Powered by GitBook

Prometheus in Kubernetes

Prometheus is a monitor and alert toolkit; it can set up on any platform.

There are some charts for Prometheus to set up in Kubernetes:

Prometheus-operator project has integrated a lot of exports and Grafana for Kubernetes, so it is easy to use.

https://github.com/coreos/prometheus-operator/tree/master/helm

Another is Kubernetes charts project, now we start set up Prometheus via charts project of Kubernetes,

Install steps:

clone kubernetes charts project, If you install prometheus by help online, the service account is default

git clone https://github.com/kubernetes/charts.git

Create persistent volume for Prometheus server, Alert manager, and Grafana.

apiVersion: v1
kind: PersistentVolume
metadata:
  name: {{prometheus-alert-pv}}
spec:
  capacity:
    storage: 10Gi
  accessModes:
    - ReadWriteOnce
storageClassName: {{ prometheus-alert}}
  nfs:
    server: {{192.168.1.100}}
    path: {{"/data/prometheus/alert"}}

Create Prometheus via helm

There are some configures need to set if you don't want use default value.

Because I adopt NFS as persistent volume, so we need to set storage class name as configures.

helm install ./stable/prometheus --name k8s-prometheus \
    --set alertmanager.persistentVolume.storageClass=prometheus-alert \
    --set server.persistentVolume.storageClass=prometheus-server \
    --namespace monitoring

Check Prometheus server is started success.

> kubectl get po -n=monitoring
NAME                                                            READY     STATUS    RESTARTS   AGE
k8s-prometheus-prometheus-alertmanager-784f7946bf-ffzz7         2/2       Running   0          1m
k8s-prometheus-prometheus-kube-state-metrics-675dbc4dd7-hgjz7   1/1       Running   0          1m
k8s-prometheus-prometheus-node-exporter-bnnf8                   1/1       Running   0          1m
k8s-prometheus-prometheus-node-exporter-xnnb9                   1/1       Running   0          1m
k8s-prometheus-prometheus-pushgateway-84b4bb947f-w9fqx          1/1       Running   0          1m
k8s-prometheus-prometheus-server-5fcfd56cd5-gmg48               2/2       Running   0          1m

> kubectl logs k8s-prometheus-prometheus-server-cd5c844d8-mhwx7 -n=monitoring
..
level=info ts=2018-03-12T06:07:58.827303886Z caller=kubernetes.go:191 component="discovery manager notify" discovery=k8s msg="Using pod service account via in-cluster config" 
level=info ts=2018-03-12T06:07:58.827659506Z caller=main.go:486 msg="Server is ready to receive web requests." 
level=info ts=2018-03-12T06:07:58.827695217Z caller=manager.go:59 component="scrape manager" msg="Starting scrape manager..."

Install Grafana

helm install ./stable/grafana --name k8s-grafana \
    --set server.persistentVolume.storageClass=prometheus-grafana \
    --set server.adminPassword=abcd@1234 \
    --set server.service.type=NodePort \
    --set server.service.nodePort=30266 \
    --namespace monitoring

Check Grafana is running

> kubectl get po -n=monitoring -l app=k8s-grafana-grafana
k8s-grafana-699fff8bb8-pqp5z   1/1       Running   0          <invalid>

Connect Grafana by browser 192.168.1.100:30266
Add data source to Grafana, URL enter kubernetes service domain name

Setting your dashboard to start monitor!

Reference:

results matching ""

No results matching ""