Repairing storage pools from container-copy storage pool volumes after a disaster

If a disaster occurs on a source server, you can repair deduplicated data extents in a directory-container storage pool from offsite container-copy storage pool tape volumes. The directory-container storage pool is repaired on a target server at a recovery site.

Before you begin

Complete the following steps:
  1. Back up the IBM Spectrum Protect server database by using one of the following methods:
    • On the Operations Center Overviews page, click Servers, select a server, and click Back Up.
    • Issue the administrative command, BACKUP DB.
  2. Review the latest information about repairing and recovering data in technote 2013682.
  3. To plan the next steps, review the following restrictions about using the AUDIT CONTAINER command.
    Attention:
    • If you issue the AUDIT CONTAINER command with the ACTION=MARKDAMAGED setting for an entire storage pool, referenced data is unavailable for restore operations until the storage pool is repaired. Depending on the database size, network bandwidth, media speed, and other factors, the REPAIR STGPOOL command might run for hours or days. For this reason, if some of the data in the storage pool is available, or the status of data in the storage pool is unknown, follow these guidelines:
      1. Consider running the AUDIT CONTAINER command with the ACTION=SCANALL setting first. The ACTION=SCANALL setting identifies database records that refer to data extents with inconsistencies. Only those data extents are marked as damaged in the database.
      2. After the extents are marked as damaged, you can run the REPAIR STGPOOL command.
    • If you plan to run the AUDIT CONTAINER command with the ACTION=REMOVEDAMAGED setting, follow these guidelines:
      1. Considering running the QUERY DAMAGED command first to determine the scope of damaged data extents in the storage pool.
      2. After that, you can run the REPAIR STGPOOL command to repair damaged extents in the storage pool.
      3. Finally, you can run the AUDIT CONTAINER command with the ACTION=REMOVEDAMAGED setting to remove any damaged data extents that remain in the storage pool.

About this task

Use the procedure to repair the following types of major damage:
  • Complete loss of all container storage pools on the source server
  • Complete loss of the primary site
The following assumptions are made for this disaster recovery scenario:
  • You were using the PROTECT STGPOOL command to back up data to offsite container-copy storage pools from a source server. You retrieved the offsite tape volumes and have them at your recovery site.
  • You were not using the PROTECT STGPOOL command to back up data to a target replication server.
  • You used the IBM Spectrum Protect Blueprints to configure the IBM Spectrum Protect source server, and you also used the Blueprint configuration scripts to restore the environment by setting up a new target server at a recovery site. The scripts copied backup versions of the IBM Spectrum Protect database, the server options file (dsmserv.opt), the volume history file (volhist.out), and device configuration file (devconfig.out) to their original locations on the recovery server. After the scripts run, you see the newly created, empty directories on the recovery server.
When you attempt to repair a directory-container storage pool from container-copy pools, the REPAIR STGPOOL command fails if any of the following conditions occur:
  • The container-copy storage pool is unavailable.
  • The container-copy storage pool is damaged.
  • The container-copy storage pool volumes are unavailable or damaged.

Procedure

  1. Mark all data extents in the container storage pool as damaged by issuing the AUDIT CONTAINER command for the container storage pool at the storage pool level, and specifying the ACTION=MARKDAMAGED parameter.
    For example, to audit a storage pool that is named STGPOOL1 and mark it as damaged, issue the following command:
    audit container stgpool=stgpool1 action=markdamaged
  2. If you protected the directory-container storage pool by using both onsite and offsite container-copy storage pools, issue the UPDATE STGPOOL command for the onsite copy of the container-copy storage pools, and specify the ACCESS=UNAVAILABLE parameter.
  3. When the offsite container-copy storage pool volumes are back onsite, check them into the library by issuing the CHECKIN LIBVOLUME command and specifying the STATUS=PRIVATE parameter.
  4. Update the status of the volumes by issuing the UPDATE STGPOOL command and specifying the ACCESS=READWRITE parameter.
  5. Repair the storage pool by issuing the REPAIR STGPOOL command and specifying the SRCLOCATION=LOCAL parameter.
    For example, to repair a storage pool that is named STGPOOL1 from offsite container-copy storage pools, issue the following command:
    repair stgpool stgpool1 srclocation=local

    When you issue the REPAIR STGPOOL command, the damaged extents are deleted from the volume immediately after they are repaired. The damaged extents are not retained according to the value specified by the REUSEDELAY parameter.

  6. Confirm that there are no additional damaged extents by issuing the QUERY DAMAGED command.
  7. Repeat this procedure to repair all of your storage pools.