Restoring the backup that was taken through fdbbackup results in database lock

Using FoundationDB
1

I am running foundationDB on Kubernetes using the foundationDB operator and took backup using fdbbackup command. When the backup is successful, I try to restore the backup using fdbrestore command and in output I get that restore is successful. But when I go ahead try to get a key I get below error

ERROR: Database is locked (1038)

All these backup and restore commands are being executed through a process that EXECs into the pod and then run those command. And I am EXECing into the pod manually to check if the entries are restored as expected.

2

The fdbrestore start command only starts a restore. It does not wait for the restore to complete unless you add the -w option. During restore, your cluster will be locked, but it will be unlocked once the restore is complete. Check the status of your cluster’s restore process with fdbrestore status on your cluster. If your restore still has not completed, it is likely due to the restore workers (which are also the backup_agents) being unable to read some or all of the backup data. Note that the backup folder must be accessible from ALL hosts running the backup_agent process.

3

Hi @SteavedHams
Thanks for the response, I was actually going to update this topic with how I was able to resolve this.
So what I was trying to do is running the backup_agent, taking the backup, stop the backup_agent and then restore the backup; in that case the restore was successful (weird) and I was getting Database locked error but when I don’t stop the backup_agent and restore the backup I was able to restore that successfully.
I am actually running this on Kubernetes and we dont have any mechanism for backup and restore if the cluster is deployed on Kubernetes, thats why was experiencing some weird issues.