Horst Gutmann

Hi :) To quote the README: > This has only been tested with Redis so far. Could you perhaps give some more details regarding the behavior you're experiencing?

I'm not completely sure what you mean. You could just expose custom metrics from the application itself. Can you perhaps give a concrete example?

Do you get any output when you run `celery-prometheus-exporter --help`?

Hi :) Sorry for the late response! Yes, the exporter tries to connect to the broker during the startup-phase but doesn't log anything until it has succeeded there. I've added...

Do you get any error messages or do the metrics simply not update?

So what you're proposing is deprecating `celery_tasks_by_name` and `celery_tasks` in favor or `celery_tasks_count` which in turn should have a `state` and `name` dimension?

I'd be +1 for for `_total` eventhough it is a gauge as this would be the same behaviour as with a histogram, wouldn't it? +1 on all your proposals 😄

Sorry, didn't have the time to look into your request :( Did you already solve it on your own?

I‘m currently afk, but I’d guess that the information about the queue is also included in the submitted event. You could then add the queue as additional label in the...

Hi :) IMHO for this example it doesn't really matter esp. as each service basically customizes the executable and so command would also be fitting :)