What exactly do you mean by failing? Search keywords are very sensitive to details. If you have a specific event that fails, open a couple of terminal windows, do a `tail -f /path/to/error_log` and then do the thing that causes the failure while watching the log.
Also, cat isn't the best approach for checking a potentially large log file. I would typically use less. You can then search, jump to the end of the file and search backwards, etc. Also, one would assume that the error_log would be relatively short compared to the access_log. If not, then you have some real problems. My inclination would be to try to fix everything you can that causes messages to error_log.
If you start by reading the logs as things happen, you'll get a sense of what to expect and what is unusual. The more familiarity you develop with the logs, the easier it will be for you to troubleshoot.
|