Just run rake build in the unicorn_hunter directory.
unicorn_hunter <application> {start,restart,stop,deploy,worker,status}
Starts the application and starts monitoring the app. This also works when the app is running.
If there is another unicorn_hunter instance already monitoring the app it will be replaced with the new one.
Tells the currently running unicorn to restart all workers. If you use preload_app you need to use "deploy" for restarting the app.
Stops the application and the monitoring unicorn_hunter.
Performs a no-downtime restart by replacing the master process with a new one.
Add or remove workers.
Shows the status of the application.
To get started copy unicorn.conf.example to /etc/unicorn/unicorn_APPNAME.conf.
You also need to put unicorn.rb.example into your applications config directory.
All configuration options from the .conf file will be passed along to the unicorn master process as environment variables.
APP_ENV (defaults to "production") - environment of the application
APP_ROOT (required) - path to your application's root directory
APP_RUBY (defaults to "ruby") - override default ruby version
APP_TYPE (defaults to "rails") - can be "rails" or "rack"
BUNDLER (defaults to "false") - set to "true" if you're using gem bundler (and need to run bin/unicorn)
UNICORN_LISTEN (defaults to 3000) - port or socket to listen on (example: "/var/run/unicorn/unicorn_test.sock" or "80")
UNICORN_PID_DIR (defaults to /var/run/unicorn) - path to store unicorn..pid and unicorn..hunter.pid
UNICORN_WORKERS (defaults to 3) - number of unicorn workers to start