The Veritas Volume Replicator (VVR) Storage Replicator Log (SRL) stores blocks until they are committed to the replicated volume on the secondary site. Since VVR replicates data as blocks, instead of files, the SRL must also store data as blocks. The SRL contains no filesystem and is seen by the Operating System as a "raw" volume. In the event of a disk read or write error, the SRL may be corrupted as it lacks the journaling capabilities of a filesystem.

To recover from a corrupt storage replicator log (SRL), dissociate and then associate it from the replicated volume group (RVG). The SRL may need to be re-associated on either the primary or the secondary, or both, depending on which site contains corrupt SRL. Use vxprint to determine the status of the SRL. Information on using vxprint is found here: 

 

https://www.veritas.com/support/en_US/article.000088736

The steps to recover from SRL corruption depend on the version of Storage Foundation for Windows (SFW) and whether the following patch is installed:

 

https://www.veritas.com/support/en_US/article.TECH63945

SECTION I: For SFW 5.1, or earlier versions with the patch discussed in article 311479 already installed.
SECTION II: For SFW 5.0 and earlier without the patch discussed in article 311479 already installed.


SECTION I: For SFW 5.1, or earlier versions with the patch mentioned in article 311479 already installed.

Recovering from primary SRL corruption
 

Note: Dissociating and associating an SRL that is located on the primary site triggers a full resynchronization of the replicated volumes. This is unavoidable if the primary SRL reports corruption.
 

These actions must be taken from an instance of Veritas Enterprise Administrator (VEA) that is connected locally on the actual server. This cannot be performed from an instance of VEA that is remotely connected to another server.
 

1. From Replication Network, expand the Replicated Data Set (RDS).
 

2. Expand the RVG.
 

3. Right-click on the SRL and select Dissociate Replicator Log.
 

4. Right-click on the RVG and select Associate Replicator Log.
 

5. Select the SRL.
 

6. Start replication.
 

 

Recovering from secondary SRL corruption

Note: The following steps do not result in a full resynchronization. Only follow these steps if SFW 5.1 is installed or if the patch that is discussed in article 311479 was already installed prior to the SRL corruption.
 

1. From Replication Network, expand the Replicated Data Set
 

2. Right-click the Replicated Volume Group and select Disable Data Access.
 

3. Run the following command from the secondary site:
 

vxrlink det
 

Note: Use vxprint -VPl to determine the secondary rlink name.
 

4. Expand the RVG.
 

5. Right-click the SRL and select Dissociate Replicator log.
 

6. Right on the RVG and select Associate Replicator Log.
 

7. Select the SRL volume.
 

8. Run the following command from the secondary site:
 

vxrlink -f att (secondary_rlink_name>
 

9. Right-click on the RVG and select Enable Data Access.

SECTION II: For SFW 5.0 and earier without the patch mentioned in article 311479 already installed.
 

Note: For Section II, the steps are the same for both the primary and secondary sites. This results in a full resynchronization of the replicated volumes.
 

These actions must be taken from an instance of Veritas Enterprise Administrator (VEA) that is connected locally on the actual server. This cannot be performed from an instance of VEA that is remotely connected to another server.
 

1. From Replication Network, expand the Replicated Data Set (RDS).
 

2. Expand the RVG.
 

3. Right-click on the SRL and select Dissociate Replicator Log.
 

4. Right-click on the RVG and select Associate Replicator Log.
 

5. Select the SRL.
 

6. Start replication.