How can issues on a single server affect all the service? Is this server (db1040) a SPOF? If yes, why?
The server started lagging due to an ongoing schema change; however, that schema change, to avoid issues, only happens on a server at a time.
This is a 'vslow' slave, but that should have little user impact (except for statistics)- as far as I know, job queue is run on all slaves, and only this one was affected by lag.