Puma workers无法在Ubuntu 20.04 VM上的Rails 5.2中启动

wgeznvg7  于 2023-10-17  发布在  其他
关注(0)|答案(1)|浏览(203)

我正在学习使用Capistrano部署一个演示rails应用程序，其中Puma作为应用服务器，Nginx作为web服务器。我在一个文件_stage.rb中设置了必要的puma配置，然后将puma设置为/etc/init.d/puma_myarticles_staging服务。可执行文件puma_init.sh.erb后来作为puma_init.sh写入远程服务器，看起来像，

  1. #!/usr/bin/env bash
  3. PATH=/usr/local/bin:/usr/local/sbin/:/sbin:/usr/sbin:/bin:/usr/bin
  4. DESC="Puma rack web server"
  5. NAME=puma_<%=fetch(:full_app_name)%>
  6. SCRIPT_NAME=/etc/init.d/${NAME}
  7. APP_ROOT=<%=current_path%>
  8. PIDFILE=<%= fetch(:puma_pid) %>
  9. STATE_FILE=<%= fetch(:puma_state) %>
  11. log_daemon_msg() { echo "$@"; }
  12. log_end_msg() { [ $1 -eq 0 ] && RES=OK; logger ${RES:=FAIL}; }
  14. run_pumactl(){
  15.   [ $# -lt 1 ] && echo "$# params were given, Expected 1" && exit 1
  16.   cd ${APP_ROOT} && <%= fetch(:rbenv_prefix) %> bundle exec pumactl -F <%=fetch(:puma_conf)%> $1
  17. }
  19. # Function that starts the puma
  20. #
  21. start_task() {
  22.   if [ -e ${PIDFILE} ]; then
  23.     PID=`cat ${PIDFILE}`
  24.     # If the puma isn't running, run it, otherwise restart it.
  25.     if [ "`ps -A -o pid= | grep -c ${PID}`" -eq 0 ]; then
  26.       do_start_task
  27.     else
  28.       restart_task
  29.     fi
  30.   else
  31.     do_start_task
  32.   fi
  33. }
  34. do_start_task() {
  35.   log_daemon_msg "--> Woke up puma ${APP_ROOT}"
  36.   run_pumactl start
  37. }
  39. # Function that stops the daemon/service
  40. #
  41. stop_task() {
  42.   log_daemon_msg "--> Stopping puma in path: ${APP_ROOT} ..."
  43.   if [ -e ${PIDFILE} ]; then
  44.     PID=`cat ${PIDFILE}`
  45.     if [ "`ps -A -o pid= | grep -c ${PID}`" -eq 0 ]; then
  46.       log_daemon_msg "--> Puma isn't running in path: ${APP_ROOT}."
  47.     else
  48.       log_daemon_msg "--> About to kill puma with PID: `cat $PIDFILE` ..."
  49.       if [ "`ps -A -o pid= | grep -c ${PID}`" -eq 0 ]; then
  50.         log_daemon_msg "--> Puma isn't running in path: ${APP_ROOT}."
  51.         return 0
  52.       else
  53.         run_pumactl stop
  54.         log_daemon_msg "--> Waiting for status ..."
  55.         sleep 5
  56.         if [ "`ps -A -o pid= | grep -c ${PID}`" -eq 0 ]; then
  57.           log_daemon_msg "--> Puma with pid ${PID} stopped successfully."
  58.           rm -f ${PIDFILE} ${STATE_FILE}
  59.         else
  60.           log_daemon_msg "--> Unable to stop puma with pid ${PID}."
  61.         fi
  62.       fi
  63.     fi
  64.   else
  65.     log_daemon_msg "--> Puma isn't running in path: ${APP_ROOT}."
  66.   fi
  67.   return 0
  68. }
  70. # Function that sends a SIGUSR2 to the daemon/service
  71. #
  72. restart_task() {
  73.   if [ -e ${PIDFILE} ]; then
  74.     log_daemon_msg "--> About to restart puma in path: ${APP_ROOT} ..."
  75.     run_pumactl restart
  76.   else
  77.     log_daemon_msg "--> Your puma was never playing... Let's get it out there first ..."
  78.     start_task
  79.   fi
  80.   return 0
  81. }
  83. # Function that sends a SIGUSR2 to the daemon/service
  84. #
  85. status_task() {
  86.   if [ -e ${PIDFILE} ]; then
  87.     log_daemon_msg "--> About to status puma ${APP_ROOT} ..."
  88.     run_pumactl status
  89.   else
  90.     log_daemon_msg "---> Puma isn't running in path: ${APP_ROOT}."
  91.   fi
  92.   return 0
  93. }
  95. case "$1" in
  96.   start)
  97.     [ "$VERBOSE" != no ] && log_daemon_msg "Starting ${DESC}" "${NAME} ..."
  98.     start_task
  99.     case "$?" in
  100.       0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
  101.       2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
  102.     esac
  103.     ;;
  104.   stop)
  105.     [ "$VERBOSE" != no ] && log_daemon_msg "Stopping ${DESC}" "${NAME} ..."
  106.     stop_task
  107.     case "$?" in
  108.       0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
  109.       2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
  110.     esac
  111.     ;;
  112.   status)
  113.     log_daemon_msg "Status ${DESC}" "${NAME} ..."
  114.     status_task
  115.     case "$?" in
  116.       0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
  117.       2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
  118.     esac
  119.     ;;
  120.   restart)
  121.     log_daemon_msg "Restarting ${DESC}" "${NAME} ..."
  122.     restart_task
  123.     case "$?" in
  124.       0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
  125.       2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
  126.     esac
  127.     ;;
  128.   *)
  129.     echo "Usage:" >&2
  130.     echo "  ${SCRIPT_NAME} {start|stop|status|restart}" >&2
  131.     exit 3
  132.     ;;
  133. esac
  134. :

puma.rb是，

  1. #!/usr/bin/env puma
  3. directory '/app/myarticles_staging/current'
  4. environment 'staging'
  6. pidfile '/app/myarticles_staging/shared/tmp/pids/puma.pid'
  7. state_path '/app/myarticles_staging/shared/tmp/states/puma.state'
  8. stdout_redirect '/app/myarticles_staging/shared/log/puma_access.log', '/app/myarticles_staging/shared/log/puma_error.log', true
  10. daemonize
  12. threads 4, 8
  14. bind 'unix:///app/myarticles_staging/shared/tmp/sockets/puma.myarticles_staging.sock'
  16. activate_control_app 'unix:///app/myarticles_staging/shared/tmp/sockets/pumactl.myarticles_staging.sock'
  18. workers '4'
  20. preload_app!
  22. on_worker_boot do
  23.   require "active_record"
  24.   ActiveRecord::Base.connection.disconnect! rescue ActiveRecord::ConnectionNotEstablished
  25.   ActiveRecord::Base.establish_connection(YAML.load_file('/app/myarticles_staging/shared/config/database.yml')['staging'])
  26. end
  28. # Allow puma to be restarted by the `rails restart` command.
  29. plugin :tmp_restart

我从一个名为_stage.rb的文件中获取所有的puma配置，

  1. set :stage, :staging
  2. set :branch, :staging
  4. set :server_port, 80
  6. set :full_app_name, "#{fetch(:application)}_#{fetch(:stage)}"
  7. set :rails_env, :staging
  9. set :deploy_to, "/app/#{fetch(:full_app_name)}"
  11. set :puma_user, fetch(:deploy_user)
  12. set :puma_state, "#{shared_path}/tmp/states/puma.state"
  13. set :puma_pid, "#{shared_path}/tmp/pids/puma.pid"
  14. set :puma_rackup, -> { File.join(current_path, 'config.ru')}
  15. set :puma_bind, "unix://#{shared_path}/tmp/sockets/puma.#{fetch(:full_app_name)}.sock"
  16. set :puma_default_control_app, "unix://#{shared_path}/tmp/sockets/pumactl.#{fetch(:full_app_name)}.sock"
  17. set :puma_conf, "#{shared_path}/config/puma.rb"
  18. set :puma_workers, 4
  19. set :puma_threads, [4, 8]
  20. set :puma_role, :app
  21. set :puma_env, :staging
  22. set :puma_preload_app, true
  24. set :puma_enable_socket_service, true
  25. set :puma_access_log, "#{shared_path}/log/puma_access.log"
  26. set :puma_error_log, "#{shared_path}/log/puma_error.log"
  28. set :nginx_access_log, "#{shared_path}/log/nginx_access.log"
  29. set :nginx_error_log, "#{shared_path}/log/nginx_error.log"

当我以/etc/init.d/puma_myarticles_staging start启动puma服务时，它输出，

  1. Starting Puma rack web server puma_myarticles_staging ...
  2. --> Woke up puma /app/myarticles_staging/current
  3. [3955] Puma starting in cluster mode...
  4. [3955] * Version 4.3.12 (ruby 2.7.0-p0), codename: Mysterious Traveller
  5. [3955] * Min threads: 4, max threads: 8
  6. [3955] * Environment: staging
  7. [3955] * Process workers: 1
  8. [3955] * Preloading application
  9. [3955] * Listening on unix:///app/myarticles_staging/shared/tmp/sockets/puma.myarticles_staging.sock
  10. [3955] ! WARNING: Detected 1 Thread(s) started in app boot:
  11. [3955] ! #<Thread:0x000055c692868bf8 /app/myarticles_staging/shared/bundle/ruby/2.7.0/gems/activerecord-6.1.7.4/lib/active_record/connection_adapters/abstract/connection_pool.rb:323 sleep> - /app/myarticles_staging/shared/bundle/ruby/2.7.0/gems/activerecord-6.1.7.4/lib/active_record/connection_adapters/abstract/connection_pool.rb:329:in `sleep'
  12. [3955] * Daemonizing...

没有留下新的美洲狮或国家档案。最终，puma服务工作人员没有 Boot ，因为我检查了puma.pid和puma.state没有文件正在写入。当我运行rbenv exec bundle exec rails s在current_path上手动测试时，它工作正常。
我使用ps ax | grep puma检查了被妖魔化的puma进程，但没有找到实际的puma工作人员，

  1. 1516 pts/0    S+     0:00 grep --color=auto puma

有什么建议我可能做错了吗？先谢了。

ubuntu

1条答案

按热度按时间
r1zk6ea1

r1zk6ea11#

也许问题出在美洲狮开始的方式上。我很确定这里没有美洲狮。您还没有发布有关systemctl设置的任何详细信息，但最佳做法是使用用户服务文件，这样puma将始终在 Boot 时启动，并且不需要密码
我使用的标准puma配置是这样的

  1. # Puma can serve each request in a thread from an internal thread pool.
  2. # The `threads` method setting takes two numbers: a minimum and maximum.
  3. # Any libraries that use thread pools should be configured to match
  4. # the maximum value specified for Puma. Default is set to 5 threads for minimum
  5. # and maximum; this matches the default thread size of Active Record.
  6. #
  7. max_threads_count = ENV.fetch("RAILS_MAX_THREADS") { 5 }
  8. min_threads_count = ENV.fetch("RAILS_MIN_THREADS") { max_threads_count }
  9. threads min_threads_count, max_threads_count
  11. # Change to match your CPU core count
  12. workers 0
  14. # Specifies the `port` that Puma will listen on to receive requests; default is 3000.
  15. #
  16. port ENV.fetch("PORT") { 3000 }
  18. # Specifies the `environment` that Puma will run in.
  19. #
  21. rails_env = ENV['RAILS_ENV'] || "production"
  22. environment rails_env
  24. app_dir = File.expand_path("../..", __FILE__)
  25. #shared_dir = "#{app_dir}/shared"
  26. shared_dir = "/home/project/apps/comtech/shared" # Use your projects path
  28. # Specifies the `pidfile` that Puma will use.
  29. #pidfile ENV.fetch("PIDFILE") { "pids/server.pid" }
  31. pidfile "#{shared_dir}/pids/puma.pid"
  32. state_path "#{shared_dir}/pids/puma.state"
  33. activate_control_app
  35. # Set up socket location
  36. bind "unix://#{shared_dir}/sockets/comtech_puma.sock"
  38. on_worker_boot do
  39.   require "active_record"
  40.   ActiveRecord::Base.connection.disconnect! rescue ActiveRecord::ConnectionNotEstablished
  41.   ActiveRecord::Base.establish_connection(YAML.load_file("#{app_dir}/config/database.yml")[rails_env])
  42. end
  44. # Specifies the number of `workers` to boot in clustered mode.
  45. # Workers are forked web server processes. If using threads and workers together
  46. # the concurrency of the application would be max `threads` * `workers`.
  47. # Workers do not work on JRuby or Windows (both of which do not support
  48. # processes).
  49. #
  50. # workers ENV.fetch("WEB_CONCURRENCY") { 2 }
  52. # Use the `preload_app!` method when specifying a `workers` number.
  53. # This directive tells Puma to first boot the application and load code
  54. # before forking the application. This takes advantage of Copy On Write
  55. # process behavior so workers use less memory.
  56. #
  57. # preload_app!
  59. # Allow puma to be restarted by `rails restart` command.
  60. plugin :tmp_restart

显然，您需要在服务器上将ENVIRONMENT变量设置为staging
相应地调整pids文件夹以使用tmp。
要使用用户服务而不是系统服务，请查看以下内容
cd ~/.config/systemd/user创建不存在的文件夹
nano name_of_your_puma.service并相应地粘贴以下内容调整路径。

  1. [Unit]
  2. Description=Puma Rack Server
  3. # Puma supports systemd's `Type=notify` and watchdog service
  4. # monitoring, if the [sd_notify](https://github.com/agis/ruby-sdnotify) gem is installed,
  5. # as of Puma 5.1 or later.
  6. # On earlier versions of Puma or JRuby, change this to `Type=simple` and remove
  7. # the `WatchdogSec` line.
  9. [Service]
  10. Type=simple
  12. # If your Puma process locks up, systemd's watchdog will restart it within seconds.
  13. #WatchdogSec=10
  14. RestartSec=10
  16. WorkingDirectory=/home/comtech/apps/comtech/current
  17. #PIDFile=/home/comtech/apps/comtech/shared/pids/puma.pid
  18. #User=comtech
  20. ExecStart=/home/comtech/.rvm/bin/rvm 3.1.3@comtech_cms_app do bundle exec puma -C /home/comtech/apps/comtech/current/config/puma_production.rb
  21. ExecStop=/home/comtech/.rvm/bin/rvm 3.1.3@comtech_cms_app do bundle exec pumactl -S /home/comtech/apps/comtech/shared/pids/puma.state stop
  22. ExecReload=/home/comtech/.rvm/bin/rvm [email protected]@comtech_cms_app do bundle exec pumactl -S /home/comtech/apps/comtech/shared/pids/puma.state restart
  23. Restart=always
  25. [Install]
  26. WantedBy=multi-user.target

然后运行

  1. $ systemctl --user enable name_of_your_puma.service

来检查服务

  1. $ systemctl --user status

停止服务

  1. $ systemctl --user stop

启动服务

  1. $ systemctl --user start

该服务将在 Boot 时自动启动，并在puma失败时重新启动puma。还要注意我使用的是RVM，因此需要相应地调整start命令

展开查看全部
赞(0）分享  回复(0）举报  2023-10-17
我来回答

相关问题

    查看更多

    热门标签

    更多
    JavaquerypythonNode开发语言requestUtil数据库Table后端算法LoggerMessageElementParser

    最新问答

    更多
    • 蜀ICP备13028337号-1 大数据知识库 https://www.saoniuhuo.com © All rights reserved
    • 本站内容来源互联网,如果侵犯您的权益请联系我们删除, 联系方式：448109455@qq.com