resque - 如何为多个进程创建一个 monit 循环来监控？

Question

这个例子展示了如何监控单个resque队列

check process resque_worker_QUEUE
  with pidfile /data/APP_NAME/current/tmp/pids/resque_worker_QUEUE.pid
  start program = "/usr/bin/env HOME=/home/user RACK_ENV=production PATH=/usr/local/bin:/usr/local/ruby/bin:/usr/bin:/bin:$PATH /bin/sh -l -c 'cd /data/APP_NAME/current; nohup bundle exec rake environment resque:work RAILS_ENV=production QUEUE=queue_name VERBOSE=1 PIDFILE=tmp/pids/resque_worker_QUEUE.pid >> log/resque_worker_QUEUE.log 2>&1'" as uid deploy and gid deploy
  stop program = "/bin/sh -c 'cd /data/APP_NAME/current && kill -9 $(cat tmp/pids/resque_worker_QUEUE.pid) && rm -f tmp/pids/resque_worker_QUEUE.pid; exit 0;'"
  if totalmem is greater than 300 MB for 10 cycles then restart  # eating up memory?
  group resque_workers

其中 QUEUE 通常是队列的索引。 monit 本身是否有能力创建一个循环，以便 QUEUE 可以是索引或迭代器，所以如果我有 6 个工作人员要创建，我仍然可以在一个 block 中有一个配置代码块？还是我必须创建一个 monit 配置构建器来进行迭代以生成一组硬编码的工作监视器作为输出？

所以代替

check process resque_worker_0
  with pidfile /data/APP_NAME/current/tmp/pids/resque_worker_0.pid
  start program = "/usr/bin/env HOME=/home/user RACK_ENV=production PATH=/usr/local/bin:/usr/local/ruby/bin:/usr/bin:/bin:$PATH /bin/sh -l -c 'cd /data/APP_NAME/current; nohup bundle exec rake environment resque:work RAILS_ENV=production QUEUE=queue_name VERBOSE=1 PIDFILE=tmp/pids/resque_worker_0.pid >> log/resque_worker_0.log 2>&1'" as uid deploy and gid deploy
  stop program = "/bin/sh -c 'cd /data/APP_NAME/current && kill -9 $(cat tmp/pids/resque_worker_0.pid) && rm -f tmp/pids/resque_worker_0.pid; exit 0;'"
  if totalmem is greater than 300 MB for 10 cycles then restart  # eating up memory?
  group resque_workers

check process resque_worker_1
  with pidfile /data/APP_NAME/current/tmp/pids/resque_worker_1.pid
  start program = "/usr/bin/env HOME=/home/user RACK_ENV=production PATH=/usr/local/bin:/usr/local/ruby/bin:/usr/bin:/bin:$PATH /bin/sh -l -c 'cd /data/APP_NAME/current; nohup bundle exec rake environment resque:work RAILS_ENV=production QUEUE=queue_name VERBOSE=1 PIDFILE=tmp/pids/resque_worker_1.pid >> log/resque_worker_1.log 2>&1'" as uid deploy and gid deploy
  stop program = "/bin/sh -c 'cd /data/APP_NAME/current && kill -9 $(cat tmp/pids/resque_worker_1.pid) && rm -f tmp/pids/resque_worker_1.pid; exit 0;'"
  if totalmem is greater than 300 MB for 10 cycles then restart  # eating up memory?
  group resque_workers

我可以做这样的事情(我知道循环的伪代码)

[0..1].each |QUEUE|
    check process resque_worker_QUEUE
      with pidfile /data/APP_NAME/current/tmp/pids/resque_worker_QUEUE.pid
      start program = "/usr/bin/env HOME=/home/user RACK_ENV=production PATH=/usr/local/bin:/usr/local/ruby/bin:/usr/bin:/bin:$PATH /bin/sh -l -c 'cd /data/APP_NAME/current; nohup bundle exec rake environment resque:work RAILS_ENV=production QUEUE=queue_name VERBOSE=1 PIDFILE=tmp/pids/resque_worker_QUEUE.pid >> log/resque_worker_QUEUE.log 2>&1'" as uid deploy and gid deploy
      stop program = "/bin/sh -c 'cd /data/APP_NAME/current && kill -9 $(cat tmp/pids/resque_worker_QUEUE.pid) && rm -f tmp/pids/resque_worker_QUEUE.pid; exit 0;'"
      if totalmem is greater than 300 MB for 10 cycles then restart  # eating up memory?
      group resque_workers
end

Answer 1

我找不到任何证据表明 monit 可以自行执行此操作，因此我编写了一个 ruby​​ monit resque 配置文件构建器并插入到 capistrano 部署任务中。

在 config/deploy/production.rb 中

set :resque_worker_count, 6

在 lib/capistrano/tasks/monit.rake

def build_entry(process_name,worker_pid_file,worker_config_file,start_command,stop_command)
<<-END_OF_ENTRY
check process #{process_name}
  with pidfile #{worker_pid_file}
  start program = \"#{start_command}\" with timeout 90 seconds
  stop program = \"#{stop_command}\" with timeout 90 seconds
  if totalmem is greater than 500 MB for 4 cycles then restart # eating up memory?
  group resque
END_OF_ENTRY
end

namespace :monit do
  desc "Build monit configuration file for monitoring resque workers"
  task :build_resque_configuration_file do
    on roles(:app) do |host|
      # Setup the reusable variables across all worker entries
      rails_env = fetch(:rails_env)
      app_name = fetch(:application)
      monit_resque_config_file_path = "#{shared_path}/config/monit/resque"
      resque_control_script = "#{shared_path}/bin/resque-control"
      monit_wrapper_script = "/usr/local/sbin/monit-wrapper"
      config_file_content = []
      (0..((fetch(:resque_worker_count)).to_i - 1)).each do |worker|
        # Setup the variables for the worker entry
        process_name = "resque_#{worker}"
        worker_config_file = "resque_#{worker}.conf"
        worker_pid_file = "/var/run/resque/#{app_name}/resque_#{worker}.pid"
        start_command = "#{monit_wrapper_script} #{resque_control_script} #{app_name} start #{rails_env} #{worker_config_file}"
        stop_command = "#{monit_wrapper_script} #{resque_control_script} #{app_name} stop #{rails_env} #{worker_config_file}"
        # Build the config file entry for the worker
        config_file_content << build_entry(process_name,worker_pid_file,worker_config_file,start_command,stop_command)
      end
      # Save the file locally for inspection (debugging)
      temp_file = "/tmp/#{app_name}_#{rails_env}_resque"
      File.delete(temp_file) if File.exist?(temp_file)
      File.open(temp_file,'w+') {|f| f.write config_file_content.join("\n") }
      # Upload the results to the server 
      upload! temp_file, monit_resque_config_file_path
    end
  end
end