During a storage resize on a Flex Appliance application instance, the Activity Monitor shows a Failed status after one hour

Problem

If you resize the storage on an application instance to a smaller size (shrink the volume), the process can take an extended amount of time to complete. The amount of time depends on how much storage needs to be resized, how full the storage is, and the overall workload on the appliance.

If the storage resize takes longer than an hour, the Activity Monitor times out and the resize status shows as Failed. However, the resize operation continues running in the background and may still complete after the timeout. If it is ultimately successful, the status changes to Successful in the Activity Monitor.

This issue is much less likely to occur in version 1.3 and 1.3.1 due to an increase of the timeout to two weeks. If the resize takes longer than two weeks you could still encounter this issue.

If the container is started while the resize is in process, it may result in data loss.

Error Message

The task status in the Activity Monitor shows as Failed.

Solution

If you see a Failed status after an hour, do not restart the resize operation and do not start the instance. Monitor the status and if it has not changed to Successful within 12 hours, contact Veritas Technical Support for assistance. Ask your representative to reference this article.

To check the current status of the resize operation, log in to the appliance as the hostadmin user and run the following command in the Veritas Appliance Shell

     > support shell

     $ ps -ef | grep fsadm

IMPORTANT: If you have an HA Flex cluster, you must check both nodes as the fsadm process will only run on one of the nodes. 

If the fsadm process is running, the storage resize operation is still in progress.

If the fsadm process is not running, the Failed status is correct and requires additional troubleshooting.

Contact the Avengers team with any questions about this article.
<DL-VTAS-AS-Team-Avengers@veritas.com>