Запуск Prometheus (мониторинг) и Grafana (визуализация) в kuberntes cluster с помощью heml

Делаю: 11.04.2019

По материалам из видео индуса:

https://www.youtube.com/watch?v=CmPdyvgmw-A&list=PL34sAs7_26wNBRWM6BDhnonoA5FMERax0&index=27

prometheus and grafana 01

Подготовили кластер и окружение
Подняли Dynamic NFS
Инсталлировали helm

UPD. Heml2 выпилен (как ненужное), предлагаю попробовать Helm3 как здесь.

Устанавливаем prometheus

// если нужно посмотреть параметры по умолчанию
$ helm inspect values stable/prometheus > /tmp/prometheus.values

// я просто создаю файл, чтобы не копаться в куче кода
$ vi /tmp/prometheus.values.final

 server:
    service:
      nodePort: 32323
      type: NodePort

$ helm install stable/prometheus --name myprometheus --values /tmp/prometheus.values.final --namespace prometheus

$ kubectl get all -n prometheus
NAME                                                   READY   STATUS    RESTARTS   AGE
pod/myprometheus-alertmanager-55f5594766-kjzcp         2/2     Running   0          4m19s
pod/myprometheus-kube-state-metrics-668df79bd8-dgqpm   1/1     Running   0          4m19s
pod/myprometheus-node-exporter-5vx6c                   1/1     Running   0          4m19s
pod/myprometheus-node-exporter-v57qb                   1/1     Running   0          4m19s
pod/myprometheus-pushgateway-56fccfb787-mkssw          1/1     Running   0          4m19s
pod/myprometheus-server-6b48bd7c95-74dfj               2/2     Running   0          4m19s

NAME                                      TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
service/myprometheus-alertmanager         ClusterIP   10.110.83.127    <none>        80/TCP         4m19s
service/myprometheus-kube-state-metrics   ClusterIP   None             <none>        80/TCP         4m19s
service/myprometheus-node-exporter        ClusterIP   None             <none>        9100/TCP       4m19s
service/myprometheus-pushgateway          ClusterIP   10.108.225.88    <none>        9091/TCP       4m19s
service/myprometheus-server               NodePort    10.102.139.197   <none>        80:32323/TCP   4m19s

NAME                                        DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR   AGE
daemonset.apps/myprometheus-node-exporter   2         2         2       2            2           <none>          4m19s

NAME                                              READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/myprometheus-alertmanager         1/1     1            1           4m19s
deployment.apps/myprometheus-kube-state-metrics   1/1     1            1           4m19s
deployment.apps/myprometheus-pushgateway          1/1     1            1           4m19s
deployment.apps/myprometheus-server               1/1     1            1           4m19s

NAME                                                         DESIRED   CURRENT   READY   AGE
replicaset.apps/myprometheus-alertmanager-55f5594766         1         1         1       4m19s
replicaset.apps/myprometheus-kube-state-metrics-668df79bd8   1         1         1       4m19s
replicaset.apps/myprometheus-pushgateway-56fccfb787          1         1         1       4m19s
replicaset.apps/myprometheus-server-6b48bd7c95               1         1         1       4m19s

$ kubectl get pvc -n prometheus
NAME                        STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS          AGE
myprometheus-alertmanager   Bound    pvc-c34dbaed-5c72-11e9-bbe3-525400261060   2Gi        RWO            managed-nfs-storage   4m53s
myprometheus-server         Bound    pvc-c34f1d62-5c72-11e9-bbe3-525400261060   8Gi        RWO            managed-nfs-storage   4m53s

$ kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM                                  STORAGECLASS          REASON   AGE
pvc-c34dbaed-5c72-11e9-bbe3-525400261060   2Gi        RWO            Delete           Bound    prometheus/myprometheus-alertmanager   managed-nfs-storage            4m58s
pvc-c34f1d62-5c72-11e9-bbe3-525400261060   8Gi        RWO            Delete           Bound    prometheus/myprometheus-server         managed-nfs-storage            4m58s

http://node1:32323

prometheus and grafana 02

Устанавливаем Graphana

// если нужно посмотреть параметры по умолчанию
$ helm inspect values stable/grafana > /tmp/grafana.values

// я просто создаю файл, чтобы не копаться в куче кода
$ vi /tmp/grafana.values.final

service:
  nodePort: 32324
  type: NodePort

adminUser: admin
adminPassword: admin

persistence:
  enabled: true

$ helm install stable/grafana --name mygrafana --values /tmp/grafana.values.final --namespace grafana

$ kubectl get all -n grafana
NAME                            READY   STATUS            RESTARTS   AGE
pod/mygrafana-d47947d89-cdmfx   0/1     PodInitializing   0          15s

NAME                TYPE       CLUSTER-IP    EXTERNAL-IP   PORT(S)        AGE
service/mygrafana   NodePort   10.99.20.18   <none>        80:32324/TCP   15s

NAME                        READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/mygrafana   0/1     1            0           15s

NAME                                  DESIRED   CURRENT   READY   AGE
replicaset.apps/mygrafana-d47947d89   1         1         0       15s

$  kubectl get pvc -n grafana
NAME        STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS          AGE
mygrafana   Bound    pvc-6cdf745f-5c74-11e9-bbe3-525400261060   10Gi       RWO            managed-nfs-storage   79s

http://node1:32324

prometheus and grafana 03

prometheus and grafana 04

prometheus and grafana 05

prometheus and grafana 06

prometheus and grafana 07

Подбираем под себя:
https://grafana.com/dashboards/8588

prometheus and grafana 08

prometheus and grafana 09

prometheus and grafana 10

Удаление из kubernetes

$ helm list
$ helm delete myprometheus --purge
$ helm delete mygrafana --purge