Debugging system-related application errors in Linux

Question

I am now faced with trying to debug some odd crashes for a commercial application, the dreaded SAS, and I am looking for some clues that might help to diagnose the issue.

Of course I check the SAS runtime logs, and I also check /var/log/messages. Nothing in /var/log/messages, but the SAS runtime log seems to indicate a memory or disk space issue:

ERROR: Insufficient space in file WORK.'#tf0533'n.UTILITY.
ERROR: Insufficient space in file WORK.20110609_BCH.DATA.

There is surely enough disk space, so I am lead to believe it might be a memory issue. What other system logs or even tools can I use for looking for further information (such as writing to a full disk device, spikes in memory usage, swapfs errors, etc.)?

Answer 1

You can trace system calls via strace and calls to library functions via ltrace.