[k8s-stack] Default grafana dashboard "Kubernetes / Compute Resources / Pod" not working

[nikmel2803]

I ran into a problem that the "Kubernetes / Compute Resources / Pod" dashboard does not display RAM consumption:

During the research, I found that the metric does not have an image label, which is the request:

I also found the following config in values (link):

    metricRelabelConfigs:
      - action: labeldrop
        regex: (uid)
      - action: labeldrop
        regex: (id|name|image)
      - action: drop
        source_labels: [__name__]
        regex: (rest_client_request_duration_seconds_bucket|rest_client_request_duration_seconds_sum|rest_client_request_duration_seconds_count)

It can be seen that the label image is removed, which is why the dashboard does not work.

If you change this config to the following, then everything works:

        metricRelabelConfigs:
          - action: labeldrop
            regex: (uid)
          - action: labeldrop
            regex: (id|name)
          - action: drop
            source_labels: [ __name__ ]
            regex: (rest_client_request_duration_seconds_bucket|rest_client_request_duration_seconds_sum|rest_client_request_duration_seconds_count)

Graph displayed:

Does this mean the default values need to be changed to fix this issue

[f41gh7]

Thanks for reporting, will fixed. Workaround for now - change default value.

[sergeyshevch]

@f41gh7 Hi! Any progress on this?

[f41gh7]

Should be fixed at the latest release

[nikmel2803]

thanks! I confirm the fix