Anyone know of a tool to detect and report on repeating patterns in a log file? [closed]

Solution 1:

I've heard of people applying Bayesian filtering on log files to spot interesting stuff versus routine log entries. They used spam filters, where the routine uninteresting entries were considered "good" while the unusual ones were considered as "spam" and using that coloring they were able to shift through.

It sounds a lot like machine learning stuff to me, but then again I've not seen it in action, only heard of it over beers.

Solution 2:

Splunk works wonders for this sort of stuff. I use it internally to gather all the logs and do quick searches via its excellent browser-based interface.