Skip to content

Latest commit

 

History

History
81 lines (57 loc) · 2.7 KB

FAQ.md

File metadata and controls

81 lines (57 loc) · 2.7 KB
Raw

Workhorse FAQ

Why should I use workhorse over X?

There exist a variety of job backends for ruby, delayed_job probably being the closest one to workhorse.

Some key advantages we feel workhorse has over other Gems:

  • Workhorse is less than 500 lines of code at its core. The code is supposed to be easily readable, understandable, and modifiable.

  • Workhorse allows you to run multiple jobs simultaneously in the same process. This capability is what inspired the creation of workhorse in the first place.

We encourage you to have a look at the other projects as well and carefully figure out which one best suits your needs.

My code is not thread safe. How can I use workhorse safely?

Job code that is not thread safe can not be executed safely in multiple threads of the same process. In these cases, set the pool_size to 1 and, if you still want to execute multiple jobs simultaneously, the daemon count to a number greater than 1:

Workhorse::Daemon::ShellHandler.run count: 5 do
  Workhorse::Worker.start_and_wait(pool_size: 1)
end

I'm using jRuby. How can I use the daemon handler?

As java processes in general can not be forked safely, the daemon handler provided with this Gem does not support jRuby platforms.

If your jRuby application consists of one single application process, it is recommended to just start the job backend in the same process:

worker = Workhorse::Worker.new(pool_size: 5)
worker.start

This code is non-blocking, which means that it will run as long as the process is up. Make sure to trap INT and TERM and call worker.shutdown when the process stops.

If you have multiple application processes though, you may want to start the worker in a separate process. For this purpose, adapt the startup script bin/workhorse.rb so that it is blocking:

Workhorse::Worker.start_and_wait(pool_size: 5)

This can then be started and "daemonized" using standard linux tools.

I'm getting random autoloading exeptions

Make sure to always start the worker in production mode, i.e.:

RAILS_ENV=production bin/workhorse.rb start

I'm getting "No live threads left. Deadlock?" exceptions

Make sure the Worker is logging somewhere and check the logs. Typically there is an underlying error that leads to the exception, e.g. missing migration in production mode.

Why does workhorse not support timeouts?

Generic timeout implementations are a dangerous thing in Ruby. This is why we decided against providing this feature in Workhorse and recommend to implement the timeouts inside of your jobs - i.e. via network timeouts.