We just experienced what appears to be a random failover of a Postgres CloudSQL instance. In looking at the logs, nothing stands out: the instance appeared to be operating correctly via its heartbeat checs, there were no sudden spikes of CPU or memory usage, and there weren't any deadlocks, but an automatic failover happened.
Is there anything that we can look for in the logs to determine what the root cause of the failover was? Has something like this happened to anybody else?
LinkedIn
Twitter
