I've got a server to host a basic Rails app ( < 2MB in total size). The server is running Apache 2.2.14 and Passenger 2.2.7 Testing on my web server goes like this: - Apache stopped: 8% memory allocated (of 1GB) - Apache started: 51% memory now allocated, ~400MB gets allocated after starting apache - Visit my server to look at my app: memory allocation goes from 51% to 90% from just that one request, causes a 10 second delay on the first page load, spikes CPU use up to 100% for a number of seconds, and memory allocation sits at 90%. - The site is then really responsive and runs fine. Once passenger processes start to time out due to inactivity memory allocation goes back down to 51%. - Re-visit app after an hour or so: Another large sustained spike in memory use, short burst of CPU use Is this normal? I've tried to find some good resources on Passenger + Apache and memory use yet nothing really seems to explain both why passenger sucks up every drop of available memory and how to limit the memory passenger consumes.
on 2012-08-20 17:26