Missing Available Restore Period

[arosas]

Case # 05075915
While I wait for support to get back to me, wondering if anyone has had a SQL DB stop giving the Available Restore Period on a single DB, even with logs successfully backing up every 20 minutes.
The DB failed an integrity check on 10/12 from the SQL side, the DBA has fixed the integrity issue and the DB is now fine.
Since this failed integrity check on 10/12, the backup from 10/13 and onwards no longer gives us point in time restore. The DB continues to backup successfully and logs continue to back up successfully.

The Veeam GUI shows the recovery model as Full but gives the message "Not Available" under the Available Restore Period section.

[foggy]

Hi Tony, doesn't look expected, please continue working with our support engineers.

[arosas]

I wanted to add more information and context to see if anyone can provide any further info.
I missed mentioning that this issue is with databases that are in an availability group and we have found this is what's causing it.
When the DB failed the integrity check the DB was removed from the availability group to perform this maintenance and was going to be left out of the availability group for the time being.
If we remove a DB from an availability group, Veeam is expecting the DB to be there, therefore the point in time is no longer available until the DN is added back. We have added the DB back and we now have the point in time restore which to me sounds like this should not be the case and maybe be either an oversight or a bug.

Let’s use an example:
I have 1 availability group with 10 DBs
I remove 3 out of the 10 DBs in the availability group because they no longer meet requirements for those DBs to be failed over.
The availability group will still exist with 7 out of the 10 DBS, but now the 3 DBs no longer have point in time restore capabilities?

[foggy]

In this case, I believe, the behavior is expected. Most likely, logs were collected from another node while the DB was out of the AG.

[arosas]

This is not possible for a couple reasons.
1. We have our failover set to manual so the secondary node cannot have ownership of the DB or Logs.
2. We also drop the secondary DB after removal from the AG.

https://docs.microsoft.com/en-us/sql/da ... rver-ver15

[foggy]

Then I recommend continuing the investigation with the help of our engineers.

[arosas]

Thanks for the reply. I'll post back here what they find in case it helps anyone else out.