Hey There,
I manage a couple of rails production environments, both monitoring
app servers (one thin one mongrels) using god.
I sshed in to the mongrels server today for the first time in a couple
of months, and tried to restart them manually via god, but couldn't
allocate mem. Turns out god was using 1GB of ram.
The other environment is a slicehost VPS. I regularly (once every two
weeks or so, depending on blog traffic) receive notifications of heavy
swap activity from slicehost, always to find out god has grown in
memory to 100MB+.
I guess the reason that I'm posting this is: could this be a config
issue on my part? or is it a major leak in god? is there any remedy,
work-around, etc?