Running databases with a good Disaster/Recovery strategy is a hot topic today and MariaDB is not exception to this. Things get more complicated when you need to think about recovery without any data loss.
- How do you do that and are you sure you can recover to any point in time when your productive MariaDB cluster deployment crashes and there is no other option than a restore?
- What about human errors such as accidental data deletion or data corruption?
- Are you prepared for such a situation and did you test it?
- What if we could use Docker containers for that purpose?
Having the production MariaDB Galera cluster running as usual and then use one or more delayed replicas in Docker containers can help to recover quickly
and online from disasters or human errors without the requirement for a full restore.
In this session we will have a look at the implementation details and then go through a demo to actually proof that the concept is working as expected
and greatly simplifies the various recovery scenarios?