Metrics initializes four times

[InboxViktorV]

Hi everyone. Is it a bug, or expected behaviour? The roadrunner initialize kernel four times -> metrics also initializes four times -> as a result system throws errors:

weather    | 2021/12/15 05:38:42    INFO    metrics         declaring new metric    {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    INFO    metrics         metric successfully added   {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    INFO    metrics         declaring new metric    {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    ERROR   metrics         metric with provided name already exist {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    INFO    metrics         declaring new metric    {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    ERROR   metrics         metric with provided name already exist {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    INFO    metrics         declaring new metric    {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}
weather    | 2021/12/15 05:38:42    ERROR   metrics         metric with provided name already exist {"name": "metno_provider_req_duration", "type": "histogram", "namespace": ""}

How to avoid those errors? Or somehow fix them?

• symfony: 5.3
• baldinof/roadrunner-bundle: 2.1.3
• roadrunner: 2.4.2

[Baldinof]

Hi!

I don't know if it's expected but it's an upstream issue.

If you folllow the documentation here you get the same logs. Despite the messages, everything work well.

[eightfourseventwo]

I'm guessing your CPU has 4 cores? I have the amount of workers set to 10 and I get the error 10 times. I suppose that is to be expected because each RR worker runs in isolation so it starts your app again for each worker.

You can get around this by defining the metrics in rr.yaml instead of the bundle configuration.

I suppose there is no way to fix this. It could perhaps just be a README update.