Comment by Nextgrid

I've had an RDS instance fail in an equally-weird way that required manual AWS operator intervention 24 hours later, until which we were effectively locked out of our data and had to restore from a recent backup & rebuild the missing data from logs to bring the service back up in the meantime.

All the supposed "savings" of using managed services to save on staff costs evaporated immediately. No refund from the provider obviously despite it being an edge-case in their implementation.