使用helm安装prometheus-operator

爱被打了一巴掌 2022-02-26 10:33 247阅读 0赞

下载 https://kubernetes-charts.storage.googleapis.com/index.yaml文件，因为你懂得的原因，很多情况下不能直接访问，需要使用一定技巧。helm的搭建方法参考[https://blog.csdn.net/liukuan73/article/details/79319900][https_blog.csdn.net_liukuan73_article_details_79319900]

*  打开index.yaml文件，找到[prometheus-operator的配置包，比如https://kubernetes-charts.storage.googleapis.com/prometheus-operator-5.0.3.tgz][prometheus-operator_https_kubernetes-charts.storage.googleapis.com_prometheus-operator-5.0.3.tgz]，并将其下载下来
 *  将其拷贝到/root/.helm/repository/local/下，并且将其解压缩，编辑values.yaml文件，将其中所有repository项目的url改为阿里云的源。并且，需要将requirements.yaml中提到的依赖配置包也一同下载，其中有镜像不能直接docker pull获取到，也要将其改为其他源，并修改value.yml，重新打包放入/root/.helm/repository/local/
 *  执行如下命令，重新生成本地的源。如果提示端口占用，需要将之前的程序kill掉

nohup  helm serve --address 0.0.0.0:8879 --repo-path /root/.helm/repository/local/ > /dev/null &

*  之后，执行 helm  repo list，提示如下内容说明正确：

NAME     	URL                                             
    stable   	https://kubernetes-charts.storage.googleapis.com
    local    	http://127.0.0.1:8879/charts                    
    monocular	https://helm.github.io/monocular                
    bitnami  	https://charts.bitnami.com/bitnami

*  最后执行：

kubectl delete crd prometheuses.monitoring.coreos.com
    kubectl delete crd prometheusrules.monitoring.coreos.com
    kubectl delete crd servicemonitors.monitoring.coreos.com
    kubectl delete crd alertmanagers.monitoring.coreos.com
    helm install local/prometheus-operator   --debug   --name  prometheus-operator  --namespace=default  -f values.yaml

将会使用我们更改过的源，来安装。可以参考[https://github.com/helm/charts/tree/master/stable/prometheus-operator][https_github.com_helm_charts_tree_master_stable_prometheus-operator]，其中grafana的初始化登录密码就在该页面中。

如果不执行kubectl delete相关命令，会提示

failed: customresourcedefinitions.apiextensions.k8s.io "alertmanagers.monitoring.coreos.com" already exists

*  完成后，执行kubectl get pod提示：

prometheus-operator-grafana-544db856cf-krl4p                 2/2     Running   0          29h
    prometheus-operator-kube-state-metrics-68f698d5b6-2k5kr      1/1     Running   0          29h
    prometheus-operator-operator-68d894d485-n4kd6                1/1     Running   0          29h
    prometheus-operator-prometheus-node-exporter-2ntpd           1/1     Running   0          29h
    prometheus-operator-prometheus-node-exporter-5zw2f           1/1     Running   0          29h
    prometheus-operator-prometheus-node-exporter-6h69p           1/1     Running   0          29h
    prometheus-operator-prometheus-node-exporter-kck5d           1/1     Running   0          29h
    prometheus-operator-prometheus-node-exporter-m8hnz           1/1     Running   0          29h
    prometheus-operator-prometheus-node-exporter-pgbvc           1/1     Running   0          29h
    prometheus-prometheus-operator-prometheus-0                  3/3     Running   1          29h

说明安装完成。

*  创建grafana-web.yml，输入如下内容：

apiVersion: v1
    kind: Service
    metadata:
     name: grafana-web
     namespace: default
     labels:
      app: grafana
    spec:
     type: NodePort
     ports:
     - port: 3000
       nodePort: 30081
     selector:
      app: grafana

执行kubectl create -f grafana-web.yml，之后，可以通过浏览器访问k8s集群中任意一台服务器的30081端口，登录grafana的界面。grafana已经自动配置了prometheus的数据源，仅需要添加dashboard即可，打开[https://grafana.com/dashboards?dataSource=prometheus&search=kubernetes][https_grafana.com_dashboards_dataSource_prometheus_search_kubernetes] ，选择一个dashboard的id号。进入grafana界面，选择左边栏的+号，再选择import按钮，输入id号，点击load，即可将dashboard加载。

参考 [https://blog.csdn.net/ygqygq2/article/details/82682040\#3-%E9%85%8D%E7%BD%AEgrafana][https_blog.csdn.net_ygqygq2_article_details_82682040_3-_E9_85_8D_E7_BD_AEgrafana]

[https_blog.csdn.net_liukuan73_article_details_79319900]: https://blog.csdn.net/liukuan73/article/details/79319900
[prometheus-operator_https_kubernetes-charts.storage.googleapis.com_prometheus-operator-5.0.3.tgz]: https://kubernetes-charts.storage.googleapis.com/prometheus-operator-5.0.3.tgz
[https_github.com_helm_charts_tree_master_stable_prometheus-operator]: https://github.com/helm/charts/tree/master/stable/prometheus-operator
[https_grafana.com_dashboards_dataSource_prometheus_search_kubernetes]: https://grafana.com/dashboards?dataSource=prometheus&search=kubernetes
[https_blog.csdn.net_ygqygq2_article_details_82682040_3-_E9_85_8D_E7_BD_AEgrafana]: https://blog.csdn.net/ygqygq2/article/details/82682040#3-%E9%85%8D%E7%BD%AEgrafana