Updating the RASD

The RASD is usually updated automatically during normal replication activity. The following procedure to force an update of the RASD should only be used with the recommendation of Sybase Technical Support when the RASD is suspected of being corrupt.

After the data in the RASD is created by initializing the Mirror Replication Agent instance, the only way to update the RASD is to re-initialize the instance using the ra_init command with the force keyword.

Before you re-initialize the Mirror Replication Agent, the primary database must be quiesced.

To update the RASD

Use the following command to determine the state of the Mirror Replication Agent instance:

ra_status

If the Mirror Replication Agent is in Admin state, skip this step and go to step 4.

If the Mirror Replication Agent is in Replicating state:

Use the following command to suspend replication by the Mirror Replication Agent instance:
```
suspend
```
Use the following command to verify that the Mirror Replication Agent is in Admin state:
```
ra_status
```

After you verify that the Mirror Replication Agent is in Admin state, use the following command to initialize the primary database:

pdb_init

WARNING! Do not use the move_truncpt option when you re-initialize a primary database that was previously initialized by the Mirror Replication Agent.

If you do not preserve the existing secondary truncation point in the primary database, you will have to re-materialize the standby database before you can resume replication.

Before you re-initialize the RASD, you must quiesce the primary database, or otherwise prevent any DDL operations that can change the database objects or schema.

Log in to the primary data server, using a user login that has appropriate permissions or authority, and then quiesce the primary database (or execute the commands that will prevent any DDL operations that change the database objects or schema).

Use the following command to re-initialize the Mirror Replication Agent and force it to update its RASD:

ra_init force

The ra_init force command does not overwrite any existing log device path in the RASD, if all of the following log device information in the RASD matches that returned by the primary data server:

For ASE:

Database name
Device ID
Device name
Device path

For Oracle:

Redo group number
Redo name
Redo path

For each log device identified in the RASD, if any item in the preceding list does not match the information returned by the primary data server, ra_init force overwrites the RASD record for that log device with the information returned by the primary data server.

After the Mirror Replication Agent is initialized, you must release the quiesce on the primary database to restore normal client access to the database.

Log in to the primary data server, using a user login that has appropriate permissions or authority, and then release the quiesce on the primary database (or execute the commands necessary to restore normal client access to the database).

Resume replication in the Mirror Replication Agent:

resume

Verify that the Mirror Replication Agent is in Replicating state:

ra_status

If the Mirror Replication Agent does not return to Replicating state, see Chapter 4, “Troubleshooting Mirror Replication Agent,” for more information.