NEC IS015-9E User Manual

NEC Storage Manager

Data Replication User’s Manual

(Function Guide)

IS015-9E

No part of the contents of this book may be reproduced or

transmitted in any form without permission of NEC Corporation.

The contents of this book may be modified without notice in the future.

Preface

This manual describes how to use the data replication function provided by NEC Storage DynamicDataReplication

Ver2, NEC Storage RemoteDataReplication Ver2, and NEC Storage ReplicationControl.

The data replication function consists of the replication volume creation function provided in a disk array and

software to manage and operate it. It utilizes replication volume to make business operation more effective.

Refer to the “NEC Storage Manager Manual Guide” (IS901) for the overview of NEC Storage and the related

manuals. Refer to the “NEC Storage Manager Data Replication User’s Manual (Disaster Recovery System

Installation and Operation Guide)” (IS027) for the usage of the remote data replication functions provided by NEC

Storage RemoteDataReplication/DisasterRecovery and NEC Storage ReplicationControl/DisasterRecovery.

Remarks 1. This manual explains functions implemented by the following program products:

• NEC Storage Manager and NEC Storage BaseProduct

• NEC Storage DynamicDataReplication

• NEC Storage ReplicationControl

• NEC Storage RemoteDataReplication

2. This manual is applicable to the program products of the following versions:

• NEC Storage Manager Ver3.3

• NEC Storage BaseProduct Ver3.3

• NEC Storage ReplicationControl Ver3.3

3. The NEC Storage Manager is referred to as iSM or Storage Manager in the text of this manual.

Also, the NEC Storage series disk array subsystem is referred to as a disk array.

4. The following descriptions in the text of this manual refer to the corresponding products.

Storage Manager NEC Storage Manager

AccessControl NEC Storage AccessControl

DynamicDataReplication NEC Storage DynamicDataReplication

RemoteDataReplication NEC Storage RemoteDataReplication

RemoteDataReplication/DisasterRecovery NEC Storage RemoteDataReplication/DisasterRecovery

ReplicationControl NEC Storage ReplicationControl

SnapControl NEC Storage SnapControl

Description Corresponding Product

5. The following descriptions in the text of this manual refer to the corresponding manuals.

Description Corresponding Manual

User's Manual (UNIX) NEC Storage Manager User's Manual (UNIX) (IS001)

User's Manual NEC Storage Manager User's Manual (IS004)

Configuration Setting Tool User’s Manual (GUI)

Data Replication User’s Manual (Installation and Operation Guide for Windows)

Data Replication Command Reference

NEC Storage Manager Configuration Setting Tool User’s Manual (GUI) (IS007)

NEC Storage Manager Data Replication User’s Manual (Installation and Operation Guide for Windows) (IS016)

NEC Storage Manager Data Replication Command Reference (IS021)

Description Corresponding Product

Data Replication User’s Manual (Disaster Recovery System Installation and Operation Guide)

Snapshot User’s Manual (Function Guide)

6. Trademarks and registered trademarks

• Microsoft® and Windows® are trademarks or registered trademarks of Microsoft Corporation in the

United States and other countries.

• HP-UX is a registered trademark of Hewlett-Packard Co. in the United States.

• UNIX is a registered trademark of The Open Group in the United States and other countries.

• VERITAS, VxVM, VxFS, NetBackup, VERITAS Volume Manager, VERITAS File System, and

VERITAS NetBackup are trademarks or registered trademarks of VERITAS Software Corporation in

the United States and other countries.

• Legato NetWorker is a registered trademark of Legato Systems, Inc. in the United States.

• Sun is a registered trademark of Sun Microsystems, Inc. in the United States and other countries.

• Solaris is a trademark or a registered trademark of Sun Microsystems, Inc. in the United States and

other countries.

• Linux is a trademark or registered trademark of Mr. Linus Torvalds in the United States and other

countries.

• AIX is a trademark of IBM Corporation.

Other product names and company names, etc. are trademarks or registered trademarks of the

associated companies.

7. In this document, matters to which careful attention needs to be paid will be described as follows:

Be sure to observe the contents.

If the indications are ignored and the system is improperly operated, settings which have been already

made might be affected.

NEC Storage Manager Data Replication User’s Manual (Disaster Recovery System Installation and Operation Guide) (IS027)

NEC Storage Manager Snapshot User’s Manual (Function Guide) (IS030)

Type of Indication

Type Description

Describes contents which require special attention during operation.

The First Edition in March 2003

The Ninth Edition in November 2004

Chapter 1 Data Replication Overview

To manage an enormous amount of information accumulated in business in a unified way and promote effective

and efficient utilization of the information, a high-throughput, large-capacity, and high-reliability storage system is

required. Data Replication provides functions to build and manage such a storage system.

This chapter describes overview of Data Replication, hardware configuration, and software configuration.

1

.

1

D

a

t

a

R

e

p

l

i

c

a

t

i

o

n

1

.

1

D

a

t

a

R

e

p

l

i

c

a

1

.

1

D

a

t

a

R

e

p

l

Data Replication is a function that creates Replication Volumes (RV) of a business volume (called

Master Volume (MV) in the data replication). It is installed in the disk array. Replication volumes

can be connected to or separated from the master volume at any time. Operations such as connection

and separation can be instructed from the business server and the iSM Client (e.g. Windows system).

The following two methods are provided to create replication volumes.

(1) Creating replication volumes within the same disk array (DDR: DynamicDataReplication)

(2) Creating replication volumes in different disk arrays (RDR: RemoteDataReplication)

Synchronize/Separate

Master Server

Operation

Master

(MV)

Replication

(RV)

DynamicDataReplication

Master

(MV)

Figure 1-1 Data Replication

Production Task

t

i

c

a

t

(Online)

RemoteDataReplication

i

o

n

i

o

n

Parallel Processing

(Backup/Search)

Replication

(RV)

1

Chapter 1 Data Replication Overview

r

1

.

2

E

x

a

m

p

l

e

1

.

2

E

x

a

1

.

2

E

R

When you introduce Data Replication and use replication volumes which can be separated, you can get

the following benefits.

• The system down time during data backup is largely reduced. Lowered access performance to the

business database during data backup in system operation can be prevented.

• A test environment using the actual business data can be built more easily.

• Processing becomes more efficient due to parallel processing of data update tasks and data reference

tasks.

In this way, Data Replication makes system construction and system management easier and more

effective.

The following sections illustrate some applications of Data Replication.

x

e

a

p

m

l

i

l

i

l

i

c

p

a

s

l

e

s

l

e

s

t

i

o

t

i

o

t

i

o

n

o

n

f

A

p

l

y

i

n

g

D

a

t

a

f

A

p

l

y

i

n

g

f

A

p

l

y

i

n

g

D

a

t

a

t

a

1

.

2

.

1

B

a

c

k

u

p

1

.

2

.

1

B

a

c

1

.

2

.

1

B

This section describes an application for backup using replication volumes.

[Backing Up from the Replication Volume to Magnetic Tape Media]

In this method, a replication volume of the master volume is backed up to the magnetic tape.

In this case, operations are suspended only for the time it takes to separate replication volumes from the

master volume. Therefore, the suspension time can be substantially reduced.

Because backup is done from the replication volume, it does not affect the master volume.

Master Serve

Production Task

k

a

c

Connect

u

p

k

u

p

Backup Serve

Maste

(MV)

Backup

Separate

Replication

(RV)

Figure 1-2 Backup from Replication Volume

Tape

2

Chapter 1 Data Replication Overview

r

The procedure for backup from a replication volume is described below.

(1) During operation, the master volume (MV) and replication volume (RV) are connected.

(2) Suspend the production task and separate the replication volume (RV). Resume the task after

separation is complete.

(3) Use the separated replication volume (RV) to perform backup and the task in parallel. After

backup is complete, reconnect the replication volume (RV) (Reconnection takes only a short time

because only updated parts in the master volume are reflected to the replication volume (RV)).

[Using the Replication Volume as Disk Backup]

In this method, the replication volume is used as backup of the master volume. In this case, you do

not have to manage the existing magnetic tapes because they are not used as storage media.

When the restoration instruction is complete, you can use the backup data even if the actual data

replication has not been completed. If data to be accessed is not restored to the master volume, the

data in the replication volume is accessed. The user does not have to be aware of using which of the

master volume or replication volume.

This reduces the data restoration time substantially.

Connect

Maste

(MV)

Separate

The procedure for using a replication volume for backup is described below.

(1) Connect the replication volume 1 (RV1) to the master volume.

(2) Suspend the production task and separate the replication volume 1 (RV1). Resume the task after

connecting the replication volume 2 (RV2).

(3) After that, use the replication volume 1 (RV1) and replication volume 2 (RV2) alternately to

perform backup.

Replication 1

(RV1)

Maste

(MV)

Replication 2

(RV2)

Figure 1-3 Using Replication Volume as Backup

Separate

Connect

Replication 1

(RV1)

Replication 2

(RV2)

3

Chapter 1 Data Replication Overview

1

.

2

.

2

T

e

s

t

E

n

v

i

1

.

2

.

2

T

e

s

t

1

.

2

.

2

T

e

You can set the same environment as the production task environment easily by creating replication

volumes using the Data Replication function. You can evaluate an application program by using data

used in the production task, which makes evaluation of application programs more efficient.

Furthermore, in building an evaluation environment, operations are suspended only for the time which

takes to separate replication volumes from the master volume. Therefore, the suspension time can be

substantially reduced.

Production

Task

s

E

t

E

n

v

r

i

r

i

o

n

m

e

n

t

S

e

t

i

n

g

o

n

m

e

n

t

S

e

t

r

o

n

m

e

n

t

S

e

t

Production

i

n

i

n

Task

g

Test

Master

(MV)

Connect

The procedure for using a replication volume as a test environment is described below.

(1) Connect the master volume (MV) and replication volume (RV).

(2) Suspend the production task, separate the replication volume (RV), and then resume the task.

(3) Perform evaluation of the application program by using the separated replication volume (RV).

Replication

(RV)

Master

(MV)

Figure 1-4 Test Environment Setting

Separate

Replication

(RV)

4

Chapter 1 Data Replication Overview

1

.

2

.

3

P

a

r

a

l

e

l

P

r

o

c

e

s

i

n

g

o

f

S

e

a

r

c

h

1

.

2

.

3

P

a

r

a

l

e

l

P

r

o

c

e

s

i

n

g

o

f

S

e

1

.

2

.

3

P

a

r

a

l

e

l

P

r

o

c

e

s

i

n

g

o

f

O

p

e

r

a

t

i

o

n

O

p

e

r

a

O

p

By creating replication volumes of the master database, you can separate the database and carry out

update tasks and search tasks using different volumes. This allows you to carry out database update

tasks without affecting database search tasks.

Nighttime

Master

(MV)

Master

Database

The procedure for parallel processing of search operation is described below.

(1) In the nighttime, suspend search tasks and connect the master volume (MV) and replication

(2) In the daytime, separate the master volume (MV) and replication volume (RV). Then perform

Connect

volume (RV).

update tasks and search tasks in parallel (RV contains data for the prior day).

t

e

r

a

t

Replication

Reflection of

Updated Parts

Figure 1-5 Parallel Processing of Search Task

i

o

i

o

(RV)

n

Daytime

Master

(MV)

Master

Database

a

S

e

a

Update Task

Separate

r

c

h

r

c

h

Data for Previous

Day

Search Task

Replication

(RV)

5

Chapter 1 Data Replication Overview

1

.

3

S

y

s

t

e

m

1

.

3

S

y

s

t

.

3

S

y

s

e

t

e

m

C

o

n

f

i

g

u

r

a

t

i

o

n

f

i

g

u

r

a

u

t

r

a

t

f

i

g

i

o

n

i

o

n

Business Server

ReplicationControl

DB

• AccessControl

• DynamicDataReplication

• RemoteDataReplication

Disk Array

Figure 1-6 System Configuration

To install and use the Data Replication function, the following hardware devices are required.

• Disk array

Disk arrays on which the Data Replication function either with DynamicDataReplication or

RemoteDataReplication is installed are required.

• Management server

iSM is installed in the management server that monitors disk arrays. This server controls disk arrays

and the Data Replication function.

• Business server/backup server

Performs Data Replication operation or backup operation in cooperation with business.

Software to run Data Replication consists of the following components.

Backup Server

Backup Software

Tape Drive

ReplicationControl

Management

Server

Replication

LAN

Storage Manager

6

Chapter 1 Data Replication Overview

• Storage Manager

Provides the disk array configuration and state display functions.

Installing DynamicDataReplication and/or RemoteDataReplication allows the replication

management function (hereinafter, referred to as the Replication Management) incorporated in iSM.

The Replication Management provides setting and operating functions such as state display, pair

setting, and replication operation for DynamicDataReplication or RemoteDataReplication.

• ReplicationControl

Provides commands for checking replication operations and replication states from the business

server and also provides library functions.

• AccessControl

The function to set the logical disks that can be accessed, for each business server.

• DynamicDataReplication

The function to realize data replication within the same disk array.

• RemoteDataReplication

The function to realize data replication in the different disk arrays.

7

Chapter 2 Data Replication

This chapter describes the types and state transitions of volumes that are necessary to perform operation using the

Data Replication function.

2

.

1

V

o

l

u

m

e

C

l

a

s

i

f

i

c

a

t

i

o

n

2

.

1

V

o

l

u

m

e

C

l

a

s

i

f

i

c

a

2

.

1

V

o

l

u

m

e

C

l

a

s

i

f

To create replication volumes using the Data Replication function, you must set the relation between

the original volume and the target volume (replication volume) first. In replication control, the

original volume is called MV (Master Volume), and replication volume is called RV (Replication

Volume). Furthermore, MV and RV are set as a pair.

In a disk array to which the Data Replication function is installed, the volume classification in the disk

array is categorized into the following three types. Figure 2-1 describes sample volume classification.

• Isolated Volume (IV)

Volume with no pair setting.

By specifying the pair setting to IV, you can set it to MV or RV.

• Master Volume (MV)

Volume with the pair setting. The original volume in the pair. Normally, volumes used in

operation are set as MVs.

To distinguish the uppermost MV from other MVs when multiple pairs are set in series hierarchically,

it is called the Primary Volume (PV).

• Replication Volume (RV)

Volume with the pair setting. The target volume in the pair. Normally, volumes used as backup or

in test operation are set as RVs.

The pair setting can be between the volumes within the same disk array, or between the volumes in

different disk arrays. To distinguish them, the former is called a Dynamic Replication Volume

(dRV) and the latter is called a Remote Replication Volume (rRV).

t

i

c

a

t

i

o

n

i

o

n

8

Chapter 2 Data Replication

Volume A1/A2/B1 : IV

Volume A3 : MV (PV) of the pair A3/A4

Volume A4 : RV (dRV) of the pair A3/A4, and MV of the pair A4/B2

Volume B2 : RV (rRV) of the pair A4/B2

Disk Array A Disk Array B

FC

A1

IV

A3

MV

(PV)

A2

IV

A4

RV/MV

(dRV)

B1

IV

B2

RV

(rRV)

Pair Setting

Figure 2-1 Example of Volume Classification

9

Chapter 2 Data Replication

2

.

2

R

e

p

R

S

e

p

l

2

.

2

.

2

The replication operations include Replicate that replicates data from MV to RV, Separate that

separates between MV and RV, and Restore that replicates data from RV to MV.

2

.

2

.

2

.

2

This operation copies data from MV to RV.

It is performed to replicate the latest data to the replication volume used in a test environment or search

tasks. When Replicate is executed, the data in MV is copied to RV. In addition, any update made to

MV after Replicate is reflected to RV.

2

.

2

.

2

.

2

R

.

1

.

1

.

1

.

2

.

2

.

2

i

c

a

t

i

o

n

O

p

e

r

a

t

i

o

n

s

l

i

c

a

t

i

o

n

O

p

e

r

a

t

i

l

a

l

a

i

c

a

t

i

o

n

O

p

e

r

i

c

a

t

e

a

t

e

t

e

t

e

t

e

t

e

i

c

i

c

r

a

r

a

o

t

i

o

n

s

This operation separates MV and RV.

It is performed to suspend data replication between MV and RV to use RV in a test environment or

search tasks.

When Separate is started, all the difference between the MV and RV contents at the point of starting

Separate is reflected into the RV, and then data replication is suspended and the RV is separated. The

updates made to MV after starting Separate are not reflected to RV and stored in the disk array as

update differences.

When executing Separate, you can determine when to make the RV available by choosing either of the

following:

• Separate for making RV available after completion of separation: Separate(completion)

Reflects all the difference between the MV and RV contents into the RV, and makes the RV available

after completion of separation.

Even though Separate is executed immediately after Replicate starts, RV cannot be used while the

difference between MV and RV is being reflected to RV. RV becomes available upon completion

of separation.

• Separate for immediately making RV available: Separate(immediate)

While reflecting the difference between the MV and RV contents into the RV, the Separate function

makes the RV available even during separation. You can instantly create RV and make it available

by executing Separate(immediate).

This function is available only for performing data replication in the same disk array. The product

“DynamicDataReplication Ver2“ is necessary for using the function.

10

Chapter 2 Data Replication

2

.

2

.

3

R

e

s

t

o

r

e

2

.

2

.

3

R

e

s

2

.

2

.

3

R

This operation copies data from RV to MV.

It is performed to restore data from the backup volume (RV) when a failure occurs in MV.

When Restore is executed, the RV contents at the point of starting Restore are reflected copied into the

MV. At this time, you can determine whether to reflect the updated data of the MV into the RV by

choosing either of the following:

• Restore with RV being updated: Restore(update)

Restores the MV while automatically reflecting the updated data of the MV into the RV. Even after

the difference between MV and RV is resolved and Restore is completed, any data update made to

MV is reflected to RV.

• Restore without RV being updated: Restore(protect)

Restores the MV without reflecting the updated data of the MV into the RV. After the difference

between MV and RV is resolved and Restore is completed, Separate is automatically executed. The

Restore(protect) function enables you to save the RV data in the state before the restoration.

The product DynamicDataReplication Ver2 or RemoteDataReplication Ver2 is necessary for using

the function.

e

s

t

o

r

e

r

e

11

Chapter 2 Data Replication

•

2

.

3

R

e

p

r

e

r

a

p

n

l

2

.

3

2

.

3

This section describes the replication operations and state transitions.

R

T

i

c

a

t

i

o

n

O

p

e

r

a

t

i

o

n

s

a

n

d

S

t

a

t

e

l

i

c

a

t

i

o

n

O

p

e

r

a

t

i

o

n

s

a

n

d

S

l

i

c

a

t

i

o

n

O

p

e

r

a

t

i

o

n

s

a

n

d

s

i

t

i

o

n

s

i

t

i

o

i

t

i

o

n

s

S

t

a

t

e

t

e

The pair setting is specified but MV and RV are separated.

separated

• MV and RV can be updated.

MV RV

Difference between MV and RV is being reflected.

MV

Replicate

• RV cannot be updated.

rpl/exec

sep/exec

Separate

• As a rule, RV cannot be updated.

rst/exec

Restore

• RV cannot be updated.

MV

RV

The pair setting is specified and synchronization has been established.

sync

Updates to MV are reflected to RV.

• As a rule, access to RV is not allowed.

rpl/sync

MV RV

rst/sync

When Restore(update) is

executed, updates to MV are reflected to RV.

• When Restore(protect) is executed, updates to MV are not reflected to RV.

• As a rule, access to RV is not allowed.

exec : Executing Replicate, Separate, or Restore sync : Replicate or Restore synchronization state rpl : replicate (Replicate)

rpl/exec ............. Executing synchronization

rpl/sync ............. Synchronization state

sep : separate (Separate)

sep/exec ............ Executing Separate

separated .......... Separate completion state

rst : restore (Restore)

rst/exec ............. Executing Restore

rst/sync ............. Synchronization state

Figure 2-2 Replication Operations and State Transitions

12

When Restore(protect) is executed,

Synchronous State (rst/sync) is placed and then automatically changed to Separated State (separated).

Chapter 2 Data Replication

2

.

3

.

1

R

e

p

l

i

c

a

t

e

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

o

n

s

2

.

3

.

1

R

e

p

l

i

c

a

t

e

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

2

.

3

.

1

R

e

p

l

i

c

a

t

e

a

n

d

S

t

a

t

e

T

r

a

n

When Replicate is performed, data copy from MV to RV starts to reflect the content of MV to RV.

Any update to MV after Replicate is also reflected to RV.

After Replicate is started, the difference between MV and RV gradually decreases, and eventually the

content of MV at the beginning of Replicate is completely reflected to RV (The difference is zero).

The state from the beginning of Replicate to the content of MV is completely reflected to RV is called

the “Replicate execution”. The state where the difference between MV and RV is zero is called the

state synchronized by Replicate, or simply the “synchronous state“. Replicate execution and the state

synchronized by Replicate are collectively called the Replicate state.

Update to MV

Replicate

MV

RV

operation

starts

Update to MV

MV

RV

Until the difference is zero

s

o

i

t

i

o

n

s

n

s

Update to MV

MV

RV

Synchronous Execution

Replicate State

Separate State

Synchronous State

Figure 2-3 Replicate and State Transitions

13

Chapter 2 Data Replication

2

.

3

.

2

S

e

2

.

3

.

2

When Separate is performed, the difference between MV and RV at the time of executing the Separate

start instruction is reflected to RV and RV is separated. No update to MV after Separate is reflected to

RV.

After Separate is performed, data copy to RV is performed if the contents of MV and RV at the

beginning of Separate do not match, and all updates to MV before the Separate start instruction are

reflected to RV. The state from the beginning of Separate to the content of MV at the beginning of

Separate is completely reflected to RV is called the “Separate execution state“. The state where all

updates to MV are reflected to RV is called the state separated by Separate, or the “separated state“.

Separate execution and the separated state are collectively called the Separate state.

When Separate is executed under specification of immediate use of RV (Separate(immediate)), the RV

contents can be referred to or updated immediately after the Separate start instruction is issued,

regardless of whether or not all the MV contents have been reflected into the RV. This feature is

implemented as follows.

When an update/reference request for the RV is made and access to an area where difference copy from

the MV into RV is not completed is to be made, control is performed for copying the difference from

the MV into RV before permitting access to the area.

The updates made to MV until Separate is started are reflected to RV. The updates made to MV after

Separate is started are not reflected to RV and managed as update difference.

The updated states of MV and RV are managed in Separate State so that the difference between the MV

and RV contents is reflected when Replicate/Restore is executed.

.

3

.

2

Update to MV

S

e

p

a

r

a

t

e

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

o

n

s

r

a

t

e

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

r

a

t

e

a

n

d

S

t

a

t

e

T

r

a

n

Update to MV

s

o

i

t

i

o

n

s

n

s

Update to MV

Until the difference between MV and

Synchronous State

MV

RV

Separate operation starts

Separate Execution

Figure 2-4 Separate and State Transitions

MV

RV

14

RV is fully reflected to RV

Separated State

Separate State

MV

RV

Chapter 2 Data Replication

2

.

3

.

3

R

e

s

t

o

r

e

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

o

n

s

2

.

3

.

3

R

e

s

t

o

r

e

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

2

.

3

.

3

R

e

s

t

o

r

e

a

n

d

S

t

a

t

e

T

r

a

n

When Restore is performed, data copy from RV to MV starts to reflect the content of RV at the

beginning of Restore to MV. When Restore(update) is executed, any update to MV after Restore is

also reflected to RV.

After Restore is started, the difference between MV and RV gradually decreases, and eventually the

content of RV at the beginning of Restore is completely reflected to MV (The difference is zero). The

state from the beginning of Restore to the content of RV at the beginning of Restore is completely

reflected to MV is called the “Restore execution”. The state where the difference between MV and

RV is zero is called the state synchronized by Restore, or simply the “synchronous state“. Restore

execution and the state synchronized by Restore are collectively called the Restore state.

When Restore(protect) is executed, the updated data of the MV is not reflected into the RV. In this

case, the updated information of the MV is managed as the difference between the MV and RV contents

so that the difference can be reflected into the RV when Replicate/Restore is executed subsequently.

When the Synchronous State (sync) is placed after Restore(protect) is executed, it is automatically

changed to the Separated State (separated).

When data of MV is referred to during Restore execution, the user can refer to the content of RV

immediately after the instruction to start Restore even if the content of RV has not been completely

reflected to MV. This is done by obtaining data from RV when the area where difference copy from

RV to MV has not been completed is accessed in response to a reference request to MV.

Update to MV

MV

RV

Separate State

Restore operation starts

Update to MV

MV

RV

Restore Execution

Until the difference is zero

Only when updating RV

o

s

i

t

i

o

Synchronous State

n

s

n

s

Update to MV

MV

RV

Restore State

Figure 2-5 Restore and State Transitions

15

Chapter 2 Data Replication

2

.

3

.

4

A

c

t

i

v

i

t

y

S

t

a

t

e

a

n

d

S

2

.

3

.

4

A

c

t

i

v

i

t

y

S

t

a

t

e

a

2

.

3

.

4

A

c

t

i

v

i

t

y

S

t

a

t

S

t

a

t

e

S

t

S

In data replication, Replicate, Restore, and Separate states are called “activity states”, or simply

“activities”.

The execution states indicating state transitions and the state in which the state transition is complete

are called “synchronous states”.

Table 2-1 shows the activity states and synchronous states which transit as a result of each replication

operation. For information on handling MV and RV access in the activity states, refer to 2.6 “RV

Access Restriction”.

Activity State Synchronous State Description

Separate State

Replicate State

Restore State

Separate Execution (sep/exec)

Separated(separated)

Forced Separate(cancel)

Failure Separation (fault)

Synchronous Execution (rpl/exec)

Synchronous State(rpl/sync)

Restore Execution(rst/exec)

Synchronous State(rst/sync)

a

t

e

t

a

t

e

Table 2-1 Activity State and Synchronous State

• Temporal state until the difference between MV and RV

• As a rule, read and write to RV are not allowed.

• Data copy between MV and RV is not performed. This

• Normally, read and write to RV are allowed.

• MV and RV are forcibly separated by Forced Separate.

• Read and write to RV are allowed.

• MV and RV are forcibly separated in the disk array due

• Read and write to RV are allowed.

• Reflection of the difference between MV and RV at

• Updates to MV is reflected to RV.

• As a rule, read and write to RV are not allowed.

• Reflection of the difference between MV and RV at

• Updates to MV is reflected to RV.

• As a rule, read and write to RV are not allowed.

• Reflection of the difference between MV and RV at

• When Restore(update) is executed, the updated data of

• When Restore(protect) is executed, the updated data of

• As a rule, read and write to RV are not allowed.

• Reflection of the difference between MV and RV at

• When Restore(update) is executed, the updated data of

• When Restore(protect) is executed, the updated data of

• As a rule, read and write to RV are not allowed.

n

e

a

n

becomes zero after Separate

state occurs immediately after the pair setting.

to a copy fault.

Replicate has not been completed (The difference is being reflected from MV to RV).

Replicate has been completed.

Restore has not been completed (The difference is being reflected from RV to MV).

the MV is reflected into the RV.

the MV is not reflected into the RV.

Restore has been completed.

the MV is reflected into the RV.

the MV is not reflected into the RV and the Separate completion state is automatically entered.

d

y

S

y

S

y

n

c

h

r

o

n

o

u

s

o

u

s

r

o

n

o

r

o

n

16

2

.

4

2

.

4

2

.

4

If the activity between volumes with the pair setting is the Replicate or Restore state, you can change the

copy method of data between MV and RV according to the load status of the disk array. The state to

which a transition is made by the instruction to change the copy method is called the “copy control state“.

There are the following two types of copy between MV and RV in the Replicate or Restore state.

• Copy for reflecting difference

Copy to reflect the content of MV (RV for Restore) at Replicate or Restore to RV (MV for Restore).

• Copy to reflect updates in MV to RV

Copy to reflect updates in MV to RV after Replicate or Restore. However, if Restore(protect) is

executed, the updated data of the MV is not reflected into the RV.

You can change the copy method and state by changing the copy control state.

There are the following copy states in the copy control states as shown in Table 2-2.

Copy Control State Copy State

Foreground Copy

Background Copy -

Suspend

C

o

p

y

C

o

Synchronous Copy Mode

Semi-synchronous Copy Mode

Suspend

Suspend due to a failure

p

C

y

C

y

C

Table 2-2 Copy Control State

Chapter 2 Data Replication

o

n

t

r

o

l

S

t

a

t

e

o

n

t

r

o

l

S

o

n

t

r

o

• Copy for reflecting difference is performed.

• Updates to MV are reflected RV sequentially.

• Copy for reflecting difference is performed.

• I/O of updates to MV is completed when data is written to MV,

and the data is copied to RV immediately after that.

• Can be set for a RemoteDataReplication pair.

• Copy for reflecting difference is performed.

• I/O of updates to MV is completed when data is written to MV,

and the data is accumulated as difference information. For the accumulated difference, data is copied to RV asynchronously. The copy interval to RV (Background Copy Level) can be changed in units of disk arrays.

• Copy for reflecting difference is not performed.

• I/O of updates to MV is completed when data is written to MV,

and the data is accumulated as difference information. Reflection to RV is not performed.

• Forcefully suspended in the disk array due to a copy fault.

• Copy for reflecting difference is not performed.

• I/O of updates to MV is completed when data is written to MV,

and the data is accumulated as difference information. Reflection to RV is not performed.

t

l

S

t

a

t

e

t

e

17

Chapter 2 Data Replication

py

Copy control states can be specified when Replicate or Restore is performed. You can also change the

copy control state you specified at Replicate or Restore as required.

When Restore with RV protection specified is executed, only copy for reflecting the difference is

executed, thus the updated data of the MV is not reflected into the RV. Therefore, specifying or

changing a copy control state (Synchronous Copy Mode, Semi-synchronous Copy Mode, or

Background Copy) has no effect.

There are the following five instructions to change the copy control state.

• Synchronous Copy instruction

• Semi-synchronous Copy instruction

• Resume instruction

• Background Copy instruction

• Suspend instruction

The Resume instruction changes Background Copy or Suspend to preceding Foreground Copy

(Synchronous Copy Mode, Semi-synchronous Copy Mode).

Figure 2-6 shows the state transition diagram of the copy control states.

Background Copy

Background

Copy Instruction

Foreground Copy

Sync Copy

Mode

Synchronous Copy or

Semi-synchronous

Copy or Resume

Background Copy Instruction

Suspend Instruction

Semi-

synchronous

Mode*

Co

Suspend Instruction

Copy Fault

* Semi-synchronous Copy Mode

can be used only for a RemoteDataReplication pair

Synchronous Copy or Semi-synchronous Copy or Resume Instruction

Suspend

normal abnormal

Figure 2-6 State Transition Diagram of Copy Control State

18

Chapter 2 Data Replication

2

.

5

R

e

l

a

t

i

o

n

s

h

i

p

b

e

t

w

e

n

C

o

p

y

2

.

5

R

e

l

a

t

i

o

n

s

h

i

p

b

e

t

w

e

n

C

p

y

C

o

C

2

.

5

R

e

l

a

t

i

o

n

s

h

i

p

b

e

t

w

e

n

P

e

r

f

o

r

m

a

n

c

e

a

n

d

C

P

e

r

f

o

r

m

a

n

c

e

a

n

P

e

r

f

o

r

m

a

n

c

e

S

t

a

t

e

S

t

S

t

If the copy control state is set to the Synchronized Copy state, the difference between MV and RV is

not accumulated because updates to MV are immediately reflected to RV. However, the write time to

MV is longer because it waits for reflection of the updates to RV. If the Copy Control state is set to

the Suspend state, the difference between MV and RV is accumulated because updates to MV are not

reflected to RV, but the write time to MV is the same as normal I/O.

In this way, there are correlations between the amount of accumulated differences in MV and RV and

write performance in different copy control states. The correlation for each copy control state is

shown in Table 2-3.

Copy Control State Difference between MV and RV Write Overhead to MV

a

t

e

a

t

e

Table 2-3 Correlations in Different Copy Control States

d

a

n

d

C

o

p

o

y

n

t

r

o

l

t

r

o

l

r

o

l

Foreground (Synchronous) Small Exists

Foreground (Semi-synchronous)

Background

Suspend Large Does not exist

If the difference between MV and RV is large in the Replicate state, we recommend to select a copy

control state which gives the difference between MV and RV smaller in a system without sufficient

suspension time for Separate execution because the processing time of Separate increases. Also, we

recommend you to select a copy control state without write overhead to MV in a system where write

performance to MV must be maintained and improved.

19

Chapter 2 Data Replication

2

.

6

R

V

A

c

e

s

2

.

6

R

V

A

c

2

.

6

R

V

A

In the Replicate state and Restore state, MV is synchronized with RV to always match the volume data

between MV and RV. Since the difference between MV and RV is reflected in the disk array

independently of the operating system or file system, however, the volume of RV may become

inconsistent. Therefore, the data replication function cannot impose access restrictions for MV, and

MV can be referred to and updated at any time. The function can impose access restrictions for RV

for activity state to prevent malfunction.

Table 2-4 shows the states which can be specified as RV access restrictions.

Access Restriction Description

Read/Write (R/W) RV is enabled for read/write.

Read Only (RO) RV is enabled only for reading.

Not Ready (NR) RV is disabled for read/write.

RV is not recognized by the

Not Available (NA)

operating system or the LU (Logical Unit) is invalid.

e

c

e

Table 2-4 RV Access Restriction

s

R

e

s

t

r

i

c

t

i

o

n

t

r

i

c

t

r

i

c

t

rpl rst sep/exec separated

- -

∆

Note 2 ∆ Note 2 ∆ Note 2 ∆ Note 3

{ { {

∆

Note 4 ∆ Note 4 ∆ Note 4 ∆ Note 4

i

o

n

i

o

n

Activity State

∆

Note 1

{

-

{: Available ∆: Available with administrative restrictions -: Unavailable

rpl: Replicate state RW: Read/Write

rst: Restore state RO: Read Only

sep/exec: Separate execution NR: Not Ready

separated: Separated state NA: Not Available

Note 1: For Separate(immediate), “Read/Write (RW)” is set even during execution of Separate.

However, keep the following operational influence in mind:

1. Data is being copied from the MV into RV during Separate execution. Therefore, if the

I/O load on the RV is high, I/O performance on the MV side may lower.

Note 2: You can set “Read Only (RO)” to RV in the Replicate state, Restore state, or Separate

execution. In this case, note the following.

1. No update to MV should be done when RV is referred to in the Replicate state or Restore

state.

2. For updates to MV, I/O processing is done to the disk by the operating system control of

the file system. Even if the application has completed the update process to the disk, it

has not necessarily completed the update process to MV. Reflection of the update to RV

is processed in the disk array independent of the operating system.

Therefore, RV cannot be referred to normally because it is not consistent as a volume.

You can use it if consistency is assured in the specific operation.

20

Chapter 2 Data Replication

Note 3: If “Read Only (RO)” is set for RV for which Separate was completed, keep the following

operational notes in mind.

1. If NTFS is used as a file system, reference to the RV is disabled.

2. If FAT16/FAT32 is used as a file system, associate the file system with the drive by using

the mount command of the disk control operation commands or by starting [Disk

Management] (Windows).

3. If FAT16/FAT32 is used as a file system, an attempt to write to RV ends up with an error.

Therefore, do not use any application for automatically performing write operation for the

drive. Do not perform any operation (e.g., changing of a partition configuration) in

which write to RV is performed through [Disk Management] (Windows).

When mounting a file system, it is necessary to mount it by a read-only specification.

Note 4: The “Not Available (NA)” state has meaning when the VSS (Volume Shadow copy Service)

function is used. A transition to the Not Available state is automatically controlled by the

VSS function. The user need not perform the operation normally.

2

.

7

C

o

p

y

F

a

u

l

t

s

a

n

d

S

t

a

t

e

T

r

a

n

s

i

t

i

o

2

.

7

C

o

p

y

F

a

u

l

t

s

a

n

d

S

t

a

t

e

T

r

a

n

2

.

7

C

o

p

y

F

a

u

l

t

s

a

n

d

S

t

a

t

e

T

If copy operation between MV and RV is not performed normally due to a connection failure between

them, a transition to the following states may occur depending on the timing and type of the failure.

• Separate state due to a fault (failure separation)

Forcefully separated in the disk array due to a copy fault. The contents of MV and RV are

completely different.

To cancel the Separate state due to a failure, remove the cause of the copy fault and perform

restoration by using Replicate and Restore.

• Suspend state due to a fault (abnormal suspend)

Forcefully suspended in the disk array due to a copy fault in the Replicate or Restore state. Copy

between MV and RV is suspended.

To cancel the Suspend state due to a failure, change the copy control state as you do to cancel the

normal Suspend state after removing the cause of the copy fault (Refer to 2.4 “Copy Control State”).

s

r

a

n

s

n

i

t

i

o

n

i

t

i

o

n

s

21

Chapter 2 Data Replication

2

.

8

F

r

e

2

.

8

2

.

8

If the power to the disk array is turned off for maintenance, access to the disk array is disabled,

disallowing to continue copy operation for the paired volumes in the disk array. In this situation, the

data replication function for the disk array stops replication operations of the whole disk array.

Freezing of replication operations for the disk array due to power down of the disk array is called

freezing of the disk array and the state is called the freeze state of the disk array.

When a disk array is in the freeze state, replication operations between volumes to which the pair

setting with the volume in the disk array is specified are also suspended, and the copy control state of

the pair becomes the freeze state.

When the pair is in the freeze state, no new replication operation can be performed.

For pairs in the freeze state, note the following.

• Freeze in the Separate state

If the pair goes into the freeze state in Separate execution, it transits to the Separate state due to a fault

(failure separation) when the following operation is performed.

In this case, to use RV after the freeze state is cancelled, you must perform Replicate again to copy

data, and perform Separate.

• Freeze in the Replicate state

If the pair goes into the freeze state in the Replicate state, copy operation between MV and RV is

suspended. Copy operation is automatically resumed when the freeze state is cancelled.

• Freeze in the Restore state

If the pair goes into the freeze state in the Restore state, I/O terminates abnormally when reference or

update is performed to the area where copy from RV to MV has not been completed.

If the pair goes into the freeze state in the restored state, copy operation between MV and RV is

suspended.

Copy operation is automatically resumed when the freeze state is cancelled.

F

e

r

e

r

e

z

e

o

f

D

i

s

k

A

r

a

y

s

o

f

D

i

s

k

A

r

o

f

D

i

s

k

A

a

r

a

y

s

22

NEC IS015-9E User Manual

Specifications and Main Features

Frequently Asked Questions

User Manual

Preface

Contents

Chapter 1 Data Replication Overview

Backup

Chapter 2 Data Replication

Copy Control State Copy State