Small daemon framework for ruby, with logging, error handler, scheduling and much more.
Small daemon framework for ruby, with logging, error handler, scheduling and much more.
My inspiration was forever for node.js written by Charlie Robbins.
My scheduling inspiration was taken from clockwork written by Adam Wiggins.
There are some alternatives, one of the best is resque, so why another daemons framework?
In my servers I’ve several daemons and what I need is:
As like sinatra and padrino I need a
thin framework to do these jobs in few seconds. This mean that:
So, if you have my needs, Forever can be the right choice for you.
$ gem install foreverb
Place your script under your standard directory, generally on my env is bin or scripts.
In that case is: bin/foo
#!/usr/bin/ruby
require 'rubygems' unless defined?(Gem)
require 'forever'
require 'mail'
Forever.run do
##
# You can set these values:
#
# dir "foo" # Default: File.expand_path('../../', __FILE__)
# file "bar" # Default: __FILE__
# log "bar.log" # Default: File.expand_path(dir, '/log/[file_name].log')
# pid "bar.pid" # Default: File.expand_path(dir, '/tmp/[file_name].pid')
#
on_error do |e|
Mail.deliver do
delivery_method :sendmail, :location => `which sendmail`.chomp
to "[email protected]"
from "[email protected]"
subject "[Foo Watcher] #{e.message}"
body "%s\n %s" % [e.message, e.backtrace.join("\n ")]
end
end
before :each do # or if you prefer before :all
require 'bundler/setup'
require 'foo'
Foo.start_loop
end
end
Assign right permission:
$ chmod +x bin/foo
start the daemon:
$ bin/foo
you should see an output like:
$ bin/foo
=> Process demonized with pid 19538
you can stop it:
$ bin/foo stop
=> Found pid 19538...
=> Killing process 19538...
You can use every method to schedule repetitive tasks.
Every allow the option :at to specify hour or minute and the option :last to specify when the every must start to loop.
:last: can be nil or a Time class. Default is 0.
:at: can be nil, a string or an array of formatted strings. Default is nil.
every 1.second, :at => '19:30' # => every second since 19:30
every 1.minute, :at => ':30' # => every minute but first call wait xx:30
every 5.minutes, :at => '18:' # => every five minutes but first call was at 18:xx
every 1.day, :at => ['18:30', '20:30'] # => every day only at 18:30 and 20:30
every 60.seconds, :last => Time.now # => will be fired 60 seconds after you launch the app
Remember that :at:
:So looking our example:
Forever.run do
dir File.expand_path('../', __FILE__) # Default is ../../__FILE__
before :all do
puts "All jobs will wait me for 1 second"; sleep 1
end
every 10.seconds, :at => "#{Time.now.hour}:00" do
puts "Every 10 seconds but first call at #{Time.now.hour}:00"
end
every 1.seconds, :at => "#{Time.now.hour}:#{Time.now.min+1}" do
puts "Every one second but first call at #{Time.now.hour}:#{Time.now.min}"
end
every 10.seconds do
puts "Every 10 second"
end
every 20.seconds do
puts "Every 20 second"
end
every 15.seconds do
puts "Every 15 seconds, but my task require 10 seconds"; sleep 10
# This doesn't block other jobs and your queue !!!!!!!
end
every 10.seconds, :at => [":#{Time.now.min+1}", ":#{Time.now.min+2}"] do
puts "Every 10 seconds but first call at xx:#{Time.now.min}"
end
on_error do |e|
puts "Boom raised: #{e.message}"
end
on_exit do
puts "Bye bye"
end
end
Running the example with the following code:
$ examples/sample; tail -f -n 150 examples/log/sample.log; examples/sample stop
you should see:
=> Pid not found, process seems doesn't exist!
=> Process demonized with pid 11509 with Forever v.0.2.0
[14/07 15:46:56] All jobs will will wait me for 1 second
[14/07 15:46:57] Every 10 second
[14/07 15:46:57] Every 20 second
[14/07 15:46:57] Every 15 seconds, but my task require 10 seconds
[14/07 15:47:00] Every one second but first call at 15:47
[14/07 15:47:00] Every 10 seconds but first call at xx:47
[14/07 15:47:01] Every one second but first call at 15:47
[14/07 15:47:02] Every one second but first call at 15:47
[14/07 15:47:03] Every one second but first call at 15:47
[14/07 15:47:04] Every one second but first call at 15:47
[14/07 15:47:05] Every one second but first call at 15:47
[14/07 15:47:06] Every one second but first call at 15:47
[14/07 15:47:07] Every 10 second
[14/07 15:47:07] Every one second but first call at 15:47
[14/07 15:47:08] Every one second but first call at 15:47
[14/07 15:47:09] Every one second but first call at 15:47
[14/07 15:47:10] Every 10 seconds but first call at xx:47
[14/07 15:47:10] Every one second but first call at 15:47
[14/07 15:47:11] Every one second but first call at 15:47
[14/07 15:47:12] Every 15 seconds, but my task require 10 seconds
...
[14/07 15:47:42] Every 15 seconds, but my task require 10 seconds
[14/07 15:47:42] Every one second but first call at 15:47
[14/07 15:47:43] Every one second but first call at 15:47
[14/07 15:47:44] Every one second but first call at 15:47
[14/07 15:47:45] Every one second but first call at 15:47
[14/07 15:47:46] Every one second but first call at 15:47
[14/07 15:47:47] Every 10 second
^C
=> Found pid 11509...
=> Killing process 11509...
[14/07 15:48:40] Bye bye
In foreverb we have a couple of filters, before and after, like rspec you should be able to filter before :all or before :each.
before :all do
puts "This will be ran only at start"
end
before :each do
puts "Do that before each job"
end
# ... here jobs ...
after :all do
puts "This will be ran only at shutdown"
end
after :each do
puts "Do that after each job"
end
$ foreverb help
Tasks:
foreverb help [TASK] # Describe available tasks or one specific task
foreverb list # List Forever running daemons
foreverb restart [DAEMON] [--all] [--yes] # Restart one or more matching daemons
foreverb start [DAEMON] [--all] [--yes] # Start one or more matching daemons
foreverb stop [DAEMON] [--all] [--yes] # Stop one or more matching daemons
foreverb tail [DAEMON] # Tail log of first matching daemon
foreverb update [DAEMON] [--all] [--yes] # Update config from one or more matching daemons
foreverb version # show the version number
$ foreverb list
RUNNING /Developer/src/Extras/githubwatcher/bin/githubwatcher
RUNNING /Developer/src/Extras/foreverb/examples/sample
Reading config from: /Users/DAddYE/.foreverb
$ foreverb list -m
PID RSS CPU CMD
5528 168 Mb 0.1 % Forever: /Developer/src/Extras/githubwatcher/bin/githubwatcher
5541 18 Mb 0.0 % Forever: /Developer/src/Extras/foreverb/examples/sample
$ foreverb stop foo
Do you want really stop Forever: bin/foo with pid 19538? y
Killing process Forever: bin/foo with pid 19538...
$ foreverb stop --all -y
Killing process Forever: /usr/bin/githubwatcher with pid 2824
Killing process Forever: examples/sample with pid 2836
$ foreverb start github
Do you want really start /Developer/src/Extras/githubwatcher/bin/githubwatcher? y
=> Found pid 5528...
=> Killing process 5528...
=> Process demonized with pid 14925 with Forever v.0.2.2
as for stop we allow --all and -y
$ foreverb restart github
Do you want really restart /Developer/src/Extras/githubwatcher/bin/githubwatcher? y
=> Found pid 5528...
=> Killing process 5528...
=> Process demonized with pid 14925 with Forever v.0.2.2
as for stop we allow --all and -y
$ foreverb tail github
[22/07 11:22:17] Quering git://github.com/DAddYE/lipsiadmin.git...
[22/07 11:22:17] Quering git://github.com/DAddYE/lightbox.git...
[22/07 11:22:17] Quering git://github.com/DAddYE/exception-notifier.git...
[22/07 11:22:17] Quering git://github.com/DAddYE/lipsiablog.git...
[22/07 11:22:17] Quering git://github.com/DAddYE/purple_ruby.git...
you can specify how many lines show with option -n, default is 150
This command would be helpful if you change pid log path, in this way the global config file ~/.foreverb will be update
using latest informations from yours deamons
Note that you can personalize the config file setting FOREVER_PATH matching your needs.
$ foreverb update github
Do you want really update config from /Developer/src/Extras/githubwatcher/bin/githubwatcher? y
as for stop we allow --all and -y
Bundler has the bad behavior to load Gemfile from your current path, so if your daemons (ex: githubwatcher)
is shipped with their own Gemfile to prevent errors you must insert that line:
ENV['BUNDLE_GEMFILE'] = File.expand_path('../../Gemfile', __FILE__) # edit matching your Gemfile path
I highly suggest to use fork and before filters when you are using forever with frameworks, this since running same job on our ruby will eat a lot of
ram, so the better way that I found is that:
Forever.run :fork => true do
before :each do
require '/config/boot' # here the rails/padrino environment
end
every 10.seconds, :at => ['12:00', '00:00'] do
Project.all(&:perform_long_task)
end
every 1.minute do
Account.all.map(&:send_emails)
end
end
This is similar to create a new process i.e.:
Process.fork do
require '/config/boot'
my_long_jobs
Project.all(&:perform_long_task)
end
Process.waitall
Use the following script if you want foreverb to fire up all of your daemons at boot time in Linux:
### BEGIN INIT INFO
# Provides: foreverb
# Required-Start: $local_fs $remote_fs
# Required-Stop: $local_fs $remote_fs
# Default-Start: 2 3 4 5
# Default-Stop: S 0 1 6
# Short-Description: foreverb initscript
# Description: foreverb
### END INIT INFO
# Do NOT "set -e"
DAEMON="foreverb"
USER="username"
SCRIPT_NAME="/etc/init.d/foreverb-username"
case "$1" in
start)
su -l $USER -c "$DAEMON start --all --yes"
;;
stop)
su -l $USER -c "$DAEMON stop --all --yes"
;;
restart)
su -l $USER -c "$DAEMON restart --all --yes"
;;
*)
echo "Usage: $SCRIPT_NAME {start|stop|restart}" >&2
exit 3
;;
esac
:
You’ll have to create one script per each user foreverb runs on.
After creating the file, make it executable:
chmod +x /etc/init.d/foreverb-username
and add it to the system’s boot:
RedHat:
sudo /sbin/chkconfig --level 345 foreverb-username on
Debian/Ubuntu:
sudo /usr/sbin/update-rc.d -f foreverb-username defaults
Gentoo:
sudo rc-update add foreverb-username default
To see a most comprensive app running foreverb + growl see githubwatcher gem
DAddYE, you can follow me on twitter @daddye or take a look at my site daddye.it
Copyright © 2011 Davide D’Agostino - @daddye
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 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.