The server has been having problems since the last few hours of intermittently becoming unresponsive.
We think the problem is due to faulty RAM and have scheduled an immediate RAM
replacement to solve it.
We will update this post regularly with more information.
2012-02-20 19:44 UTC: the server RAM was replaced, but the problem persists. We’ll continue to troubleshoot and will update this post when we have more information.
2012-02-20 01:16 UTC: We’ve isolated the problem down to a out of memory killer error that is being triggered by numerous processes (different processes each time). We’ve disabled all non-essential services on the machine. The server seems stable for now and we’re continuing to monitor it.