Description of problem:

Currently, we can't configure the worker_timeout used by puma workers. The default of 60s is used.

Changing the worker_timeout would allow users to workaround specific issues happening during peak moments. Here's an example where increasing the timeout would help:


~~~
Oct 14 06:08:45 satellite.example.com foreman: [1290] ! Terminating timed out worker (worker failed to check in within 60 seconds): 2420
~~~

Version-Release number of selected component (if applicable):

Actual results:

Not possible to change the default 60s timeout in a supported way.

Expected results:
Have an installer option to define the worker_timeout.

Additional info:

The timeout can be defined with the following line on the file /usr/share/foreman/config/puma/production.rb

~~~
worker_timeout 120
~~~

My suggestion is to add a line on the file /usr/share/foreman/config/puma/production.rb that will read a variable to be defined on the systemd unit, like we do for the number of workers, and min/max threads.

Like this:

~~~
worker_timeout ENV.fetch('FOREMAN_PUMA_WORKER_TIMEOUT', 60).to_i
~~~

Then, setting the value FOREMAN_PUMA_WORKER_TIMEOUT on the systemd unit would be enough:

~~~
# cat /etc/systemd/system/foreman.service.d/installer.conf 
[Service]
User=foreman
Environment=FOREMAN_ENV=production
Environment=FOREMAN_HOME=/usr/share/foreman
Environment=FOREMAN_PUMA_THREADS_MIN=5
Environment=FOREMAN_PUMA_THREADS_MAX=5
Environment=FOREMAN_PUMA_WORKERS=6
Environment=FOREMAN_PUMA_WORKER_TIMEOUT=120 <============ new variable to define the timeout
~~~