It sounds very much your server got under heavy load that time since the ping timed out for 60 seconds. Did both instances of the Wrapper time out same fashion?

The Wrapper requested on both a shutdown. Instance 1 was shutting down in timely manner and got restarted. Whereas instance 2 didn't and got killed forcibly.
What is the shutdown timeout (wrapper.shutdown.timeout)? It might help to extend the timeout for instance 2 in case this application needs more time for shutdown.

Hope this information helps you out.

Best Regards,