Timber
Rake tasks for dealing with large rails logs, e.g. a memory-efficient task to interleave logs from multiple machines into a single chronological file. Also provides a foundation for extending custom log parsing. Does NOT require Hodel 3000.
Install / Use
/learn @cainlevy/TimberREADME
= Timber
A memory-efficient Rails log parser that does not require Hodel 3000 formatting. Just feed it the production logs you've already got.
Intended to work as a foundation for special logging tools like Oink (http://github.com/noahd1/oink), which add custom lines and then need to parse the extra variables back out for analysis.
= Examples
== Combining
Suppose that you're running a cluster of machines that don't share a file system (e.g. EY Cloud). Each has its own log file. That'd be much easier to grep if it were combined into one, yeah? First, grab the logs and unpack them into your working directory. Then:
rake timber:combine LOGS=production*.log-20100718
Then wait a moment. Shouldn't take long (my laptop will combine ~375 mbs from four logs in about 2 minutes).
Copyright (c) 2010 Lance Ivy, released under the MIT license
