colinsurprenant / raad Goto Github PK

I can't start a simple service. Here is the output:

$ ruby lib/scheduler/runner.rb start

Raad service wrapper v0.5.0 starting
[4738:INFO] 2013-02-03 21:43:30 :: starting runner service in development mode
[4738:INFO] 2013-02-03 21:43:30 :: stopping runner service
[4738:INFO] 2013-02-03 21:43:30 :: >> Raad service wrapper stopped

Hi,

It would be great to have a restart command ? Do you plan to support it ?

I have a background job in which I need to access ActiveRecord models. If I start it without '-d', things work fine - I can access the database through AR models; but once I put '-d' in the command, the script doesn't work anymore.

I wonder if anybody has seen this before?

sh rvm_setup.sh
yields
rvm_setup.sh: 13: [[: not found rvm_setup.sh: 13: [[: not found ERROR: An RVM installation was not found.

Changing the [[]]'s to [] seems to help in some cases.

Also,
function waitline {...}
yields
function: not found

Changing it to
assert() {...}
lets it work

Can't get it working.

Here is a few more details about my approach: http://stackoverflow.com/questions/11457346/using-raad-and-eventmachine-in-cli-gem

Is it possible to use it like this, or am I using the wrong tool for this job?

Hello! Thanks for this gem!

I have an existing Thor-based command-line interface that I want to have start|stop to daemonize, but also needs a variety of other CLI options. The way the Raad gem is currently set up with bootstrapping, it appears to assume that the only command-line options are start/stop. It also seems to makes assumptions based on the current filename for service initiation that, in my case, would not make sense.

I can hack what I need together with some ideas from this gem, but this is what I would have loved to see:

in bin/my_app:

require 'raad'
require 'thor'

class MyApp < Thor

  default_task :start

  desc "Start My App", "Starts App Server"
  def start

    daemonize = true

    if daemonize
      Raad::Runner.new('ruby path_to_my_service_starter.rb').daemonize
    else
      MyServiceStarter.run
    end
  end

  desc "Some other option I need"
  def my_option
  end

end

Or something to this effect. Not really an issue, just a suggestion. Thanks again.

In my application I did setup the environment in the initialize method.
So i changed it to 'my_env = Raad.env'
But it's always the default 'development' value.
If i move the statement to the start method it return the correct value (from the -e option)

any idea ?

first ^C will call the stop method and if the start thread still hangs, it will wait up to the defined timeout (default 60s) to force kill. a second ^C should provide a force kill shortcut.

it could be really helpful to provide a template /etc/init.d script of include a generator for it.

it could be useful to have an option to set the daemon working directory

I'm using raad along with jruby-jms to consume queues from activemq, it would be handy for deployment if I could specify where the log and pid files would be

postpone service class initialization after services setup to allow usage of logger in particular in the service initialize method.

currently there is no way to completely disable logging: in console mode by default logging is in the console and in daemon mode, logging is in the default log file.

Just a heads up and not directly related to RAAD.

Whenever RAAD is initiated, the following warning is generated:

.../gems/log4r-1.1.10/lib/log4r/formatter/patternformatter.rb:126 warning: Ambiguous first argument; make sure.

I sent in a pull request to fix it it in log4r: nbrochu/log4r@e54e97146c474cabc56e1da616b65393cc2b91bf

stop_service should only be called once, regardless if SIGINT or other signals are sent multiple times

NameError: uninitialized constant MyServiceClass
  const_missing at org/jruby/RubyModule.java:2715
         (eval) at (eval):1
    module_eval at org/jruby/RubyModule.java:2354
           run! at /home/jenko/.rvm/gems/jruby-1.7.0.preview1/gems/raad-0.5.0/lib/raad/bootstrap.rb:43
           Raad at /home/jenko/.rvm/gems/jruby-1.7.0.preview1/gems/raad-0.5.0/lib/raad/bootstrap.rb:63

this is when there is a mismatch between the command line service name and the actual class name which must match the camelcasing of the service name.

When the pid file is empty pid and read_pid_file methods return 0, because en empty string is converted to zero by to_i. And since process with pid 0 always exists in the system the daemon can't be started or restarted. That also happens for other non-integer content of the pid file

Empty pid file can occur when there's no space left on device. Additionally, number of bytes could be checked to distinguish this case (and maybe delete the pid file?).

Sorry if this is stupid question, but I am trying to pack a daemon in a JAR using warble. The script runs as expected in jruby, but once packed using warble and running:

java -jar Script.jar start

produces no output, no log, and just exists gracefully without any warning/error. how can I produce some logs to check why the daemon is not firing up?

Thanks in advance.

It would be nice to pass parameters to the service.

ex: before radd I did run the ruby with "ruby GetResponseByPop3.rb test"
with raad I run it with "ruby GetResponseByPop3.rb start"

but it would be nice to add -- test to pass my parameters

any idea , or I am wrong ?