Rosa Gutierrez

This is [now better documented in the README](https://github.com/rails/solid_queue?tab=readme-ov-file#workers-dispatchers-and-scheduler), with the addition of a new environment variable to completely skip the scheduler as well, added in https://github.com/rails/solid_queue/pull/580.

Oh wait, I got this the other way around. You wanted to have just the scheduled, instead of no scheduler 😅 Going to reopen for that `--only-scheduler` option.

I think the supervisor might need 3 connections as well: one for the main thread, one for the process maintenance task and one for its own heartbeat. The dispatcher, with...

No, the scheduler should have enough with 2 (heartbeat + task scheduling).

Going to close this one as the estimation and error shown should be ok regardless of whether you configure workers or not.

Hmm... interesting that you get that query. That code results in the following query for us: ```sql DELETE FROM `solid_queue_jobs` WHERE `solid_queue_jobs`.`finished_at` IS NOT NULL AND `solid_queue_jobs`.`finished_at` < '2024-11-04 19:42:00.055618'...

Ahh, I realised why this is the case. PostgreSQL doesn't support `LIMIT` on `DELETE` queries 😅 I don't have a good alternative for this one at the moment, I'm afraid....

Huh, yes, but I'm curious to know when you reach this line with `process` being `nil` 🤔 In theory, this should never be executed before the supervisor has been registered....

Thanks a lot for these extra logs, @HashNuke! These are very helpful as I haven't been able to reproduce this 🙏

Oh! @tahirwaseer, that's very interesting because you're hitting a different error (besides this one), that SEGFAULT in the supervisor. Do you know where that comes from exactly? It needs to...