In case it's relevant, we run a single-replica pod that exposes this and other kinds of metrics that don't belong to a specific pod. We like this a bit better, it avoids having the same metric (queue latency is a "global" metric, not specific to any pod) exposed by multiple targets.
This PR keeps the existing default, with the all_queues option enabling the desired behavior.
We'd appreciate if this was considered for incorporation 😄
Hi! I added the the Sidekiq Queue instrumentation back in https://github.com/discourse/prometheus_exporter/pull/114 and when trying to upgrade this gem on our apps I found https://github.com/discourse/prometheus_exporter/pull/126, which breaks our setup. We'd really prefer it if we could get metrics on all queues, not just the ones from the current process.
In case it's relevant, we run a single-replica pod that exposes this and other kinds of metrics that don't belong to a specific pod. We like this a bit better, it avoids having the same metric (queue latency is a "global" metric, not specific to any pod) exposed by multiple targets.
This PR keeps the existing default, with the
all_queues
option enabling the desired behavior.We'd appreciate if this was considered for incorporation 😄