LogJuicer

LogJuicer takes a Rails production log and attempts to squeeze some useful information out of it. Mainly intended for sampling purposes, LogJuicer might help identify which controllers or actions are used most often and which are suffering from long loading times in the real world.

Screenshots and Demo

Screenshots can be found on the LogJuicer home page (http://logjuicer.org) and a live demo can be seen at http://demo.logjuicer.org.

Installation/Usage from README:

1. Create a database (production), alter config/database.yml as necessary.

(create development and test if you want to run tests)

2. Bootstrap the database (rake db:schema:load RAILS_ENV=production)

3. Run rake task to import a rails production log file:

(rake tasks from: 'rake -D juice')

rake juice:import
    Parse a production log file and import into database, specifing with 
LOG_FILE=/path/to/production.log and PARAMS_TO_SKIP=password,secret (and you 
probably want RAILS_ENV=production)

rake juice:read_test
    Parse a production log file without importing (test run), specifying 
LOG_FILE=/path/to/production.log and PARAMS_TO_SKIP=password,secret (and you 
probably want RAILS_ENV=production)

4. Start server and point browser to server

$ mongrel_rails start -e production
$ firefox http://localhost:3000

Notes

LogJuicer is going to use some page and action caching, so you need to make sure write permissions are set for the proper user on the public and tmp/cache directories. If you are running mod_rails or fcgi, this means the user that your web server runs as (e.g. apache) needs write permissions. If you are using mongrel, the whatever user that is running the mongrel process needs write permissions.

When you run the juice:import task, all cached files are going to be purged automatically.