Kubernetes Monitoring

is done with prometheus. Webinterface is grafana. https://prometheus.acc.gsi.de/grafana/

Service Metrics

Prometheus is doing auto discovery of pods and services it should monitor. For all pods a basic monitoring (cpu, memory) is active. If the pod or service is instrumented and exposes additional information (java heap, etc) on a http url prometheus can collect this data. The pod or service must have the following annotations
metadata:
  annotations:
    prometheus.io/scrape: 'true'
    prometheus.io/port: '80'
    prometheus.io/path: '/metrics'
    prometheus.io/scheme: 'http'

  • prometheus.io/path: '/metrics'annotation prometheus.io/scrape declares that this service should be monitored. the value is a string. All annotations are always strings.
  • path /metrics is the default, overwrite where necessary for example with /actuator/prometheus
  • scheme can switch between http and https.
  • port is the port where prometheus can query for metrics. For services this defaults to the service ports. If a service exposes multiple ports, for example one http port and one rmi port, overwrite it to only scrape the http port.

-- ChristophHandel - 21 Jun 2022
Topic revision: r5 - 26 Jul 2022, HannoHuether
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback