implement report_after_job_retries in ActiveJob adapter

[modosc]

we are using sentry-rails to report errors from ActiveJob (which is running on sidekiq). we want the ability to only report on errors when retries are exhausted. there doesn't seem to be a way to do this directly in the ActiveJob integration (sentry-sidekiq, sentry-delayed_job, and sentry-resque all support this configuration option).

is it possible to implement this directly in the ActiveJob adapter?

[sl0thentr0py]

Yes, we should implement :report_after_job_retries for active job too, feel free to make a PR if you want it quicker!

[modosc]

@sl0thentr0py i gave it a shot in #2500 but ran into several issues:

1. tests work locally but not in CI or in CI but not locally, similar to Fix ActiveJob tests for Rails 7.2 and 8.0 #2487
2. rails 5.0 didn't support retry_on and i'm unclear on how you expect that to be handled (print a deprecation warning? throw an error?).

anyway - i'm closing the pr out but i'll make sure the branch still exists. i think the approach is valid (and it seems to work) but i can't spend anymore time figuring out the tests.

[sl0thentr0py]

@modosc thanks a lot! I will reopen the PR and get it into a mergeable state myself. You don't have to do further work on it.