如何启用 Prometheus/Grafana 以显示我的 rabbitmq 指标？ 目前没有显示数据

Question

I am using the prometheus-community stack.

I have a rabbitmq pod running in my K8s cluster, I have enabled rabbitmq_prometheus from within the pod so I am able to see the time series data of of my rabbitmq from localhost:15692. I have also installed a dashboard in grafana (this enter link description here ) in order to be able to see this data. However I get No Data showing in my RabbitMQ dashboard. as shown below. .

I have also installed prometheus-community/prometheus-rabbitmq-exporter, saved the values file locally so that i can make changes to it, currently it looks like this below

prometheus:
  monitor:
    enabled: true
    additionalLabels: {}
    interval: 15s
    namespace: []
  rules:
    enabled: false
    additionalLabels: {}

I have applied the values file to my helm release however when i do an inspect on it, i still get enabled: false I am not sure what I am missing or doing wrong.

How do i show my rabbitmq metrics in grafana?

Answer 1

I've not really understood the steps you took to make it work, but I understand your problem is not being able to import RabbitMQ metrics in Prometheus. So here's the solution.

1.Deploy prometheus using kube-prometheus-stack helm chart . As the helm chart comes with "prometheus operator", it's best to use PodMonitor and/or ServiceMonitor CRD's as they provide far more configuration options. Here's some documentation around that.

Keep default values (as the purpose is just to import RabbitMQ metrics) for most of the stuff, except that set "prometheus.prometheusSpec.serviceMonitorSelector.matchLabels" with a label value. Something like this

serviceMonitorSelector:
  matchLabels:
    monitoring-platform: core-prometheus

2.Deploy RabbitMQ using the bitnami rabbitmq helm chart . Again, as the purpose here is just to expose metrics for prometheus to scrape, you can use default values for much of it. Here's the values file:

## Prometheus Metrics
metrics:
  enabled: true
  plugins: 'rabbitmq_prometheus'
  podAnnotations:
    prometheus.io/scrape: 'true'
    prometheus.io/port: '{{ .Values.service.metricsPort }}'

  ## Prometheus Service Monitor
  ## ref: https://github.com/coreos/prometheus-operator
  serviceMonitor:
    enabled: true
    interval: 30s
    additionalLabels:
      monitoring-platform: core-prometheus

So using the serviceMonitor label (which matches the label of prometheus serviceMonitorSelector ) Prometheus identifies RabbitMQ as a target.