zarqman

I'm seeing duplicate cron executions in our production environment too. Confirmed on v3.6.2, but it goes back into v2.x. We use `preserve_job_records: :on_unhandled_error`. In testing, it seems like the issue...

One more thought: given our use of `preserve_job_records: :on_unhandled_error`, setting `cleanup_preserved_jobs_before_seconds_ago` to 5 minutes is less than ideal. We currently have it set to several days so we have time...

@bensheldon It's "more than expected", but not operationally dangerous. With 5 goodjob daemons, it runs 5x at each interval. Given that it runs against an indexed field, I wouldn't anticipate...

This sparked some thoughts--hope it's okay to wade in. 😄 Don't know if it applies in @mitchellhenke's case, but I believe Postgres dedupes `NOTIFY`s when inside a transaction and the...

I've seen a couple more instances of this. It turns out jobs only sometime report Running. Other times they report Queued. So, the reported state may be more of a...

@bensheldon, I've explored this too and ran into the same issue with `logger=` not working as expected. If I recall correctly, it's because logging actually happens via LogSubscriber instrumentation which...

FWIW, just happened to see that SolidQueue is using `silence`, so that should be viable: https://github.com/basecamp/solid_queue/blob/main/lib/solid_queue/processes/poller.rb

@eric-christian, rails/rails#44127 was the fix for the mentioned issue. As you noted, it's only available in 7.1. However, there's a monkeypatch listed in that PR that works on 6.1 and...

Thanks for the quick reply. I tested the potential patch for `http1/server.rb` and unfortunately it did not help (or seem to change anything). I'll work on some kind of repro...

Here's a repro example. This appears to be an issue with graceful stop not actually stopping the server when it's idle and waiting for a keep-alive request, as somewhat discussed...