My server seized up. It was virtually unusable. This seemed like a likely culprit:
root 1037 0.0 92.6 1455696 958368 ? Ssl 2012 3:57 /usr/sbin/nscd
That's 92% RAM. I killed it to get my server operational again.
Given that I have now killed the process, what post-mortem diagnostics can I do to see what the cause was?
top occasionally showed high
wa values, which might suggest there might be something up with a file or network socket it was trying to read? Any suggestions?