Start multiple Resque workers with Upstart

Acknowledgement.txt

Thanks to Jason Roelofs for his article at http://jasonroelofs.com/2012/03/12/manage-and-monitor-resque-with-upstart-and-monit/ which got me most of the way there.

resque-worker.conf

description "Start a Resque worker by supplying an ID. For example: start resque ID=1. See also: resque-workers.conf"

respawn
respawn limit 5 20

instance $ID

script
  APP_PATH=/apps/my_app/current
  PIDFILE=$APP_PATH/tmp/pids/resque-$ID.pid
  LOGFILE=$APP_PATH/log/resque_workers.log
  su -c "cd $APP_PATH; bundle exec rake resque:work QUEUE=* PIDFILE=$PIDFILE RAILS_ENV=production >> $LOGFILE 2>&1" - my_user
end script

resque-workers.conf

description "Start multiple Resque workers. Change NUM_WORKERS based on your needs."

start on (local-filesystems and net-device-up IFACE=eth0)

task

env NUM_WORKERS=5

script
  for i in `seq 1 $NUM_WORKERS`
  do
    start resque-worker ID=$i
  done
end script

Sorry I didn't see your comment sooner, kesor. To restart the workers gracefully, I send a USR signal to the Resque workers via a Rake task, and then let Upstart bring them back up. Here's an example of the Rake task:

task :restart_workers => :environment do
  pids = Array.new
  Resque.workers.each do |worker|
    pids << worker.to_s.split(/:/).second
  end
  if pids.size > 0
    system("kill -QUIT #{pids.join(' ')}")
  end
end

Correction: QUIT is the signal for a graceful shutdown. USR1 immediately kills the child.