I am testing the backup/restore process for the mongodb operator.
And this is just a theoretical question:
if the cluster where the backup is created is not accessible does this means that I cannot use the backup that is already present on s3 ?
In order to restore the backup first you need to list the backups on the cluster where they are taken:
kubectl get psmdb-backup
Obviously, if the cluster is not accessible, or this command does not list the backups due to disk corruption. Or it is possible somehow to use it on another cluster when the original cluster can’t be accessed ?
You can also restore directly from s3 using backupSource field in psmdb-restore object. That way you don’t need a corresponding psmdb-backup. Please see Providing Backups
Is it possible to restore a backup of the not Kubernetes-based cluster on the Kubernetes-based cluster?
It is interesting from the point of migrating the existing Mongodb cluster to Kubernetes.
@kamuna if you use PBM on both sides, it should be possible. Please let me know if you got any problems, so I can assist you. You can even write a blog post for percona.community website!