The MySQL (RDS) server on mtpremiumus04 stack crashed. Automatic failover was triggered and service resumed to normal after a minute. Operation has engaged with AWS to investigate the root cause.
AWS support answer:
At 2019-01-13 20:22:38 UTC our internal monitoring system detected that the underlying host of your primary instance was not healthy and therefore initiated a failover to the secondary.
It has replaced the underlying host that was unhealthy.