Monitoring Merb Processes in the Wild with God

Before you start reading, this post only applies for Merb 0.5.3, any newer version is totally untested

When you take Merb out into the wild, it does, unfortunatly, suffer a lot of the same problems as the mongrel handler than runs Rails.

There is however a saviour out there - God - To clarify, im not talking about the man upstairs; rather the process monitoring tool which rocks at restarting bloated mongrel processes on *nix based OS.

The Merb handler does not have any way of restarting a running cluster, so you physically have to stop, then start a new merb process. This is somewhat out of sync with how the god process handling and restarting works in that you define start, stop and critically, restart paramaters. To get around this we have to use a hacky sleep then start method - its not ideal, but hey, it works :)

Rather than letting merb handle the process forking, what were going to do is let God handle the writing of Pids and managing of the process. Ok, less of all this talk and lets take a look at some configuration code for the God configuration.

# MIT License
# 
# Copyright (c) 2008, Tim Perrett
# 
# Permission is hereby granted, free of charge, to any person 
# obtaining a copy of this software and associated 
# documentation files (the "Software"), to deal in the Software
# without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, 
# and/or sell copies of the Software, and to permit persons 
# to whom the Software is furnished to do so, subject to the 
# following conditions:
# 
# The above copyright notice and this permission notice shall 
# be included in all copies or substantial portions of the Software.
# 
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF 
# ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED 
# TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A 
# PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT 
# SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR 
# ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN 
# ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE 
# OR OTHER DEALINGS IN THE SOFTWARE.

SITE_LOCATION = "/var/www/sites/domain.com" 
MERB_ROOT = "#{SITE_LOCATION}/current" 
MERB_CONFIG = "#{SITE_LOCATION}/shared/config/merb.yml" 
MERB_ENVIROMENT = 'production'
PROCESS_USER = 'timperrett'
PROCESS_GROUP = 'timperrett'

%w{4006 4007}.each do |port|  
  God.watch do |w|

    pid_path = File.join(MERB_ROOT, 'log',"merb.#{port}.pid")

    w.name = "merb-#{port}" 
    w.interval = 30.seconds # default
    w.start = "#{MERB_ROOT}/script/merb -u #{PROCESS_USER} \
    -G #{PROCESS_GROUP} -M #{MERB_CONFIG} -p #{port} \
    -e #{MERB_ENVIROMENT} -d" 
    w.log = "/home/timperrett/godmerb.log" 
    w.stop = "(cd #{MERB_ROOT}; merb -k #{port})" 
    w.restart = "(cd #{MERB_ROOT}; merb -k #{port}); sleep 1; \
    #{MERB_ROOT}/script/merb -u #{PROCESS_USER} -G \
    #{PROCESS_GROUP} -M #{MERB_CONFIG} -p #{port} \
    -e #{MERB_ENVIROMENT} -d" 
    w.start_grace = 5.seconds
    w.restart_grace = 20.seconds
    w.pid_file = File.join(MERB_ROOT, "log/merb.#{port}.pid")
    # w.group = "merbs" 
    w.behavior(:clean_pid_file)

    w.start_if do |start|
      start.condition(:process_running) do |c|
        c.interval = 10.seconds
        c.running = false
      end
    end

    w.restart_if do |restart|
      restart.condition(:memory_usage) do |c|
        c.above = 51.megabytes
        c.times = [3, 5] # 3 out of 5 intervals
      end

      restart.condition(:cpu_usage) do |c|
        c.above = 50.percent
        c.times = 5
      end
    end

    w.lifecycle do |on|
      on.condition(:flapping) do |c|
        c.to_state = [:start, :restart]
        c.times = 5
        c.within = 5.minute
        c.transition = :unmonitored
        c.retry_in = 5.minutes
        c.retry_times = 5
        c.retry_within = 2.hours
      end
    end

  end
end

Im not sure that this is ideal, but it certainly seems to work for me and importantly keeps the site running without problems - which is a dang sight better than them becoming unresponsive!

I hope this might help someone, somewhere….

comments powered by Disqus