I have a question about using storage replication for a MySQL Server running on LXC container and an HA active for that container.
Does anybody knows if storage replication does a consistent image (fs freeze before doing snapshot) or not?
In case of HA activating a migration to another node for the VM the mysql data volume might be in a inconsistent state.
Because using pve_zsync to replicate a mysql server LXC container to a another cluster makes the mysql data corrupted and mysql cannot recover and boot successfully.
Does anybody has a similar scenario and feedback to share?
At the moment I'm using a replica to maintain a remote mysql container instance ready in case of failure of a node to avoid risks.
By the way a backup and restore of a mysql server lxc container is able to book without problems.
Thank you for any help.
Does anybody knows if storage replication does a consistent image (fs freeze before doing snapshot) or not?
In case of HA activating a migration to another node for the VM the mysql data volume might be in a inconsistent state.
Because using pve_zsync to replicate a mysql server LXC container to a another cluster makes the mysql data corrupted and mysql cannot recover and boot successfully.
Does anybody has a similar scenario and feedback to share?
At the moment I'm using a replica to maintain a remote mysql container instance ready in case of failure of a node to avoid risks.
By the way a backup and restore of a mysql server lxc container is able to book without problems.
Thank you for any help.