HP Application Recovery Manager software A.06.10 Concepts guide (March 2008)

Creating a replica
1. A replica is created.
2. The application host is resumed. Any dismounted filesystem is remounted.
In the case of an offline backup, the database can be put back online and normal
operation started again.
In the case of an online backup, transaction log files and cached information
from the replica creation period are applied to the database.
3. The backup host environment is prepared for the replica’s disks and data. New
devices are detected by scanning. Any volume groups are imported and
activated. Filesystems are mounted.
Replicating the data objects
With the database/filesystem in the required state, the Disk Array Agents on the
application system and the backup system are triggered to perform the replication.
The two Disk Array Agents act as a pair:
• On the application system, the agent resolves the specified data to the volumes
containing it.
• On the backup system, the agent allocates the volumes required for the replica.
The array then creates the replica on the array.
The method of replication depends on the type of disk array being used, whether
LVM mirroring is required, and so on. For information on how split mirror and snapshot
replication is performed, see Chapter 2 on page 21.
Mounting the replica on the backup system
Application Recovery Manager creates mount points on the backup system and
mounts the filesystems on the replica to them. The mount paths depend on whether
an application or filesystem backup is being performed and the options you select
in the GUI. After a ZDB session is complete, you can choose to leave the filesystems
mounted on the selected paths.
With the VSS integration, the replicas are by default not mounted.
Concepts guide 53