I'm new to setting up a production server with Rails, so I tried to do what seems to be the standard setup right now: a couple mongrel servers behind Apache 2.2. I've got a test page that refreshes itself every couple of seconds, and I've noticed that I get a "Proxy Error" when that's run for about a minute or so, from what appears to be a mongrel server dieing a horrible death. If I refresh, the other one will shortly follow, forcing me to manually start the cluster again. The odd thing, is that everything runs flawlessly if I'm in "development" in my mongrel_cluster.yml file. My setup is totally standard, so I was hoping that someone would know what the difference between these two environments could be that is actually killing the server. Thanks for any help!
on 2007-03-05 23:35
on 2007-03-06 00:34
Okay, nevermind, I figured it out. It looks like an uncaught exception crashes Mongrel when it's in production. Odd... Oh well, I can live with it.