Replication went to stale state

[Niveda]

Hi ,

We are running replication between 2 source arrays and one secondary array of the same model. The replication between them is periodic and I have a set of 5 volumes in a group. They went to stale state without any reason being found. I manually started them again and now I see two snapshots created for each volume. Is this fine? Also one of the volume in the group before going stale didnt get the snap created fyi. and was still syncing.

[sbhat09]

Hello @Niveda,

They would have went to stale state due to any glitch in replication network. The volume still 'syncing' could be due to the full sync that is going on for that volume. That will take some long time. In 'showrcopy -d' command output, you can see the sync progress. If they are making progress, you can wait for completion. No worries.

Regarding the snapshots - If the snapshot creation time is matching with the most recent sync period, it is created by the remote copy. You can also confirm it by it's naming convention. In periodic Remote copy, the previous snapshot will be removed automatically when a new snapshot creates at every sync interval. It is normal to have remote copy snapshots at any point of time.
If the snapshot timestamp matches with the time when the volumes went to stale state, it is the system auto-created snapshot to protect volumes. If the volumes are out of stale state, you can delete them.

If you need further clarification, please provide me the 'showrcopy -d' outputs from both primary and secondary.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Mahesh202]

Hi Niveda

It's not uncommon for volumes to go into a stale state during replication, and restarting the replication manually is often necessary to get them back in sync. As for the two snapshots created for each volume, it depends on the replication setup and configuration. In some cases, creating multiple snapshots can be intentional to ensure data consistency during replication. However, it's important to review the replication configuration to confirm whether the creation of multiple snapshots is expected.

Regarding the volume that didn't get the snap created before going stale, it's possible that the replication for that volume was delayed or experiencing issues that prevented the creation of the snapshot. It's recommended to investigate the replication logs to determine the root cause of the issue and ensure that the replication is configured correctly.

In summary, the creation of multiple snapshots for each volume and the issue with the volume that didn't get the snap created before going stale requires a closer look at the replication configuration and logs to confirm whether it's expected behavior or an issue that needs to be addressed.

Hope this helps.!!

Regards
Mahesh.