Thanks for this introduction, Jeffrey.

As you mentioned failed jobs in your lesson, maybe you have some ideas on our problem below.

( If this is not the right place to raise this sort of question, just let me know. )

In our setup, we have multiple servers running our Horizon app as pure workers. Only one master has access to the DB.

Failed jobs are supposed to end up on Horizon/Redis as well as in the DBs failed-jobs table. I understand this is because one might want to preserve them outside of Redis, because Redis jobs might get flushed.

In our case, loosing jobs is not too big of an issue, because we can trigger our jobs again without any issues.

Checking which ones failed and understanding why seems all possible with the Horizon dashboard. So no need for the DB failed_jobs.

Yet, even if we remove the failed_jobs queue-configuration, we still see exceptions in the log file telling us that failed_jobs can’t be found. This adds unnecessary clutter to the log.

Do you know of a way how to avoid this?

A UI For Your Redis-Backed Queues