HP StorageWorks Storage Mirroring for Linux User's Guide (T2558-96078, February 2008)

ManualsBrandsHP ManualsSoftwareHP Storage Mirroring Software Windows Media Kit

5 - 3

Replication Set Operations

The following replication set operations are available:

 Create—You can create replication sets through the Management Console or the text clients by selecting icons through

the graphical interface or by specifying explicit paths through any of the clients.

 Modify—Replication set modifications are a dynamic process. While Storage Mirroring is replicating, changes can be made

to a replication set that is in use. Once the changes have been saved, the new settings will become effective immediately.

If new folders or files are added to a set that is being replicated, a remirror is necessary to synchronize the source and

target.

In certain circumstances, the Management Console will display a window stating that a modified replication set needs to

be disconnected and reconnected after a save has been completed. This is to allow Storage Mirroring to read the new

data that was added to the replication set.

Modifications in the Management Console consist of clearing and selecting alternate volumes, directories, and/or files

while modifications in the text clients consist of deleting an existing rule and adding a new one.

 Rename/Copy—You can rename or copy a replication set in the Management Console only. The text clients require

you to create a new replication set. When renaming an unsaved replication set, you supply a new name and the original

name is replaced. If the replication set has already been saved, a copy is made; you supply a new name and the original is

left intact.

 Save—A replication set must be saved before establishing a source/target connection and before exiting any of the

Storage Mirroring clients so that it will be available in future client sessions.

 Delete—If a replication set is not currently connected to a target, you can delete that replication set. After a

source/target connection has been established, a replication set cannot be deleted until the source and target have been

disconnected.

 Display/List—When using the text clients, you can display all of the replication sets associated with a specified source or

you can list the rules that are associated with a specified replication set.

 Resync/Revert—You can restore previously saved settings. In this instance, a previously saved replication set where

changes were made, but are no longer wanted, can be discarded without losing the replication set altogether. You can

revert back to the replication set that was saved before making the changes.

How Replication Sets Work

To better understand how replication sets work, the following diagram identifies the steps involved with creating and saving

a replication set as well as what happens to the replication set information after a source and target have been connected.

Each numbered step is described following the diagram.

1. The Storage Mirroring client manages all replication set modifications. The administrator creates, modifies, or deletes

replication set rules or the replication set itself from the Storage Mirroring client machine.

2. The first time a replication set is saved, a replication set database is created on the source machine. This file contains all

of the replication sets that are established for that machine.

3. When the Storage Mirroring client manages any replication set activity after the initial creation of the replication set

database, the source machine sends the entire file to the client machine.

4. When the administrator modifies a replication set, the Storage Mirroring client manages that activity and modifies the

local copy of the replication set database.

5. When the replication set is saved, the file is sent to the source machine.

6. When a source/target connection is established, a portion of the replication set database is sent to the target machine.

The portion contains only the data that is relevant to the current connection; it does not contain replication set data

that is relevant to other connections. The file on the target machine is identical to the source machine’s name so that the

file can be recovered if the source machine crashes. It is called

source_machine_name.db.

Source

Target

SourceMachine.db

Replication Set 1

Replication Set 2

Replication Set 3