We had a short load spike on our application servers a couple weeks back. A load of 28 on a 4-core machine is more than uncomfortable – it’s downright dangerous. Luckily, it only lasted for 2 minutes and, just as suddenly as it came, vanished again without a trace. Well, that’s not quite true, because this is exactly why we keep those gigabytes of Apache logs lying around!
I knew where the information was, but how to crawl through that >1GB log and find it? A few years back I had done some similar research through JBoss application logs using Splunk. It was a fairly pleasant experience and I wondered if they were still around.