Monitors system memory and scale up/down Unicorn workers. This setup can be used for applications where memory usage wildly varies based on usage so that static number of workers may not be ideal. Unicop uses TTIN and TTOU signals to scale up/down number of unicorn workers. Main purpose of this tool is to make use of spare memory to speed up request response time by dynamically managing the number of Unicorn workers.
- Clone the repo.
- Run
ruby unicop.rb
ruby unicop.rb --dry-run will output a report and actions to be performed.
Can be used with cron to run in every n minutes.
eg. */10 * * * * /home/ubuntu/unicop/start.sh (Runs in every 10 minutes)
Configuration is stored in config.yml in project directory. Update this to suite your application needs.
scale_up_trigger_mem_per: Percentage of memory when new worker gets spawned(Scale Up).
scale_down_trigger_mem_per: Percentage of memory when existing worker gets killed(Scale Down).
max_workers: Maximum number of workers allowed.
min_workers: Minimum number of workers allowed.
old_files_to_keep: Number of old log files to keep.
max_file_size_in_bytes: Maximum log file size before log rotation.
Runs on standard linux distributions, throws error and exit if tried to run on non-linux environments. Requires Ruby 1.9.3+.
Creates log file in same directory with name unicop.log.
You can contribute to this project by adding issues and raising MRs.
- Fork it ( https://github.com/albertpaulp/unicop/fork )
- Create your feature branch (
git checkout -b my-new-feature) - Commit your changes (
git commit -am 'Add some feature') - Push to the branch (
git push origin my-new-feature) - Create a new Pull Request