Filip Petkovski

Thanks for the tip Matthias, should be fixed now 👍

Any suggestions to help move this PR forward?

I am not completely convinced this is strictly caused by kubernetes discovery. Target groups for the endpoints role are always generated from scratch, and if Prometheus is getting stale reads...

@jutley if you still have the faulty Kubernetes endpoint, you can try checking if there are any `notReadyAddresses` in the `subsets` field. These IPs will not be shown in the...

To me it looks like `notReadyAddresses` are treated the same way as regular addresses, with the exception that the `kubernetes_endpoint_ready` label is set to false: https://github.com/prometheus/prometheus/blob/e239e3ee8b13b51b0f791a199813a14f74600a7e/discovery/kubernetes/endpoints.go#L304-L306 Not sure if this...

Thanks for this contribution. If I understand correctly, this PR allows specifying a label allowlist for all reasources globally. Instead of adding a new flag, could we extend the syntax...

/remove-lifecycle stale

Ideally, there should be a dedicated endpoint for the the readiness probe built into KSM. Since that's not there, I don't know if there is a right or wrong endpoint...

I agree, we should at least document the options somewhere to avoid confusion.

As an alternative solution, could you drop unwanted labels at scrape time with metric relabel configs?