Kubernetes从0到1-第五篇 部署Prometheus监控

环境要求

• helm 版本 v3.3+
• kubernetes 集群版本 1.14+

1. 添加国内chart仓库

首先，添加国内的镜像仓库源

helm repo add apphub https://apphub.aliyuncs.com
helm repo update

查看已经添加的 chart 库

helm repo list

2. 创建命名空间 namespace
   将 Prometheus 安装到独立的命名空间。

创建 namespace 的配置文件 monitoring.yaml

apiVersion: v1
kind: Namespace
metadata:
   name: monitoring
   labels:
     name: monitoring

应用刚创建的 yaml 文件

kubectl apply -f monitoring.yaml

3. 选择 prometheus-operator 版本

查询所需要的版本:

# helm search prometheus-operator -l

apphub/prometheus-operator      8.7.0           0.35.0          Provides easy monitoring definitions for Kubernetes servi...
apphub/prometheus-operator      8.2.2           0.34.0          Provides easy monitoring definitions for Kubernetes servi...
apphub/prometheus-operator      8.2.0           0.34.0          Provides easy monitoring definitions for Kubernetes servi...
apphub/prometheus-operator      8.1.2           0.34.0          Provides easy monitoring definitions for Kubernetes servi...
apphub/prometheus-operator      7.1.1           0.32.0          Provides easy monitoring definitions for Kubernetes servi...

选择8.2.2版本安装。

安装

下载chart

helm pull apphub/prometheus-operator --version 8.2.2
tar -xvf prometheus-operator-8.2.2.tgz && cd prometheus-operator/

修改父values.yaml

alertmanager:
  ingress:
    enabled: true
    hosts: 
      - alertmanager.demo.com
  service:
    loadBalancerSourceRanges: ["192.168.238.0/24"]
    type: LoadBalancer
grafana:
  # 覆盖子chart（grafana）镜像版本
  image:
    repository: grafana/grafana
    tag: 8.1.0
  enabled: true
  # 启用密码
  adminPassword: admin@123456
  ingress:
    enabled: true
    hosts:
      - grafana.demo.com
  grafana.ini:
    ## 添加数据库信息，数据库配置信息将写入grafana配置文件 grafana.ini 中。
    database:
      type: mysql
      host: 192.168.238.150
      name: grafana
      user: grafana
      password: ZdZ5GIBh
      ssl_mode: disable
prometheus:
  enabled: true
  service:
    loadBalancerSourceRanges: ["192.168.238.0/24"]
    type: LoadBalancer
  ingress:
    enabled: true
    hosts:
      - prometheus.demo.com
  prometheusSpec:
    # 持久化存储，实现TSDB数据持久化
    storageSpec: 
      volumeClaimTemplate:
        spec:
          storageClassName: managed-nfs-storage
          accessModes: ["ReadWriteOnce"]
          resources:
            requests:
              storage: 20Gi

安装prometheus-operator

helm install prometheus -n monitoring . -f ./values.yaml

查看运行的pod

kubectl get pods -n monitoring

NAME                                                     READY   STATUS    RESTARTS   AGE
alertmanager-prometheus-prometheus-oper-alertmanager-0   2/2     Running   0          4h2m
prometheus-grafana-b876dcf9c-9d92n                       2/2     Running   0          4h2m
prometheus-kube-state-metrics-6f569f64f6-x2fxl           1/1     Running   0          4h2m
prometheus-prometheus-node-exporter-4bnd2                1/1     Running   0          4h2m
prometheus-prometheus-node-exporter-4j8m9                1/1     Running   0          4h2m
prometheus-prometheus-node-exporter-5md8v                1/1     Running   0          4h2m
prometheus-prometheus-oper-operator-7687488465-cmg9l     2/2     Running   0          4h2m
prometheus-prometheus-prometheus-oper-prometheus-0       3/3     Running   1          4h1m

查看ingress

kubectl get ingress -n monitoring

NAME                                      CLASS    HOSTS                   ADDRESS           PORTS   AGE
prometheus-grafana                        <none>   grafana.demo.com        192.168.238.156   80      20h
prometheus-prometheus-oper-alertmanager   <none>   alertmanager.demo.com   192.168.238.156   80      20h
prometheus-prometheus-oper-prometheus     <none>   prometheus.demo.com     192.168.238.156   80      20h

访问grafana

访问prometheus

访问alertmanager

参考文档
[1]: https://github.com/helm/charts/blob/master/stable/prometheus-operator/README.md