How can we detect a quorum failure using metrics?

[veera-chenna]

Hi All,

Am looking for the alerting on quorum failures. But when checked at the metrics, couldn't find the correct one. Is there a way to know that quorum failure?

Thanks in advance

[michaelklishin]

Queue status is one of queue metrics, and so is the number of replicas it has online.

Note that with a large number of queues you very likely will have to revert to using aggregates and not per-object (per-queue) metrics. The doc section should explain why.

Before upgrades, there is a very specific health check that feels relevant.

[veera-chenna]

@michaelklishin , We are using it with less number of queues for now. If needed will retrieve it with metrics family using /metrics/detailed endpoint. But cannot find the Queue status metric. Could you please help with the metric name?