16.6 How to repair the DFS Replica service?

The repair process rebuilds the DFS Replica database. Upon completion, the current active database is replaced with the repaired one. If there are two replica sites, the replica automatically synchronizes with the active repaired database. Until the repair database is activated, all DFS Replica requests (except those specifically referencing the repaired database) operate against the existing database. Thus, clients can access DFS junctions even during repair for volumes that still have correct entries in the DFS Replica database.

  1. Browse and select the DFS management context you want to manage. This displays the replica sites.

  2. Select a replica site, then select Repair DFS Replica.

  3. Select one of the following repair levels, then click OK:

    • Replace with the last saved copy: Restores the last saved copy of the database using the automatically created backup file.

    • Copy from another replica site: Retrieves a copy of the database from another server that is currently running the DFS Replica service.

      This option is available only if there is more that one replica site.

    • Rebuild from the eDirectory tree: Rebuilds the database from scratch by recursively scanning the eDirectory tree down from the management context container and recording volume object information in the repaired database. This is a time-consuming activity and should be considered carefully.

  4. Click Confirm. Monitor the status of the rebuild periodically until it completes. This duration can vary from a few minutes to several days depending on the selected repair level. To view the progress, select the replica site, and then Details.

    During the repair process, the status displays as Repairing. If the option Rebuild from the eDirectory tree is selected, upon completion of the repair, DFS automatically reloads the DFS Replica service on the replica server and activates the database, changing the state to Running. If there is a second replica site, its database copy is automatically synchronized with the repaired database.

  5. If any errors occur during the repair, refer to the following log file:

    /var/opt/novell/log/dfs/vlrpr.log