Prometheus

Exporting Metrics

If you have an existing Prometheus cluster, it is very easy to export Linkerd’srich telemetry data to your cluster. Simply add the following item to yourscrape_configs in your Prometheus config file (replace {{.Namespace}} withthe namespace where Linkerd is running):

  1. - job_name: 'linkerd-controller'
  2.   kubernetes_sd_configs:
  3.   - role: pod
  4.     namespaces:
  5.       names: ['{{.Namespace}}']
  6.   relabel_configs:
  7.   - source_labels:
  8.     - __meta_kubernetes_pod_label_linkerd_io_control_plane_component
  9.     - __meta_kubernetes_pod_container_port_name
  10.     action: keep
  11.     regex: (.*);admin-http$
  12.   - source_labels: [__meta_kubernetes_pod_container_name]
  13.     action: replace
  14.     target_label: component
  16. - job_name: 'linkerd-proxy'
  17.   kubernetes_sd_configs:
  18.   - role: pod
  19.   relabel_configs:
  20.   - source_labels:
  21.     - __meta_kubernetes_pod_container_name
  22.     - __meta_kubernetes_pod_container_port_name
  23.     - __meta_kubernetes_pod_label_linkerd_io_control_plane_ns
  24.     action: keep
  25.     regex: ^linkerd-proxy;linkerd-metrics;{{.Namespace}}$
  26.   - source_labels: [__meta_kubernetes_namespace]
  27.     action: replace
  28.     target_label: namespace
  29.   - source_labels: [__meta_kubernetes_pod_name]
  30.     action: replace
  31.     target_label: pod
  32.   # special case k8s' "job" label, to not interfere with prometheus' "job"
  33.   # label
  34.   # __meta_kubernetes_pod_label_linkerd_io_proxy_job=foo =>
  35.   # k8s_job=foo
  36.   - source_labels: [__meta_kubernetes_pod_label_linkerd_io_proxy_job]
  37.     action: replace
  38.     target_label: k8s_job
  39.   # __meta_kubernetes_pod_label_linkerd_io_proxy_deployment=foo =>
  40.   # deployment=foo
  41.   - action: labelmap
  42.     regex: __meta_kubernetes_pod_label_linkerd_io_proxy_(.+)
  43.   # drop all labels that we just made copies of in the previous labelmap
  44.   - action: labeldrop
  45.     regex: __meta_kubernetes_pod_label_linkerd_io_proxy_(.+)
  46.   # __meta_kubernetes_pod_label_linkerd_io_foo=bar =>
  47.   # foo=bar
  48.   - action: labelmap
  49.     regex: __meta_kubernetes_pod_label_linkerd_io_(.+)

That’s it! Your Prometheus cluster is now configured to scrape Linkerd’smetrics.

Linkerd’s proxy metrics will have the label job="linkerd-proxy". Linkerd’scontrol-plane metrics will have the label job="linkerd-controller".

For more information on specific metric and label definitions, have a look atProxy Metrics.

原文: https://linkerd.io/2/observability/prometheus/