Adin Hodovic

@kittywaresz @iqbalaydrus The newest release will prune metrics for a worker that goes offline after 10 minutes by default (adjustable). Should result in way less active time series.

I hit the same issue when using "fieldsets" instead of "fields" in the inline, buttons don't render since `_request` is None

hey @lostanlen, I believe the dependency was renamed in the PyPi registry and the newer versions were only found under `soundfile`. ``` Breaking Changes The soundfile module has evolved rapidly...

Can you re run the workflow, think the OS mp3 dependency was missing.

The cleanup of metrics happens when workers are timed out, it helps to lower label cardinality but this can be disabled. However, it happens by default after a timeout of...

> Are your workers picking up the tasks and executing them? The metrics above just indicate that the client has sent tasks to the queue (note `celery_task_sent_total`), but actually none...

Try running it with debug logging.

> @adinhodovic > > These are the logs. It registers the heartbeat but still times out the worker. I have the workers sending a heartbeat every 10 seconds. > >...

If your workers go offline (rotate) metrics should be quickly cleaned up. Works fine for us with up to ~100 pods. On new releases all metrics get cleaned quite quickly....

v0.8.0 and above cleanup metrics of timed out / rotated workers. It might be that an exporter running longer time in environments with high cardinality on worker names (e.g pod...