I am running into an issue and now sure where to start looking. My fleet server, which is running on an EC2 keeps dying frequently. It normally comes up (not always) after a couple of restart but dies at a later point. Sometimes the webpage give 502 even when the service status is up and running.

502 is not an error returned by Fleet, so perhaps an LB health check related issue?

@zwass i missed your reply. There's no load balancer. I sometimes get 502 but mostly get server timeout. Earlier it used to happen once in a while. But now it's so frequent that the instance is unusable. I thought it was resource related issue on EC2 so i increased the ulimit but that only fixed the issue for a while.

How's the load on the MySQL server?