尝试使用USR2信号重启Unicorn时，我遇到了这个特殊问题。在干净地重新启动VPS时，我没有遇到向Unicorn发送USR2信号并让它正常重启的问题。然而，如果我再试一次，大约一个小时后，我将留下一个老主人，他们会在新的主人身边停下来阻止新主人开始。然后我被迫杀死了老主人，所以新主人可以开始。如果我重新启动VPS，它会修复它，但一小时后会再次启动问题。我在Rails 4，Ruby 2.0.0上。
I, [2014-01-07T15:37:37.118523 #19797] INFO -- : executing ["/srv/rails/current/bin/unicorn", "-c", "/srv/rails/current/config/unicorn.rb", {12=>#<Kgio::UNIXServer:fd 12>}] (in /srv/rails/releases/20140107091945)
I, [2014-01-07T15:37:37.118983 #19797] INFO -- : forked child re-executing...
I, [2014-01-07T15:37:38.998632 #19797] INFO -- : inherited addr=/srv/rails/shared/sockets/unicorn.sock fd=12
I, [2014-01-07T15:37:38.999038 #19797] INFO -- : Refreshing Gem list
I, [2014-01-07T15:37:41.927794 #19967] INFO -- : Refreshing Gem list
/srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:219:in `pid=': Already running on PID:19967 (or pid=/srv/rails/shared/pids/unicorn.pid is stale) (ArgumentError)
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:151:in `start'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/bin/unicorn:126:in `<top (required)>'
from /srv/rails/current/bin/unicorn:16:in `load'
from /srv/rails/current/bin/unicorn:16:in `<main>'
/srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:219:in `pid=': Already running on PID:21250 (or pid=/srv/rails/shared/pids/unicorn.pid is stale) (ArgumentError)
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:151:in `start'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/bin/unicorn:126:in `<top (required)>'
from /srv/rails/current/bin/unicorn:16:in `load'
from /srv/rails/current/bin/unicorn:16:in `<main>'
E, [2014-01-07T15:40:46.720131 #20878] ERROR -- : reaped #<Process::Status: pid 21075 exit 1> exec()-ed
E, [2014-01-07T15:40:46.720870 #20878] ERROR -- : master loop error: Already running on PID:21250 (or pid=/srv/rails/shared/pids/unicorn.pid is stale) (ArgumentError)
E, [2014-01-07T15:40:46.723525 #20878] ERROR -- : /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:219:in `pid='
E, [2014-01-07T15:40:46.723671 #20878] ERROR -- : /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:413:in `reap_all_workers'
E, [2014-01-07T15:40:46.723747 #20878] ERROR -- : /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:292:in `join'
E, [2014-01-07T15:40:46.723815 #20878] ERROR -- : /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/bin/unicorn:126:in `<top (required)>'
E, [2014-01-07T15:40:46.723880 #20878] ERROR -- : /srv/rails/current/bin/unicorn:16:in `load'
E, [2014-01-07T15:40:46.723930 #20878] ERROR -- : /srv/rails/current/bin/unicorn:16:in `<main>'
E, [2014-01-07T15:41:13.704700 #21250] ERROR -- : adding listener failed addr=/srv/rails/shared/sockets/unicorn.sock (in use)
E, [2014-01-07T15:41:13.704901 #21250] ERROR -- : retrying in 0.5 seconds (4 tries left)
E, [2014-01-07T15:41:14.205452 #21250] ERROR -- : adding listener failed addr=/srv/rails/shared/sockets/unicorn.sock (in use)
E, [2014-01-07T15:41:14.205597 #21250] ERROR -- : retrying in 0.5 seconds (3 tries left)
78.40.124.16, 173.245.49.122 - - [07/Jan/2014 15:41:14] "GET / HTTP/1.0" 200 28697 0.8345
E, [2014-01-07T15:41:14.706179 #21250] ERROR -- : adding listener failed addr=/srv/rails/shared/sockets/unicorn.sock (in use)
E, [2014-01-07T15:41:14.706335 #21250] ERROR -- : retrying in 0.5 seconds (2 tries left)
E, [2014-01-07T15:41:15.206834 #21250] ERROR -- : adding listener failed addr=/srv/rails/shared/sockets/unicorn.sock (in use)
E, [2014-01-07T15:41:15.206987 #21250] ERROR -- : retrying in 0.5 seconds (1 tries left)
E, [2014-01-07T15:41:15.707431 #21250] ERROR -- : adding listener failed addr=/srv/rails/shared/sockets/unicorn.sock (in use)
E, [2014-01-07T15:41:15.707563 #21250] ERROR -- : retrying in 0.5 seconds (0 tries left)
78.40.124.16, 149.154.158.74 - - [07/Jan/2014 15:41:15] "GET / HTTP/1.0" 200 32866 0.4528
E, [2014-01-07T15:41:16.208055 #21250] ERROR -- : adding listener failed addr=/srv/rails/shared/sockets/unicorn.sock (in use)
/srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/socket_helper.rb:158:in `initialize': Address already in use - "/srv/rails/shared/sockets/unicorn.sock" (Errno::EADDRINUSE)
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/socket_helper.rb:158:in `new'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/socket_helper.rb:158:in `bind_listen'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:255:in `listen'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:801:in `block in bind_new_listeners!'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:801:in `each'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:801:in `bind_new_listeners!'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/lib/unicorn/http_server.rb:146:in `start'
from /srv/rails/shared/bundle/ruby/2.0.0/gems/unicorn-4.7.0/bin/unicorn:126:in `<top (required)>'
from /srv/rails/current/bin/unicorn:16:in `load'
from /srv/rails/current/bin/unicorn:16:in `<main>'
deploy_path = "/srv/rails"
RAILS_ENV = ENV['RAILS_ENV'] || "production"
working_directory "#{deploy_path}/current"
pid "#{deploy_path}/shared/pids/unicorn.pid"
stderr_path "#{deploy_path}/shared/log/unicorn.log"
# Listen on a UNIX data socket
listen "#{deploy_path}/shared/sockets/unicorn.sock"
worker_processes 4
# Preload application before forking worker processes
preload_app true
# Restart any workers that haven't responded in 30 seconds
timeout 30
before_fork do |server, worker|
##
# When sent a USR2, Unicorn will suffix its pidfile with .oldbin and
# immediately start loading up a new version of itself (loaded with a new
# version of our app). When this new Unicorn is completely loaded
# it will begin spawning workers. The first worker spawned will check to
# see if an .oldbin pidfile exists. If so, this means we've just booted up
# a new Unicorn and need to tell the old one that it can now die. To do so
# we send it a QUIT.
#
# Using this method we get 0 downtime deploys.
old_pid = "#{server.config[:pid]}.oldbin"
if File.exists?(old_pid) && server.pid != old_pid
begin
Process.kill("QUIT", File.read(old_pid).to_i)
rescue Errno::ENOENT, Errno::ESRCH => e
log = File.open(Rails.root.join('log/unicorn.log'), "a")
log.puts "Error encountered when killing process:\n"
log.puts "#{e.message}"
log.close
end
end
# the following is recomended for Rails + "preload_app true"
# as there's no need for the master process to hold a connection
if defined?(ActiveRecord::Base)
ActiveRecord::Base.connection.disconnect!
end
end
after_fork do |server, worker|
##
# Unicorn master loads the app then forks off workers - because of the way
# Unix forking works, we need to make sure we aren't using any of the parent's
# sockets, e.g. db connection
ActiveRecord::Base.establish_connection
# Redis and Memcached would go here but their connections are established
# on demand, so the master never opens a socket
##
# Unicorn master is started as root, which is fine, but let's
# drop the workers to deployer
begin
uid, gid = Process.euid, Process.egid
user, group = 'deployer', 'deployer'
target_uid = Etc.getpwnam(user).uid
target_gid = Etc.getgrnam(group).gid
worker.tmp.chown(target_uid, target_gid)
if uid != target_uid || gid != target_gid
Process.initgroups(user, target_gid)
Process::GID.change_privilege(target_gid)
Process::UID.change_privilege(target_uid)
end
rescue => e
if RAILS_ENV == 'development'
STDERR.puts "couldn't change user, oh well"
else
raise e
end
end
end
require 'bundler/capistrano' # runs a bundle install --deployment
# https://github.com/sstephenson/rbenv/issues/101
set :keep_releases, 10
set :shared_children, shared_children + %w(public/images public/uploads)
# Multistage extension
set :stages, ["production", "staging"]
set :default_stage, "staging"
require 'capistrano/ext/multistage'
require 'underglow/capistrano'
# Whenever crontab updates
set :whenever_environment, defer { stage }
set :whenever_command, "bin/whenever"
require 'whenever/capistrano'
set :application, "rails"
set :user, "deployer"
default_run_options[:pty] = true
default_run_options[:shell] = '/bin/zsh'
set :use_sudo, false
# repository
set :repository, "XXXXXXXXXXXXXXXXX"
set :branch, fetch(:branch, "master") # can specify a branch from `cap -S branch="<branch_name>"`
set :scm, :git
set :scm_verbose, true
set :ssh_options, forward_agent: true
set :deploy_to, "/srv/rails"
set :deploy_via, :remote_cache
# We're using a rbenv user install, setup the PATH we need to access the rbenv shims
set :default_environment, {
'PATH' => "$HOME/.rbenv/shims:$HOME/.rbenv/bin:$PATH"
}
有没有人见过这个？
答案 0 :(得分：4)
您应该检查独角兽stdout / stderr日志，以获取更多证据，说明为什么旧的独角兽可能会挂起或新的麒麟未正确杀死它。
一个问题是，如果在部署新版本期间删除了较旧的capistrano版本目录，则在热交换切换期间可能会出现捆绑器错误。人们建议添加以下内容以绑定到Gemfile与特定于发布的路径的永久路径：
before_exec do |server|
ENV['BUNDLE_GEMFILE'] = "#{deploy_path}/current/Gemfile"
end
如果这是您遇到的问题，您应该看到捆绑器错误或独角兽日志失败。
答案 1 :(得分：3)
这可能对您没有帮助，但这是我为解决问题而采取的措施。
我从Unicorn 4.7.0的发布开始遇到这个问题。在4.7.0中，pid文件的写入行为发生了变化，并破坏了我的重启脚本。旧的4.7.0之前的行为是：将pid文件移动到oldpid，写入新的pid，启动worker，关闭master。最后一步当然是在我的unicorn.rb文件中。新的行为是迅速删除旧的pid，并且只在发生一些繁重的提升之后才写出新的pid。这打破了我的脚本，因为它不能相信事情正确地重新启动。这导致我的sh脚本尝试重新启动它，导致与现在刚刚开始的独角兽进程混淆，并且eh-script开始“全面启动”。两者都以各种方式丢失，因此两人都离开了，留下一位老主人仍在服务请求。
我的unicorn.rb文件中也有一个缺陷，就像有人提到的那样，没有正确设置捆绑包。
升级到最近发布的Unicorn 4.8.1解决了这个问题，因为pid文件是在4.7.0天之前编写的。