Database specialists should be on-call for database related problems
Currently database specialists are in a bit of a weird position where they are focusing on DB work but also (at times) on call for non DB issues (e.g. NFS burning down). While at times this is necessary (e.g. to fill gaps), I would like for DB specialists to not be the first responders for everything unless otherwise necessary. The "flow" that I envision is basically this:
- A production engineer gets paged
- The engineer determines the problem is database related
- The engineer escalates the problem to a database specialist
- The two (or more) engineers work together to resolve the problem
Instead of the current:
- A DB specialist gets paged because NFS is down
- The specialist has no clue what to do because NFS is not their forte
- The DB specialist has to hope the runbooks cover enough, or escalate the issue to a production engineer anyway
This does mean we'll have a DB specialist on call at all times (though not as the first responder), which in turn means we should try to find a DB specialist in the US since so far we're all in the EU.