k8s搭建prometheus+grafana监控系统

标签：kubectl kube grafana yaml prometheus dashboardDefinitions k8s

• 安装git
• yum install git -y
• 登录github，查看kube-prometheus项目
• https://github.com/prometheus-operator/kube-prometheus
• 查看kubenetes对应版本

• 拉取对应的kube-prometheus版本
• git clone https://github.com/prometheus-operator/kube-prometheus.git -b release-0.7
• cd /kube-prometheus/manifests/setup
• kubectl create -f .
• cd /kube-prometheus/manifests
• kubectl create -f .
修改三个文件，改为NodPort
• prometheus-service.yaml
• alertmanager-service.yaml
• grafana-service.yaml

• 
• 查看是否生效：kubectl get svc -n monitoring
• 有pod节点拉取镜像失败时：ImagePullBackOff
• kubectl describe pod -n monitoring
• 查看拉取镜像

• 修改 vim kubeStateMetrics-deployment.yaml
• image修改为：从https://hub.docker.com/搜索此镜像，并替换文件中地址承担

• kubectl apply -f kubeStateMetrics-deployment.yaml
• 所有失败的节点同上操作
• 修改grafana时区，否则grafana时间显示不准确

grep -i timezone grafana-dashboardDefinitions.yaml sed -i 's/UTC/UTC+8/g' grafana-dashboardDefinitions.yaml sed -i 's/utc/utc+8/g' grafana-dashboardDefinitions.yaml grep -i timezone grafana-dashboardDefinitions.yaml kubectl apply -f grafana-dashboardDefinitions.yaml  

标签：kubectl,kube,grafana,yaml,prometheus,dashboardDefinitions,k8s
From： https://www.cnblogs.com/hd92/p/17267566.html