Veritas Storage Foundation 5.1 SP1: Storage and Availability Management for Oracle (5900-1504, April 2011)

6
List checkpoints using dbed_ckptdisplay and rollback using
dbed_ckptrollback into appropriate checkpoint.
7
After rollback, start Oracle and let it recover. This procedure assumes archive
logs are also replicated and available for oracle recovery.
Creating new Database Storage Checkpoints at the remote replicated
site
To make the remote site productive and useful, some customers use the remote
site for reporting or backup. This is usually done by stopping replication and
running the backup or report, shutting down the database, and restarting
replication.
When you are using Database Storage Checkpoint to clone the database and when
you are running the cloned database, you must stop replication. To minimize the
down time for replication, Symantec recommends removing the cloned database
as soon as you complete your reporting or backup tasks and unmount both the
checkpoint file system and the database file system.
For this procedure the SFDB repository need not be part of the replicated volume
since it does not reuse any information from the primary site's SFDB repository.
To create a checkpoint at the remote site
1
Stop VVR replication.
2
Run fsck for all database volumes on the replicated site.
3
Mount the database file systems.
4
Startup Oracle and recover the database.
5
Create the SFDB repository using dbed_update.
Note: You do not need to use the -o rename option because since the SFDB
repository information from the primary site is not reused.
6
Create a checkpoint using dbed_ckptcreate.
7
Clone the database using the newly created checkpoint.
8
Run your reporting or backup tasks.
9
Shutdown the cloned database.
10
Unmount and remove the checkpoint if it is no longer required.
287Deploying Storage Foundation for Databases (SFDB) tools in a replicated environment
Using Storage Foundation for Databases (SFDB) tools in a VVR environment