VOL have the same disk signature. When the P-VOL and S-VOL have the same data, split the pair and then rewrite the disk signature so that they can retain different disk signatures. You can use the uniqueid command to rewrite a disk signature. See the Hitachi Adaptable Modular Storage Command Control Interface (CCI) User's Guide for the detail

• Command devices:

When a remote path detachment, which is caused by a controller detachment or Fibre channel failure, continues for longer than one minute, the command device may be unable to be recognized when recovery from the remote path detachment is made. To make the recovery, execute the "re-scanning of the disks" of Windows®. When Windows® cannot access the command device, although CCI becomes able to recognize the command device, restart CCI.

Plan and design—arrays, volumes and operating systems 4–9

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 66

Dynamic Disk in Windows 2000/Windows Server

n an environment of the Windows Server 2000/Windows Server, you cannot use TCE pair volumes as dynamic disk. The reason for this restriction is because in this case if you restart Windows or use the Rescan Disks command after creating or re-synchronizing a TCE pair, there are cases where the S-VOL is displayed as Foreign in Disk Management and become inaccessable.

VMWare and TCE Configuration

When creating a backup of the virtual disk in the vmfs format using TCE, shutdown the virtual machine that accesses the virtual disk, and then split the pair.

If one LU is shared by multiple virtual machines, shutdown all the virtual machines that share the LU when creating a backup. It is not recommended to share one LU by multiple virtual machines in the configuration that creates a backup using TCE.

4–10 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 67

Concurrent Use of Dynamic Provisioning

• When the array firmware version is less than 0893/A, the DP-VOLs created by Dynamic Provisioning cannot be set for a P-VOL or an S-VOL of TCE. Moreover, when the array firmware version is less than 0893/A, the DP-VOLs cannot be added to the data pool used by SnapShot and TCE.

• Depending on the installed cache memory, Dynamic Provisioning and TCE may not be unlocked at the same time. To unlock Dynamic Provisioning and TCE at the same time, add cache memories. For the capacity of the supported cache memory, refer to User Data Area of

Cache Memory on page 4-16.

• The data pool used by SnapShot and TCE cannot be used as a DP pool of Dynamic Provisioning. Moreover, the DP pool used by Dynamic Provisioning cannot be used as data pools of SnapShot and TCE.

When the array firmware version is 0893/A or more, the DP-VOLs created by Dynamic Provisioning can be set for a P-VOL, an S-VOL, or a data pool of TCE. H o w eve r, th e n ormal LU and the DP-VOL cannot coexi s t i n the same data pool.

The points to keep in mind when using TCE and Dynamic Provisioning together are described here. Refer to the Hitachi Adaptable Modular Storage Dynamic Provisioning User's Guide for detailed information about Dynamic Provisioning. Hereinafter, the LU created in the RAID group is called a normal LU and the LU created in the DP pool that is created by Dynamic Provisioning is called a DP-VOL.

• When using a DP-VOL as a DMLU

Check that the free capacity (formatted) of the DP pool to which the DP-VOL belongs is 10 GB or more, and then set the DP-VOL as a DMLU. If the free capacity of the DP pool is less than 10 GB, the DP-VOL cannot be set as a DMLU.

• LU type that can be set for a P-VOL, an S-VOL, or a data pool of TCE

The DP-VOL created by Dynamic Provisioning can be used for a P-VOL, an S-VOL, or a data pool of TCE. Table 4-1 and Table 4-2 show a combination of a DP-VOL and a normal LU that can be used for a P-VOL, an S-VOL, or a data pool of TCE.

Table 4-1: Combination of a DP-VOL and a Normal LU

TCE P-VOL TCE S-VOL Contents

DP-VOL DP-VOL Available. The P-VOL and S-VOL capacity can be

reduced compared to the normal LU.

DP-VOL Normal LU Available.

Normal LU DP-VOL Available. When the pair status is Split, the S-VOL

capacity can be reduced compared to the normal LU by deleting 0 data.

Plan and design—arrays, volumes and operating systems 4–11

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 68

Table 4-2: Combination of a DP-VOL for Data Pool and a Normal

P-VOL Data

Pool

DP-VOL DP-VOL Available. The data pool consumed capacity can

be reduced compared to the normal LU on both local side and remote side.

DP-VOL Normal LU Available. The data pool consumed capacity can

be reduced compared to the normal LU on local side.

Normal LU DP-VOL Available. The data pool consumed capacity can

be reduced compared to the normal LU on remote side.

When creating a TCE pair using the DP-VOLs, in the P-VOL, the S-VOL or the data pool specified at the time of the TCE pair creation, the DPVOLs whose Full Capacity Mode is enabled and disabled cannot be mixed.

When creating the data pool using multiple DP-VOLs, the data pool cannot be created by combining the DP-VOLs that have different setting of Enabled/Disabled for Full Capacity Mode.

• Assigning the controlled processor core of a P-VOL and a data pool that uses the DP-VOL

When the controlled processor core of the DP-VOL used for a P-VOL (SVOL) or used for a data pool of TCE differs as well as the normal LU, switch the P-VOL (S-VOL) controlled processor core assignment to the data pool controlled processor core automatically and create a pair. (In case of AMS2500)

• DP pool designation of a P-VOL (S-VOL) and a data pool which uses the DP-VOL

When using the DP-VOL created by Dynamic Provisioning for a P-VOL (SVOL) or a data pool of TCE, using the DP-VOL designated in separate DP pool of a P-VOL (S-VOL) and a data pool is recommended considering the performance.

• Setting the capacity when placing the DP-VOL in the data pool

When the pair status is Split, the old data is copied to the data pool while writing to the P-VOL. When using the DP-VOL created by Dynamic Provisioning as the data pool of TCE, the consumed capacity of the DPVOL in the data pool is increased by storing the old data in the data pool. If the DP-VOL of more than or equal to the DP pool capacity is created and used for the data pool, this processing may deplete the DP pool capacity. When using the DP-VOL for the data pool of TCE, it is recommended to set the capacity making the over provisioning ratio 100% or less so that the DP pool capacity does not deplete.

4–12 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 69

Furthermore, the threshold value of the data pool of TCE and the threshold value of the DP pool differ. Even if the data pool use rate of TCE shows 10% or less, the DP pool consumed capacity may have exceeded Depletion Alert. Check whether the actual use rate falls below the respective threshold values of the data pool and DP pool of TCE.

• Pair status at the time of DP pool capacity depletion

When the DP pool is depleted after operating the TCE pair that uses the DP-VOL created by Dynamic Provisioning, the pair status of the pair concerned may be a Failure. Table 4-3 on page 4-13 shows the pair statuses before and after the DP pool capacity depletion. When the pair status becomes a Failure caused by the DP pool capacity depletion, add the DP pool capacity whose capacity is depleted, and execute the pair operation again.

Table 4-3: Pair Statuses before the DP Pool Capacity Depletion and

Pair Statuses after the DP Pool Capacity Depletion

Pair Statuses after the

Pair Statuses before the DP

Pool Capacity Depletion

Simplex Simplex Simplex

Reverse Synchronizing Failure Reverse Synchronizing

Paired Paired

Split Split Split

Failure Failure Failure

DP Pool Capacity

Depletion belonging to

P-VOL

Failure (See Note)

Pair Statuses after the DP

Pool Capacity Depletion

belonging to data pool

Failure (See Note)

Failure

NOTE: When write is performed to the P-VOL to which the capacity

depletion DP pool belongs, the copy cannot be continued and the pair status becomes a Failure.

• DP pool status and availability of pair operation

When using the DP-VOL created by Dynamic Provisioning for a P-VOL (SVOL) or a data pool of the TCE pair, the pair operation may not be executed depending on the status of the DP pool to which the DP-VOL belongs. Table 4-4 on page 4-14 and Table 4-5 on page 4-14 show the DP pool status and availability of the TCE pair operation. When the pair operation fails due to the DP pool status, correct the DP pool status and execute the pair operation again.

Plan and design—arrays, volumes and operating systems 4–13

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 70

Table 4-4: DP Pool for P-VOL Statuses and Availability of Pair Operation

DP Pool Statuses, DP Pool Capacity Statuses, and DP Pool Optimization

Statuses

Pair Operation

Normal

Create pair 0 (1) 0x (1) 0x0

Split pair000000 Resync pair 0 (1) 0x (1) 0x0 Swap pair 0 (1) 0x (1) 000

Delete pair000000

Capacity in

Growth

Capacity

Depletion

Regressed Blocked

DP in

Optimization

Table 4-5: DP Pool for S-VOL Statuses and Availability of Pair Operation

Pair Operation

Normal

Create pair 0 (2) 0x (2) 0x0

Split pair000000 Resync pair 0 (2) 0x (2) 000 Swap pair 0 (2) 00 (2) 0x0

Delete pair000000

Capacity in

Growth

Capacity

Depletion

Regressed Blocked

DP in

Optimization

(1) Refer to the status of the DP pool to which the DP-VOL of the S-VOL

belongs. If the status exceeds the DP pool capacity belonging to the S-VOL by the pair operation, the pair operation cannot be performed.

(2) Refer to the status of the DP pool to which the DP-VOL of the P-VOL

belongs. If the status exceeds the DP pool capacity belonging to the P-VOL by the pair operation, the pair operation cannot be performed.

(3) When the DP pool was created or the capacity was added, the

formatting operates for the DP pool. If pair creation, pair resynchronization, or swapping is performed during the formatting, depletion of the usable capacity may occur. Since the formatting progress is displayed when checking the DP pool status, check if the sufficient usable capacity is secured according to the formatting progress, and then start the operation.

• Operation of the DP-VOL during TCE use

When using the DP-VOL created by Dynamic Provisioning for a P-VOL, an S-VOL, or a data pool of TCE, any of the operations among the capacity growing, capacity shrinking, LU deletion, and Full Capacity

4–14 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 71

Mode changing of the DP-VOL in use cannot be executed. To execute the operation, delete the TCE pair of which the DP-VOL to be operated is in use, and then perform it again.

• Operation of the DP pool during TCE use

When using the DP-VOL created by Dynamic Provisioning for a P-VOL, an S-VOL, or a data pool of TCE, the DP pool to which the DP-VOL in use belongs cannot be deleted. To execute the operation, delete the TCE pair of which the DP-VOL is in use belonging to the DP pool to be operated, and then execute it again. The attribute edit and capacity addition of the DP pool can be executed usually regardless of the TCE pair.

• Availability of TCE pair creation between different firmware versions

When the firmware versions of the array differ on the local side and the remote side, if the firmware version of the array including the DP-VOL is 0893/A or more, a TCE pair can be created (Figure 4-1). Table 4-6 shows pair creation availability when the firmware version of the array is 0893/A or more or less than 0893/A (including AMS500/1000).

•

Figure 4-1: Availability of TCE Pair Creation between Different

Firmware Versions

Plan and design—arrays, volumes and operating systems 4–15

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 72

Table 4-6: Availability of TCE Pair Creation between Different Firmware

Versions

P-VOL S-VOL

Normal LU Normal LU 0 0 0

Normal LU DP-VOL 0 x 0

DP-VOLNormal LU00x

DP-VOL DP-VOL 0 x x

Local: 0893/A or

Remote: 0893/A

or more

Local: 0893/A or

Remote: Less than

0893

Local: Less than

0893/A

Remote: 0893/A

or more

• Cascade connection

A cascade can be performed on the same conditions as the normal LU. However, the firmware version of the array including the DP-VOL needs to be 0893/A or more.

User Data Area of Cache Memory

If TCE is used to secure a part of the cache memory, the user data area of the cache memory decreases. Also, by using SnapShot and Dynamic Provisioning together, the user data area may further decrease. Table 4-7 through Table 4-13 on page 4-20 show the cache memory secured capacity and the user data area when using the program product. For Dynamic Provisioning, the user data area differs depending on DP Capacity Mode.

Refer to the Hitachi Adaptable Modular Storage Dynamic Provisioning User's Guide for detailed information.

Table 4-7: Supported Capacity of the Regular Capacity Mode (H/

W Rev. is 0100)

Array Type Cache Memory

AMS2100 1 GB/CTL 80 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

AMS2300 1 GB/CTL 140 mb -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

8 GB/CTL 4 GB

(1 of 2)

Management

Capacity for

Dynamic

Provisioning

Capacity

Secured for

SnapShot or TCE

4–16 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 73

Table 4-7: Supported Capacity of the Regular Capacity Mode (H/

W Rev. is 0100)

(1 of 2)

Management

Array Type Cache Memory

Capacity for

Dynamic

Provisioning

AMS2500 2 GB/CTL 300 MB 512 MB

4 GB/CTL 1.5 GB

6 GB/CTL 3 GB

8 GB/CTL 4 GB

10 GB/CTL 5 GB

12 GB/CTL 6 GB

16 GB/CTL 8 GB

Capacity

Secured for

SnapShot or TCE

Table 4-8: Supported Capacity of the Regular Capacity Mode (H/

Array Type

(the H/W

Rev. is

0100)

W Rev. is 0100)

Capacity

Secured for

Dynamic

Provisioning

and TCE or

SnapShot

User Data

Area when

Dynamic

Provisioning,

TCE, and

SnapShot are

Disabled

(2 of 2)

User Data

Area when

Provisioning

Using

Dynamic

User Data

Area when

Using

Dynamic

Provisioning

and TCE or

SnapShot

AMS2100 - 590 MB 590 MB N/A

580 MB 1,520 MB 1,440 MB 940 MB

2,120 MB 3,520 MB 3,460 MB 1,400 MB

AMS2300 - 500 MB 500 MB N/A

660 MB 1,440 MB 1,300 MB 780 MB

2,200 MB 3,280 MB 3,120 MB 1,080 MB

4,240 MB 7,160 MB 7,020 MB 2,920 MB

AMS2500 800 MB 1,150 MB 850 MB N/A

1,830 MB 2,960 MB 2,660 MB 1,130 MB

3,360 MB 4,840 MB 4,560 MB 1,480 MB

4,400 MB 6,740 MB 6,440 MB 2,340 MB

5,420 MB 8,620 MB 8,320 MB 3,200 MB

6,440 MB 10,500 MB 10,200 MB 4,060 MB

8,480 MB 14,420 MB 14,120 MB 5,940 MB

Plan and design—arrays, volumes and operating systems 4–17

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 74

Table 4-9: Supported Capacity of the Maximum Capacity Mode

(H/W Rev. is 0100) (1 of 2)

Management

Array Type Cache Memory

Capacity for

Dynamic

Provisioning

AMS2100 1 GB/CTL 210 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

AMS2300 1 GB/CTL 310 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

8 GB/CTL 4 GB

AMS2500 2 GB/CTL 520 MB 512 MB

4 GB/CTL 1.5 GB

6 GB/CTL 3 GB

8 GB/CTL 4 GB

10 GB/CTL 5 GB

12 GB/CTL 6 GB

16 GB/CTL 8 GB

Capacity

Secured for

SnapShot or TCE

Table 4-10: Supported Capacity of the Maximum Capacity Mode

(H/W Rev. is 0100) (2 of 2)

Capacity

Secured for

Array

Type

Dynamic

Provisioning

and TCE or

SnapShot

AMS2100 - 590 MB N/A N/A

710 MB 1,520 MB 1,310 MB 810 MB

2,270 MB 3,520 MB 3,310 MB 1,250 MB

AMS2300 - 500 MB N/A N/A

830 MB 1,440 MB 1,130 MB 610 MB

2,350 MB 3,280 MB 2,970 MB 930 MB

4,410 MB 7,160 MB 6,850 MB 2,750 MB

User Data

Area when

Dynamic

Provisioning,

TCE, and

SnapShot are

Disabled

User Data

Area when

Using

Dynamic

Provisioning

User Data

Area when

Using

Dynamic

Provisioning

and TCE or

SnapShot

4–18 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 75

Table 4-10: Supported Capacity of the Maximum Capacity Mode

(H/W Rev. is 0100) (2 of 2)

Capacity

Secured for

Array

Type

Dynamic

Provisioning

and TCE or

SnapShot

AMS2500 1,022 MB 1,150 MB N/A N/A

2,208 MB 2,960 MB N/A N/A

3,600 MB 4,840 MB 4,320 MB 1,240 MB

4,620 MB 6,740 MB 6,220 MB 2,120 MB

5,640 MB 8,620 MB 8,100 MB 2,980 MB

6,660 MB 10,500 MB 9,980 MB 3,840 MB

8,700 MB 14,420 MB 13,900 MB 5,720 MB

User Data

Area when

Dynamic

Provisioning,

TCE, and

SnapShot are

Disabled

User Data

Area when

Using

Dynamic

Provisioning

User Data

Area when

Using

Dynamic

Provisioning

and TCE or

SnapShot

Table 4-11: Supported Capacity of the Regular Capacity Mode

H/W Rev. is 0200) (1 of 2)

(

Management

Array Type Cache Memory

Capacity for

Dynamic

Provisioning

AMS2100 1 GB/CTL 80 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

AMS2300 1 GB/CTL 140 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

8 GB/CTL 4 GB

AMS2500 2 GB/CTL 300 MB 512 MB

4 GB/CTL 1.5 GB

6 GB/CTL 3 GB

8 GB/CTL 4 GB

10 GB/CTL 5 GB

12 GB/CTL 6 GB

16 GB/CTL 8 GB

Capacity

Secured for

SnapShot or TCE

Plan and design—arrays, volumes and operating systems 4–19

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 76

Table 4-12: Supported Capacity of the Regular Capacity Mode

H/W Rev. is 0200) (2 of 2)

(

Capacity

Secured for

Dynamic

Provisioning

and TCE or

SnapShot

AMS2100 - 590 MB 590 MB N/A

580 MB 1,390 MB 1,310 MB 810 MB

2,120 MB 3,360 MB 3,280 MB 1,220 MB

AMS2300 - 500 MB 500 MB N/A

660 MB 1,340 MB 1,200 MB 680 MB

2,200 MB 3,110 MB 2,970 MB 930 MB

4,240 MB 6,940 MB 6,800 MB 2,700 MB

AMS2500 800 MB 1,150 MB 850 MB N/A

1,830 MB 2,780 MB 2,480 MB 950 MB

3,360 MB 4,660 MB 4,360 MB 1,280 MB

4,400 MB 6,440 MB 6,140 MB 2,040 MB

5,420 MB 8,320 MB 8,020 MB 2,900 MB

6,440 MB 9,980 MB 9,680 MB 3,540 MB

8,480 MB 14,060 MB 13,760 MB 5,580 MB

User Data

Area when

Dynamic

Provisioning,

TCE, and

SnapShot are

Disabled

User Data

Area when

Using

Dynamic

Provisioning

User Data

Area when

Provisioning

and TCE or

SnapShot

Using

Dynamic

Table 4-13: Supported Capacity of the Maximum Capacity Mode

(H/W Rev. is 0200) (1 of 2)

Management

Array Type Cache Memory

Capacity for

Dynamic

Provisioning

AMS2100 1 GB/CTL 210 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

AMS2300 1 GB/CTL 310 MB -

2 GB/CTL 512 MB

4 GB/CTL 2 GB

8 GB/CTL 4 GB

4–20 Plan and design—arrays, volumes and operating systems

Capacity

Secured for

SnapShot or TCE

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 77

Table 4-13: Supported Capacity of the Maximum Capacity Mode

(H/W Rev. is 0200) (1 of 2)

Management

Array Type Cache Memory

Capacity for

Dynamic

Provisioning

AMS2500 2 GB/CTL 520 MB 512 MB

4 GB/CTL 1.5 GB

6 GB/CTL 3 GB

8 GB/CTL 4 GB

10 GB/CTL 5 GB

12 GB/CTL 6 GB

16 GB/CTL 8 GB

Capacity

Secured for

SnapShot or TCE

Table 4-14: Supported Capacity of the Maximum Capacity Mode

(H/W Rev. is 0200) (2 of 2)

User Data

Area when

Using

Dynamic

Provisioning

and TCE or

SnapShot

Array

Type

Capacity

Secured for

Dynamic

Provisioning

and TCE or

SnapShot

User Data

Area when

Dynamic

Provisioning,

TCE, and

SnapShot are

Disabled

User Data

Area when

Using

Dynamic

Provisioning

AMS2100 - 590 MB N/A N/A

710 MB 1,390 MB 1,180 MB 680 MB

2,270 MB 3,360 MB 3,150 MB 1,090 MB

AMS2300 - 500 MB N/A N/A

830 MB 1,340 MB 1,030 MB 510 MB

2,350 MB 3,110 MB 2,800 MB 760 MB

4,410 MB 6,940 MB 6,630 MB 2,530 MB

AMS2500 1,022 MB 1,090 MB N/A N/A

2,078 MB 2,780 MB N/A N/A

3,600 MB 4,660 MB 4,140 MB 1,060 MB

4,620 MB 6,440 MB 5,920 MB 1,820 MB

5,640 MB 8,320 MB 7,800 MB 2,680 MB

6,660 MB 9,980 MB 9,460 MB 3,320 MB

8,700 MB 14,060 MB 13,540 MB 5,360 MB

Plan and design—arrays, volumes and operating systems 4–21

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 78

Formatting the DMLU in the Event of a Drive Failure

When the DMLU is in a RAID group or DP pool with RAID5 or RAID6 and a drive failure occurs on the RAID group or DP pool with no redundancy, the data in the DMLU will be incomplete and unusable.

At that time, for the firmware version of 08C3/F and later, the DMLU will automatically become unformatted, so make sure to format the DMLU.

For less than 08C3/F, even though the DMLU will not automatically become unformatted, make sure to format the DMLU.

It is possible to format a DMLU without having to release the DMLU.

4–22 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 79

Maximum supported capacity

The capacity you can assign to replication volumes per controller is limited, for the following reasons:

• The TCE P-VOL and S-VOL, and the SnapShot P-VOL if used, share common data pool resources. Therefore, data pool capacity is limited.

• The maximum capacity supported by a TCE pair depends on the P-VOL capacity of SnapShot (if used), data pool capacity, and cache memory capacity.

• When using other copy systems and TCE together, the maximum supported capacity of the P-VOL may be restricted further.

In addition to this, capacity is managed by the AMS array in blocks of 15.75 KB for data volumes and 3.2 KB for data pools. For example, when a P-VOL block’s actual size is 16 KB, the array manages it as two blocks of 15.75 KB, or 31.5 KB. Data pool capacity is managed in the same way but at 3.2 KB per block.

This section provides formulas for calculating your existing or planned TCE volume capacity and comparing it to the maximum supported capacity for your particular controller and its cache memory size.

TCE capacity must be calculated for both of the following:

1. The ratio of TCE and SnapShot (if used) capacity to data pool capacity. Capacity is calculated using the following volumes:

- TCE P-VOLs and S-VOLs

- SnapShot P-VOLs (if used)

- All data pools

2. Concurrent use of TCE and ShadowImage. If SnapShot is used concurrently also, it is included in this calculation. Capacity is calculated using the following volumes:

- TCE P-VOLs

- SnapShot P-VOLs

- ShadowImage S-VOLs

•

NOTE: When SnapShot is enabled, a portion of cache memory is assigned

to it for internal operations. Hitachi recommends that you review the

appendix on SnapShot and Cache Partition Manager in the Hitachi AMS 2000 Family Copy-on-Write SnapShot User Guide.

Plan and design—arrays, volumes and operating systems 4–23

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 80

TCE and SnapShot capacity

Because capacity is managed by the array in blocks of 15.75 KB for of data volumes and 3.2 KB for data pools, the capacity of your array’s TCE and SnapShot volumes must be specially calculated.

All formulas, tables, graphs and examples pertain to one controller. On dual controller arrays, you must perform calculations for both controllers.

Managed capacity is calculated here, per controller, using the following formula:

Size of all TCE P-VOLs + Size of all TCE S-VOLs + Size all SnapShot P-VOLs (if used) / 5 + size of all data pool volumes < Max. Sup. Capacity

Maximum supported capacity is shown in Table 4-15.

•

Table 4-15: Maximum Supported Capacities, per

Cache Memory

per Controller

(TCE P-VOLs and S-VOLs, SnapShot P-VOLs, Data

Controller Cache Size

Maximum Capacity

Pools)

AMS2100 AMS2300 AMS2500

2 GB per CTL 1.4 TB Not supported Not supported

4 GB per CTL 6.2 TB 6.2 TB Not supported

8 GB per CTL Not supported 12.0 TB 12.0 TB

16 GB per CTL Not supported Not supported 24.0 TB

•

NOTE: In a dual-controller array, the calculations must be performed for

both controllers.

Example:

In this example, the array and cache memory per controller is AMS 2300/4 GB.

1. List the size of each TCE P-VOL and S-VOL on the array, and of each SnapShot P-VOL (if present) in the array. For example:

TCE P-VOL 1 = 100 GB TCE S-VOL 1 = 100 GB SnapShot P-VOL 1 = 50 GB

4–24 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 81

2. Calculate managed P-VOL and S-VOL capacity, using the formula:

ROUNDUP (P-VOL/S-VOL / 15.75) * 15.75

For example:

TCE P-VOL1: ROUNDUP (100 / 15.75) = 7 7 * 15.75 = 110.25 GB, the managed P-VOL Capacity TCE S-VOL1: ROUNDUP (100 / 15.75) = 7 7 * 15.75 = 110.25 GB, the managed S-VOL Capacity SnapShot P-VOL1: ROUNDUP (50 / 15.75) = 4 4 * 15.75 = 63 GB, the managed P-VOL Capacity

3. Add the total managed capacity of P-VOLs and S-VOLs. For example:

Total TCE P-VOL and S-VOL managed capacity = 221 GB Total SnapShot P-VOL capacity = 63 GB 221 GB + 63 GB = 284 GB

4. For each P-VOL and S-VOL, list the data pools and their sizes. For example:

TCE P-VOL1 has 1 data pool whose capacity = 70 GB TCE S-VOL1 has 1 data pool whose capacity = 70 GB SnapShot P-VOL1 has 1 data pool whose capacity = 30 GB

5. Calculate managed data pool capacity, using the formula:

ROUNDUP (data pool capacity / 3.2) * 3.2

For example:

TCE P-VOL 1 data pool: ROUNDUP (70 / 3.2 = 22) * 3.2 = 71 GB TCE S-VOL 1 data pool: ROUNDUP (70 / 3.2 = 22) * 3.2 = 71 GB SS P-VOL 1 data pool: ROUNDUP (30 / 3.2 = 10) * 3.2 = 32 GB

6. Add total data pool managed capacity. For example:

71 GB + 71 GB + 32 GB = 175 GB

7. Calculate total managed capacity using the following equation:

ROUNDUP Total TCE/SnapShot managed capacity / 5 + total data pool managed capacity < maximum supported capacity

For example:

Divide the total TCE/SnapShot capacity by 5. 284 GB / 5 = 57 GB

8. Add the quotient to data pool managed capacity. For example:

57 GB + 176 GB = 234 GB

9. Compare managed capacity to maximum supported capacity for the 4 GB cache per controller for AMS 2300, which is 6.2 TB. The managed capacity is well below maximum supported capacity.

Table 4-20 on page 4-31 through Table 4-23 on page 4-32 show how

closely capacity between data volumes and data pool volumes must be managed. These tables are provided for your information. Also, Figure 4-4

on page 4-32 shows a graph of how data volume-to-data pool volume

relates to maximum supported capacity.

Plan and design—arrays, volumes and operating systems 4–25

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 82

TCE, SnapShot, ShadowImage concurrent capacity

If ShadowImage is used on the same controller as TCE, capacity for concurrent use must also be calculated and compared to maximum supported capacity. If SnapShot is used also, it is included in concurrent-use calculations.

Concurrent-use capacity is calculated using the following formula:

Maximum TCE supported capacity of P-VOL and S-VOL (TB) = TCE maximum single capacity

- (Total ShadowImage S-VOL capacity / 51)

- (Total SnapShot P-VOL capacity / 3)

TCE maximum single capacity is shown in Table 4-16.

•

Table 4-16: TCE Maximum Single Capacity per Controller

Equipment Type

AMS2100 1 GB per CTL –

AMS2300 1 GB per CTL –

AMS2500 2 GB per CTL 10

Mounted Memory

Capacity

2 GB per CTL 15

4 GB per CTL 18

2 GB per CTL 14

4 GB per CTL 38

8 GB per CTL 77

4 GB per CTL 38

6 GB per CTL 54

8 GB per CTL 70

10 GB per CTL 93

12 GB per CTL 116

16 GB per CTL 140

Single Maximum Capacity

(TB)

Example

In this example, the array and cache memory per controller is AMS2100 and 2 GB per CTL.

Maximum TCE supported capacity of P-VOL and S-VOL (TB) = TCE maximum single capacity

- (Total ShadowImage S-VOL capacity / 51)

- (Total SnapShot P-VOL capacity / 3)

1. TCE Maximum single capacity = 15 TB

4–26 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 83

2. Calculate ShadowImage S-VOL managed capacity (ROUNDUP (S-VOL /

15.75) * 15.75). Then divide by 51. For example: All ShadowImage S-VOLs: ROUNDUP 4000 GB / 15.75 = 254 GB 254 * 15.75 = 4001 GB, the managed S-VOL Capacity 4001 GB / 51 = 79

3. Subtract the quotient from the TCE maximum single capacity. For example:

15 TB (15,000 GB) - 79 GB = 14,921 GB

4. Calculate SnapShot P-VOL managed capacity, then divide by 3. For example:

All SnapShot P-VOLs: ROUNDUP 800 GB / 15.75 = 51 GB 51 GB * 15.75 = 803 GB 803 GB / 3 = 268 GB

5. Subtract the quotient from the remaining TCE maximum single capacity. For example:

14,921 GB - 268 GB = 14,653 GB, the capacity left for TCE P-VOLs and S-VOLs on the controller.

If your system’s managed capacity exceeds the maximum supported capacity, you can do one or more of the following:

• Change the P-VOL size

• Reduce the number of P-VOLs

• Change the data pool size

• Reduce ShapShot and ShadowImage P-VOL/S-VOL size

Plan and design—arrays, volumes and operating systems 4–27

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 84

Maximum Supported Capacity of P-VOL and Data Pool

Table 4-17 to Table 4-19 show the maximum supported capacities of the P-

VOL and the data pool for each cache memory capacity, and the formula for calculating the previous capacities.

Table 4-17: Formula for Calculating Maximum Supported

Capacity Value for P-VOL/Data Pool (AMS2100)

Capacity of Cache Memory

Installed

1 GB/CTL Not supported.

2 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

4 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

Capacity Spared for the Differential

Data (Shared by SnapShot and TCE)

of TCE capacity ÷ 5 + Total data pool capacity < 1.4 TB

of TCE capacity ÷ 5 + Total data pool capacity < 6.2 TB

Table 4-18: Formula for Calculating Maximum Supported

Capacity Value for P-VOL/Data Pool (AMS2300)

Capacity of Cache Memory

Installed

1 GB/CTL Not supported.

2 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

4 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

8 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

Capacity Spared for the Differential

Data (Shared by SnapShot and TCE)

of TCE capacity ÷ 5 + Total data pool capacity < 1.4 TB

of TCE capacity ÷ 5 + Total data pool capacity < 6.2 TB

of TCE capacity ÷ 5 + Total data pool capacity < 12.0 TB

Table 4-19: Formula for Calculating Maximum Supported

Capacity Value for P-VOL/Data Pool (AMS2500)

Capacity of Cache Memory

Installed

2 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

4 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

6 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

Capacity Spared for the Differential

Data (Shared by SnapShot and TCE)

of TCE capacity ÷ 5 + Total data pool capacity < 1.4 TB

of TCE capacity ÷ 5 + Total data pool capacity < 4.7 TB

of TCE capacity ÷ 5 + Total data pool capacity < 9.4 TB

4–28 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 85

Table 4-19: Formula for Calculating Maximum Supported

Capacity Value for P-VOL/Data Pool (AMS2500)

Capacity of Cache Memory

Installed

8 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

10 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

12 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

16 GB/CTL Total P-VOL of SnapShot and P-VOL (S-VOL)

No SnapShot-TCE cascade configuration

In no SnapShot-TCE cascade configuration, you need to add all volumes of SnapShot and TCE in Table 4-17 to Table 4-19 .

1 TB×4 LU ÷ 5 + less than 0.6 TB < 1.4 TB

Capacity Spared for the Differential

Data (Shared by SnapShot and TCE)

of TCE capacity ÷ 5 + Total data pool capacity < 12.0 TB

of TCE capacity ÷ 5 + Total data pool capacity < 15.0 TB

of TCE capacity ÷ 5 + Total data pool capacity < 18.0 TB

of TCE capacity ÷ 5 + Total data pool capacity < 24.0 TB

•

Figure 4-2: No SnapShot-TCE cascade configuration

Plan and design—arrays, volumes and operating systems 4–29

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 86

SnapShot-TCE cascade configuration

In a SnapShot-TCE cascade configuration, you do not need to add volumes of TCE in Table 4-17 to Table 4-19. You need only to add volumes of SnapShot in the formula.

1 TB×2 LU ÷ 5 + less than 1 TB < 1.4 TB

•

Figure 4-3: SnapShot-TCE cascade configuration

4–30 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 87

Cache limitations on data and data pool volumes

This section provides comparisons in capacity between the data volumes and the data pool volumes under the limitations of the AMS controllers’ cache memory. The values in the tables and graph in this section are calculated from the formulas and maximum supported capacity in TCE and

SnapShot capacity on page 4-24.

•

NOTE: “Data volumes” in this section consist of TCE P-VOLs and S-VOLs

and SnapShot P-VOLs (if used).

•

Table 4-20: P-VOL to Data Pool Capacity Ratio on

AMS 2100 when Cache Memory is 2 GB/Controller

Ratio of All P-VOL Capacity

to All Data Pool Capacity

1:0.1 4.6 : 0.4

1:0.3 2.8 : 0.8

1:0.5 2.0 : 1.0

•

All P-VOL Capacity

to All Data Pool Capacity

(TB)

Table 4-21: P-VOL to Data Pool Capacity Ratio on

AMS 2300/2100 when Cache Memory is 4 GB per CTL

Ratio of All P-VOL Capacity

to All Data Pool Capacity

AMS 2100/2300/2500 AMS 2100/2300

1:0.5 20.6 : 2.0

1:1 12.4 : 3.7

1:3 8.8 : 4.4

All P-VOL Capacity

to All Data Pool Capacity (TB)

Table 4-22: P-VOL to Data Pool Capacity Ratio on

AMS 2500/2300 when Cache Memory is 8 GB per CTL

Ratio of All P-VOL Capacity

to All Data Pool Capacity

1:0.1 40.0 : 4.0

1:0.3 24.0 : 7.2

1:0.5 17.1 : 8.5

All P-VOL Capacity

to All Data Pool Capacity

(TB)

Plan and design—arrays, volumes and operating systems 4–31

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 88

Table 4-23: P-VOL to Data Pool Capacity Ratio on

AMS 2500 when Cache Memory is 16 GB per CTL

Ratio of All P-VOL Capacity

to All Data Pool Capacity

1:0.1 80.0 : 8.0

1:0.3 48.0 : 14.4

1:0.5 34.2 : 17.1

•

All P-VOL Capacity

to All Data Pool Capacity

(TB)

Figure 4-4: Relation of Data Volume, Data Pool Capacities to Cache Size

per Controller

Cautions for Reconfiguring the Cache Memory

The cautions for the cache memory reconfiguration processing in the installation, un-installation, or invalidation/validation operation occur when the firmware version of array is 0897/A or more.

• I/O processing performance

The I/O performance, in case of the sequential write pattern, deteriorates approximately 20% to 30% by releasing a part of the user data area in the cache memory and performing the memory reconfiguration of the management information storage area for TCE. In other patterns, the I/O performance deteriorates less than 10%.

• Time-out for memory reconfiguration processing

4–32 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 89

If the I/O inflow is large, data saving to the drives of the cache data takes time and may time out in 10 to 15 minutes (internal processing time is 10 minutes). In this case, the processing can be continued by executing it again at the time when I/O inflow is small.

• Inhibiting the memory reconfiguration processing performance while executing other functions

In the following items, the memory reconfiguration processing is inhibited to increase the data amount to the cache. Perform the memory reconfiguration processing again after completing the operation of other functions or recovering the failure.

- Other than master cache partition (partition 0 and partition 1) in

use

- Cache partition in changing

- DP pool in optimization

- RAID group in growing

- LU ownership in changing

- Cache Residency LU in operation

- Remote path and/or pair of TCE in operation

- SnapShot Logical Units or Data Pools in operation

- DMLU in operation

- Logical Unit in formatting

- Logical Unit in parity correction

- IP address for maintenance or management in operation

- SSL information in operation

- Array firmware in updating

- Power OFF of array in operation

- Spin-down or spin-up by Power Saving feature in operation

• Inhibiting the operation of other functions during memory reconfiguration

- When the memory reconfiguration processing fails on the way due

to the factors other then the time-out

- RAID group grown operation

- Replication Pair operation

- Dynamic Provisioning operation

- Cache Residency Manager setting operation

- Logical Unit formatting operation

- Logical Unit parity correction operation

- Cache Partition Manager operation

- Modular Volume Migration operation

- Array firmware updating operation

Plan and design—arrays, volumes and operating systems 4–33

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 90

- Installing, uninstalling, enabling, or disabling of extra-cost option operation

- Logical Unit operation

- Logical Unit unifying operation

• Un-installation/invalidation before the memory reconfiguration

When performing the un-installation or invalidation of TCE and SnapShot before the memory reconfiguration, the status of only the option operated at last is displayed in the Reconfigure Memory Status.

Table 4-24 shows the Memory Reconfiguring Statuses displayed on

Navigator 2.

Table 4-24: Memory Reconfiguring Statuses

Statuses Meaning

Normal Indicates that the memory reconfiguration processing

is completed normally.

Pending Indicates the status which is waiting for the memory

reconfiguration. Even if the memory reconfiguration instruction is executed and the message indicating the inoperable status is output, it is changed to this status because the instruction is received.

Reconfiguring(nn%) Indicates the status that the memory reconfiguration

is operating. (nn%) shows reconfiguring as a percent.

N/A Indicates that it is out of the memory reconfiguration

target.

Failed(Code-nn: error message)

Indicates the status that the memory reconfiguration failed because failures and others have occurred inside the array. Recover the status according to the following troubleshooting for each error code and each error message. If it still fails, call the Support Center.

Failed(Code-01: Time out)

Code-01 occurs when the access from the host is frequent or the amount of the unwritten data in the cache memory is large. Execute the memory reconfiguration operation again when the access from the host decreases.

Failed(Code-02: Failure of Reconfigure Memory)

Code-02 occurs when the drive restoration processing starts in the background. Execute the memory reconfiguration operation again after the drive restoration processing is completed.

Failed(Code-03: Failure of Reconfigure Memory)

Code-03 occurs when the copy of the management information in the cache memory fails. The controller replacement is required. Call the Support Center.

Failed(Code-04: Failure of Reconfigure Memory)

Code-04 occurs when the unwritten data in the cache memory cannot be saved to the drive. The restart of the array is required. Note: If the firmware version of the array is less than 0897/A, the memory reconfiguration without restart of the array is unsupported.

4–34 Plan and design—arrays, volumes and operating systems

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 91

Requirements and

specifications

This chapter provides TCE system requirements and specifications. Cautions and restrictions are also provided.

 TCE system requirements

 TCE system specifications

Requirements and specifications 5–1

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 92

TCE system requirements

Table 5-1 describes the minimum TCE requirements.

•

Table 5-1: TCE Requirements

Item Minimum Requirements

AMS firmware version Version 0832/B or later is required for AMS 2100 or

Navigator 2 version Version 3.21 or higher is required for the

AMS 2300 arrays with hardware Rev. 0100.

Version 0840/A or later is required for AMS2500 arrays with hardware Rev. 0100.

Version 0890/A or later is required for AMS2100, 2300, or 2500 arrays with hardware Rev. 0200.

Firmware version 0890/A or more is required on both local side and remote side arrays when connecting Rev. 0200 hardware.

management PC for AMS 2100 or 2300 arrays where the hardware Rev. is 0100.

Version 4.00 or higher is required for the management PC for an AMS2500 array where the hardware Rev. is 0100.

Version 9.00 or higher is required for the management PC for AMS 2100, 2300, or 2500 where the hardware Rev. is 0200.

CCI version 01-21-03/06 or later is required for Windows host

only when CCI is used for the operation of TCE.

Number of AMS arrays 2

Supported array AMS models AMS2100/2300/2500

TCE license keys One per array.

Number of controllers: 2 (dual configuration)

Volume size S-VOL block count = P-VOL block count.

Command devices per array (CCI only)

Max. 128. The command device is required only when CCI is used. The command device volume size must be greater than or equal to 33 MB.

5–2 Requirements and specifications

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 93

Displaying the hardware revision number

The hardware revision (Rev.) can be displayed when an individual array is selected from the Arrays list using Navigator 2, version 9.00 or later.

•

TCE system specifications

Table 5-2 describes the TCE specifications.

•

Table 5-2: TCE Specifications

Parameter TCE Specification

User interface • Navigator 2 GUI

•Navigator 2 CLI

•CCI

Controller configuration Configuration of dual controller is required.

Cache memory • AMS2100: 2 GB/controller

• AMS2300: 2, 4 GB/controller

• AMS2500: 2, 4, 6, 8 GB/controller

Host interface AMS 2100, 2300, and 2500: Fibre channel or iSCSI (cannot mix)

Remote path • One remote path per controller is required—totaling two for a pair.

• The interface type of multiple remote paths between local and remote arrays must be the same.

Number of hosts when remote path is iSCSI

Maximum number of connectable hosts per port: 239.

Requirements and specifications 5–3

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 94

Table 5-2: TCE Specifications (Continued)

Parameter TCE Specification

Data pool • Recommended minimum size: 20 GB

• Maximum # of data pools per array: 64

• Maximum # of LUs that can be assigned to one data pool: 64

• Maximum # of LUs that can be used as data pools: 128.

• When the array firmware version is less than 0852/A, a unified LU cannot be assigned to a data pool. If 0852/A or higher, a unified LU can be assigned to a data pool.

• Data pools must set up for both the P-VOL and S-VOL.

Port modes Initiator and target intermix mode. One port may be used for host I/O

and TCE at the same time.

Bandwidth • Minimum: 1.5 M.

• Recommended: 100M or more.

• When low bandwidth is used:

- The time limit for execution of CCI commands and host I/O

must be extended.

- Response time for CCI commands may take several seconds.

License Key is required.

Command device (CCI only)

DMLU • Required.

Maximum # of LUs that can be used for TCE pairs

Pair structure One S-VOL per P-VOL.

Supported RAID level • RAID 1 (1D+1D), RAID 5 (2D+1P to 15D+1P)

Combination of RAID levels

Size of LU The LU size must always be P-VOL = S-VOL.

Types of drive for P-VOL, S-VOL, and data pool

Supported capacity value of P-VOL and S-VOL

Copy pace User adjustable rate that data is copied to remote array. See the copy

• Required for CCI.

• Minimum size: (33 MB; 65,538 blocks (1 block = 512 bytes)

• Must be set up on local and remote arrays.

• Maximum # allowed per array: 128

• Must be set up on local and remote arrays.

• Minimum capacity per DMLU: 10 GB

• Maximum number allowed per array: 2

• If setting up two DMLUs on an array, they should belong to different RAID groups.

• AMS2100: 1,022

• AMS2300: 2,046

• AMS2500: 2,046

The maximum when different types of arrays are used for TCE (i.e. AMS500 and AMS2100) is the array with the smallest maximum.

• RAID 1+0 (2D+2D to 8D+8D)

• RAID 6 (2D+2P to 28D+2P)

Local RAID level can be different than remote level. The number of data disks does not have to be the same.

The max LU size is 128 TB.

If the drive types are supported by the array, they can be set for a P-VOL, an S-VOL, and data pools. SAS, SAS7.2K, SSD, or SAS (SED) drives are recommended. Set all configured LUs using the same drive type.

Capacity is limited. See Maximum supported capacity on page 4-23.

pace step on page 7-6 for more information.

5–4 Requirements and specifications

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 95

Table 5-2: TCE Specifications (Continued)

Parameter TCE Specification

Consistency Group (CTG) • Maximum allowed: 16

• Maximum # of pairs allowed per consistency group:

- AMS2100: 1,022

- AMS2300: 2,046

- AMS2500: 2,046

Management of LUs while using TCE

Pair creation using unified LUs

Restriction during RAID group expansion

Unified LU for data pool Not allowed.

Differential data When pair status is Split, data sent to the P-VOL and S-VOL are

Host access to a data pool A data pool LU is hidden from a host.

Expansion of data pool capacity

Reduction of data pool capacity

Failures • When the copy operation from P-VOL to S-VOL fails, TCE suspends

Data pool usage at 100% When data pool usage is 100%, the status of any pair using the pool

Array restart at TCE installation

TCE use with TrueCopy Not Allowed.

TCE use with SnapShot SnapShot can be cascaded with TCE or used separately.

TCE use with ShadowImage

TCE use with LUN Expansion

A TCE pair must be deleted before the following operations:

• Deleting the pair’s RAID group, LU, or data pool

• Formatting an LU in the pair

• Growing or shrinking an LU in the pair

• A TCE pair can be created using a unified LU.

- When array firmware is less than 0852/A, the size of each LU making up the unified LU must be 1 GB or larger.

- When the array firmware is 0852/A or later, there are no restrictions on the LUs making up the unified LU.

• LUs that are already in a P-VOL or S-VOL cannot be unified.

• Unified LUs that are in a P-VOL or S-VOL cannot be released.

A RAID group in which a TCE P-VOL or data pool exists can be expanded only when pair status is Simplex or Split. If the TCE data pool is shared with SnapShot, the SnapShot pairs must be in Simplex or Paired status.

managed as differential data.

• Data pools can be expanded by adding an LU.

• Mixing SAS/SSD and SATA drives in a data pool is not supported. Set all configured LUs using the same drive type.

Yes. The pairs associated with a data pool must be deleted before the data pool can be reduced.

the pair (Failure). Because TCE copies data to the remote S-VOL regularly, data is restored to the S-VOL from the update immediately before the occurrence of the failure.

• A drive failure does not affect TCE pair status because of the RAID architecture.

becomes Pool Full. P-VOL data cannot be updated to the S-VOL.

The array is restarted after installation to set the data pool, unless it is also used by SnapShot. Then there is no restart.

Only a SnapShot P-VOL can be cascaded with TCE.

Although TCE can be used at the same time as a ShadowImage system, it cannot be cascaded with ShadowImage.

When firmware version is less than 0852/A, it is not allowed to create a TCE pair using unified LUs, which unify the LU with 1 GB or less capacity.

Requirements and specifications 5–5

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 96

Table 5-2: TCE Specifications (Continued)

Parameter TCE Specification

TCE use with Data Retention Utility

TCE use with Cache Residency Manager

TCE use with Cache Partition Manager

TCE use with SNMP Agent Allowed. A trap is transmitted for the following:

TCE use with Volume Migration

TCE use with Power Saving

Reduction of memory The memory cannot be reduced when the ShadowImage, SnapShot,

Load balancing Not supported.

LU assigned to data pool An LU consisting of a SAS, SSD drives, and an LU with a SATA drive

Extension of influence of the TCE function installation

Remote Copy over iSCSI in the WAN environment

Allowed.

• When S-VOL Disable is set for an LU, a pair cannot be created using the LU as the S-VOL.

• S-VOL Disable can be set for an LU that is currently an S-VOL, if pair status is Split.

Allowed. However, an LU specified by Cache Residency Manager cannot be used as a P-VOL, S-VOL, or data pool.

• TCE can be used together with Cache Partition Manager.

• Make the segment size of LUs to be used as a TCE data pool no larger than the default, (16 KB).

• See Appendix D, Installing TCE when Cache Partition Manager is in

use for details on initialization.

• Remote path failure.

• Threshold value of the data pool is exceeded.

• Actual cycle time exceeds the default or user-specified value.

• Pair status changes to:

- Pool Full

- Failure.

- Inconsistent because the data pool is full or because of a

failure.

Allowed. However, a Volume Migration P-VOL, S-VOL, or Reserved LU cannot be used as a TCE P-VOL or S-VOL.

Allowed, however, pair operations are limited to split and delete.

TCE, or Volume Migration function is validated. Make the reduction after invalidating the function.

cannot coexist in a data pool. Set all the LUs configured with the same drive type.

When the firmware version of the array is less than 0897/A, you must restart the array to ensure the data pool resource. However, the restart is not required when the data pool is already used by SnapShot because TCE and SnapShot share the data pool.

We recommend using TrueCopy in the WAN environment of MTU1500 or more. However, if TCE needs to be implemented in the WAN environment of less than MTU1500, change the maximum segment size (MSS) of the WAN router to a value less than 1500, and then create a remote path. The data length transmitted from TCE of HUS100 changes to the specified value less than 1500.

When creating a remote path without changing the MSS value or not creating the remote path after changing the MSS value again, a data transfer error occurs because TCE of HUS100 transmits the MTU1500 data. To change the MSS value, request the customer or the WAN router provider.

5–6 Requirements and specifications

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 97

Installation and setup

This chapter provides TCE installation and setup procedures using the Navigator 2 GUI. Instructions for CLI and CCI can be found in the appendixes.

 Installation procedures

 Setup procedures

Installation and setup 6–1

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 98

Installation procedures

The following sections provide instructions for installing, enabling/disabling, and uninstalling TCE. Please note the following:

• TCE must be installed on the local and remote arrays.

• Before proceeding, verify that the array is operating in a normal state. Installation/un-installation cannot be performed if a failure has occurred.

In cases where the DKN-200-NGW1 (NAS unit) is connected to the disk array, check the following items in advance.

1. Prior to this operation, execute Correspondence when connecting the NAS unit if each of the following three items apply to the disk array.

- NAS unit is connected to the disk array. Ask the disk array

administrator to confirm whether the NAS unit is connected or not.

- NAS unit is in operation. Ask the NAS unit administrator to confirm

whether the NAS service is operating or not.

- A failure has not occurred on the NAS unit. Ask the NAS unit

administrator to check whether failure has occurred or not by checking with the NAS administration software, NAS Manager GUI, List of RAS Information, etc. In case of failure, execute the maintenance operation together with the NAS maintenance personnel.

2. Correspondence when connecting the NAS unit.

- If the NAS unit is connected, ask the NAS unit administrator for

termination of NAS OS and planned shutdown of the NAS unit.

3. Points to be checked after completing this operation:

- Ask the NAS unit administrator to reboot the NAS unit. After

rebooting, ask the NAS unit administrator to refer to “Recovering from FC path errors” in the check the status of the Fibre Channel path (FC path in short) and to recover the FC path if it is in a failure status.

- In addition, if there are any personnel for the NAS unit

maintenance, ask the NAS unit maintenance personnel to reboot the NAS unit.

Hitachi NAS Manager User's Guide and

6–2 Installation and setup

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 99

Installing TCE

Prerequisites

• A key code or key file is required to install or uninstall TCE. If you do not have the key file or code, you can obtain it from the download page on the HDS Support Portal, http://support.hds.com.

• The array may require a restart at the end of the installation procedure. If SnapShot is enabled at the time, no restart is necessary.

• If restart is required, it can be done either when prompted or at a later time.

• TCE cannot be installed if more than 239 hosts are connected to a port on the array.

To install TCE without rebooting

1. In the Navigator 2 GUI, click the check box for the array where you want

to install TCE, then click Show & Configure Array.

2. Under Common Array Tasks, click Install License. The Install License

screen displays.

•

3. Select the Key File or Key Code radio button, then enter the file name

or key code. You may browse for the Key File.

4. Click OK.

5. Click Confirm on the screen requesting a confirmation to install the TCE

option.

6. Click Reconfigure Memory to install the TCE option.

7. Click Close. The Licenses list appears.

8. Confirm TC-EXTENDED on the Name column of the Installed Storage

Features list, and Pending on the Reconfigure Memory Status column.

9. Check the check box of TC-EXTENDED, and click Reconfigure Memory.

10.Click Confirm in the Reconfigure Memory menu and then click Close. The Licenses list appears.

11.Confirm the Reconfigure Memory Status is Reconfiguring(nn%) or Normal.

12.When the Reconfigure Memory Status is Reconfiguring(nn%), click Refresh Information after waiting for a while, and confirm the Reconfigure Memory Status changes to Normal.

13.When the Reconfigure Memory Status is Failed(Code- 01:Timeout), click Install License, and re-execute steps 6 to 13.

Code-01 occurs when the access from the host is frequent or the amount of the unwritten data in the cache memory is large.

Installation and setup 6–3

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide

Page 100

14.When the Reconfigure Memory Status is Failed(Code-02: Failure of Reconfigure Memory), perform steps 9 to 13.

Code-02 occurs when the drive restoration processing starts in the background.

15.When the Reconfigure Memory Status is Failed(Code-04: Failure of Reconfigure Memory), click the Resource of the Explorer menu, then Arrays to return to the Arrays screen.

Code-04 occurs when the unwritten data in the cache memory cannot be saved to the drive.

16.Select the array in which you will install TCE, click Reboot Array.

17.When the Reconfigure Memory Status is Failed(Code-03: Failure of Reconfigure Memory), ask the Support Center to solve the

problem.

Code-03 occurs when the copy of the management information in the cache memory fails.

18.Installation of TCE is now complete.

To install TCE with rebooting

1. In the Navigator 2 GUI, click the check box for the array where you want

to install TCE, then click Show & Configure Array.

2. Under Common Array Tasks, click Install License. The Install

License screen displays.

•

3. Select the Key File or Key Code radio button, then enter the file name

or key code. You may browse for the Key File.

4. Click OK.

5. Click Confirm on the screen that appears, requesting a confirmation to

install TCE option.

6. Click Reboot Array to install the TCE option.

7. A message appears confirming that this optional feature is installed.

Mark the check box and click Reboot Array.

The restart is not required at this time if it is done when validating the function. However, in the case where the installation of TCE was completed before TCE is installed, the dialog box, which asks whether or not to do the restart, is not displayed at this time because the restart was done to ensure the resource for the data pool in the cache memory. When the restart is not needed, the installation of the TCE function is completed.

If the array restarts because of a spin-down instruction when Power Saving, the spin down may fail if the instruction is received immediately after the array restarts. When the spin-down fails, perform the spin-

6–4 Installation and setup

Hitachi AMS 2000 Family TrueCopy Extended Distance User Guide