Administration Guide › Set Up Replication › Multiwrite Replication with DISP Recovery (Multiwrite-DISP) › Transaction Log and Data Recovery › How to Perform a Disaster Recovery

How to Perform a Disaster Recovery

If a disaster happened to your CA Directory implementation, you can no longer rely on automatic recovery from a transaction file (single server) or from a peer DSA (multiwrite-DISP). Instead, you need to recover your data from a backup.

To recover your data in the case of a disaster, do one of the following.

• For a single server:
  1. Locate your last backup copy.
  2. Delete the transaction log (.tx file).
  3. Rename the backed up files by removing the prefix z from the filename extension. The file should now be named server.db.
  4. Restart the DSA.
• For multiple peers:

  Note: The steps below assume that you have multiple peers and peer B was unavailable for a long time. The procedure for adding a new peer is very similar.

  1. Run the following command on all hosts with peers, including host B:

     dxdisp B
     

     This sets the last update time to now (T).

  2. Remove the transaction log file on B, if it exists.
  3. Take a snapshot copy of any one of the running DSAs.

     The online dump creates a backup file: *.zdb

  4. Copy the backup file to B.
  5. Rename the copied file by removing the z file extension prefix.

     You should now have: server.db

  6. Start B

     The DSA peers send all changes since T to B to bring it in sync.

More information:

Add a DSA to a Multiwrite-DISP System

dump dxgrid-db Command—Take a Consistent Snapshot Copy of a Datastore

Online Datastore Dump