I have an Amazon EC2 instance that I use for a variety of things such as hosting an SQL server and cronning various jobs I believe are mildly CPU intensive but not memory intensive (this is a hunch and I've not actually profiled this).
I've noticed that on average once a week my instance becomes non-responsive (port 22 is unreachable, other ports that I've exposed are sometimes but not ALWAYS unreachable).
Most questions that I've come across to be similar to my issue speak of restarting the instance not solving their problem, but I'd like to know how to go about diagnosing why this happens and how to reduce these outages. I'm relatively new to using the AWS ecosystem so maybe I've missed something basic that's not showing up in the other questions
EDIT: I forgot to mention that the downtime also happens when the cronned processes aren't running.