Saving chunk status on the secondary database server

For high-availability cluster servers, if the status of a chunk (down or online) is changed on the secondary database server, and that secondary server is restarted before a checkpoint is completed, the updated chunk status is not saved.

To ensure that the new chunk status is flushed to the reserved pages on the secondary database server, force a checkpoint on the primary database server and verify that a checkpoint also completes on the secondary database server. The new chunk status is now retained even if the secondary database server is restarted.

If the primary chunk on the secondary database server is down, you can recover the primary chunk from the mirror chunk.

To recover the primary chunk from the mirror chunk:

  1. Run onspaces -s on the secondary database server to bring the primary chunk online.
  2. Run onmode -c on the primary database server to force a checkpoint.
  3. Run onmode -m on the primary database server to verify that a checkpoint was completed.
  4. Run onmode -m on the secondary database server to verify that a checkpoint was also completed on the secondary database server.

After you complete these steps, the primary chunk is online when you restart the secondary database server.


Copyright© 2019 HCL Technologies Limited