使用kube-prometheus部署k8s監控

1、概述

1.1 在k8s中部署Prometheus監控的方法

通常在k8s中部署prometheus監控可以采取的方法有以下三種

• 通過yaml手動部署
• operator部署
• 通過helm chart部署

1.2 什么是Prometheus Operator

Prometheus Operator的本職就是一組用戶自定義的CRD資源以及Controller的實現，Prometheus Operator負責監聽這些自定義資源的變化，並且根據這些資源的定義自動化的完成如Prometheus Server自身以及配置的自動化管理工作。以下是Prometheus Operator的架構圖

圖片來源：

https://raw.githubusercontent.com/prometheus-operator/prometheus-operator/master/Documentation/user-guides/images/architecture.png

1.3 為什么用Prometheus Operator

由於Prometheus本身沒有提供管理配置的AP接口（尤其是管理監控目標和管理警報規則），也沒有提供好用的多實例管理手段，因此這一塊往往要自己寫一些代碼或腳本。為了簡化這類應用程序的管理復雜度，CoreOS率先引入了Operator的概念，並且首先推出了針對在Kubernetes下運行和管理Etcd的Etcd Operator。並隨后推出了Prometheus Operator

1.4 kube-prometheus項目介紹

prometheus-operator官方地址：https://github.com/prometheus-operator/prometheus-operator kube-prometheus官方地址：https://github.com/prometheus-operator/kube-prometheus

兩個項目的關系：前者只包含了Prometheus Operator，后者既包含了Operator，又包含了Prometheus相關組件的部署及常用的Prometheus自定義監控，具體包含下面的組件

• The Prometheus Operator：創建CRD自定義的資源對象
• Highly available Prometheus：創建高可用的Prometheus
• Highly available Alertmanager：創建高可用的告警組件
• Prometheus node-exporter：創建主機的監控組件
• Prometheus Adapter for Kubernetes Metrics APIs：創建自定義監控的指標工具（例如可以通過nginx的request來進行應用的自動伸縮）
• kube-state-metrics：監控k8s相關資源對象的狀態指標
• Grafana：進行圖像展示

2、環境介紹

本文的k8s環境是通過kubeadm搭建的v 1.20.2版本，由1master+2node組合

持久化存儲為storageclass動態存儲，底層由ceph-rbd提供

➜  kubectl version -o yaml
clientVersion: buildDate: "2020-12-08T17:59:43Z" compiler: gc gitCommit: af46c47ce925f4c4ad5cc8d1fca46c7b77d13b38 gitTreeState: clean gitVersion: v1.20.0 goVersion: go1.15.5 major: "1" minor: "20" platform: darwin/amd64 serverVersion: buildDate: "2021-01-13T13:20:00Z" compiler: gc gitCommit: faecb196815e248d3ecfb03c680a4507229c2a56 gitTreeState: clean gitVersion: v1.20.2 goVersion: go1.15.5 major: "1" minor: "20" platform: linux/amd64 ➜ kubectl get nodes NAME STATUS ROLES AGE VERSION k8s-m-01 Ready control-plane,master 11d v1.20.2 k8s-n-01 Ready <none> 11d v1.20.2 k8s-n-02 Ready <none> 11d v1.20.2 ➜ manifests kubectl get sc NAME PROVISIONER RECLAIMPOLICY VOLUMEBINDINGMODE ALLOWVOLUMEEXPANSION AGE dynamic-ceph-rbd (default) ceph.com/rbd Delete Immediate false 7d23h

kube-prometheus的兼容性說明（https://github.com/prometheus-operator/kube-prometheus#kubernetes-compatibility-matrix），按照兼容性說明，部署的是最新的release-0.7版本

kube-prometheus stack	Kubernetes 1.16	Kubernetes 1.17	Kubernetes 1.18	Kubernetes 1.19	Kubernetes 1.20
release-0.4	✔ (v1.16.5+)	✔	✗	✗	✗
release-0.5	✗	✗	✔	✗	✗
release-0.6	✗	✗	✔	✔	✗
release-0.7	✗	✗	✗	✔	✔
HEAD	✗	✗	✗	✔	✔

3、清單准備

從官方的地址獲取最新的release-0.7分支，或者直接打包下載release-0.7

➜  git clone https://github.com/prometheus-operator/kube-prometheus.git ➜ git checkout release-0.7 或者 ➜ wget -c https://github.com/prometheus-operator/kube-prometheus/archive/v0.7.0.zip

默認下載下來的文件較多，建議把文件進行歸類處理，將相關yaml文件移動到對應目錄下

➜  cd kube-prometheus/manifests ➜ mkdir -p serviceMonitor prometheus adapter node-exporter kube-state-metrics grafana alertmanager operator other

最終結構如下

➜  manifests tree .
. ├── adapter │ ├── prometheus-adapter-apiService.yaml │ ├── prometheus-adapter-clusterRole.yaml │ ├── prometheus-adapter-clusterRoleAggregatedMetricsReader.yaml │ ├── prometheus-adapter-clusterRoleBinding.yaml │ ├── prometheus-adapter-clusterRoleBindingDelegator.yaml │ ├── prometheus-adapter-clusterRoleServerResources.yaml │ ├── prometheus-adapter-configMap.yaml │ ├── prometheus-adapter-deployment.yaml │ ├── prometheus-adapter-roleBindingAuthReader.yaml │ ├── prometheus-adapter-service.yaml │ └── prometheus-adapter-serviceAccount.yaml ├── alertmanager │ ├── alertmanager-alertmanager.yaml │ ├── alertmanager-secret.yaml │ ├── alertmanager-service.yaml │ └── alertmanager-serviceAccount.yaml ├── grafana │ ├── grafana-dashboardDatasources.yaml │ ├── grafana-dashboardDefinitions.yaml │ ├── grafana-dashboardSources.yaml │ ├── grafana-deployment.yaml │ ├── grafana-service.yaml │ └── grafana-serviceAccount.yaml ├── kube-state-metrics │ ├── kube-state-metrics-clusterRole.yaml │ ├── kube-state-metrics-clusterRoleBinding.yaml │ ├── kube-state-metrics-deployment.yaml │ ├── kube-state-metrics-service.yaml │ └── kube-state-metrics-serviceAccount.yaml ├── node-exporter │ ├── node-exporter-clusterRole.yaml │ ├── node-exporter-clusterRoleBinding.yaml │ ├── node-exporter-daemonset.yaml │ ├── node-exporter-service.yaml │ └── node-exporter-serviceAccount.yaml ├── operator │ ├── 0namespace-namespace.yaml │ ├── prometheus-operator-0alertmanagerConfigCustomResourceDefinition.yaml │ ├── prometheus-operator-0alertmanagerCustomResourceDefinition.yaml │ ├── prometheus-operator-0podmonitorCustomResourceDefinition.yaml │ ├── prometheus-operator-0probeCustomResourceDefinition.yaml │ ├── prometheus-operator-0prometheusCustomResourceDefinition.yaml │ ├── prometheus-operator-0prometheusruleCustomResourceDefinition.yaml │ ├── prometheus-operator-0servicemonitorCustomResourceDefinition.yaml │ ├── prometheus-operator-0thanosrulerCustomResourceDefinition.yaml │ ├── prometheus-operator-clusterRole.yaml │ ├── prometheus-operator-clusterRoleBinding.yaml │ ├── prometheus-operator-deployment.yaml │ ├── prometheus-operator-service.yaml │ └── prometheus-operator-serviceAccount.yaml ├── other ├── prometheus │ ├── prometheus-clusterRole.yaml │ ├── prometheus-clusterRoleBinding.yaml │ ├── prometheus-prometheus.yaml │ ├── prometheus-roleBindingConfig.yaml │ ├── prometheus-roleBindingSpecificNamespaces.yaml │ ├── prometheus-roleConfig.yaml │ ├── prometheus-roleSpecificNamespaces.yaml │ ├── prometheus-rules.yaml │ ├── prometheus-service.yaml │ └── prometheus-serviceAccount.yaml └── serviceMonitor ├── alertmanager-serviceMonitor.yaml ├── grafana-serviceMonitor.yaml ├── kube-state-metrics-serviceMonitor.yaml ├── node-exporter-serviceMonitor.yaml ├── prometheus-adapter-serviceMonitor.yaml ├── prometheus-operator-serviceMonitor.yaml ├── prometheus-serviceMonitor.yaml ├── prometheus-serviceMonitorApiserver.yaml ├── prometheus-serviceMonitorCoreDNS.yaml ├── prometheus-serviceMonitorKubeControllerManager.yaml ├── prometheus-serviceMonitorKubeScheduler.yaml └── prometheus-serviceMonitorKubelet.yaml 9 directories, 67 files

修改yaml，增加prometheus和grafana的持久化存儲

manifests/prometheus/prometheus-prometheus.yaml

...
  serviceMonitorSelector: {} version: v2.22.1 retention: 3d storage: volumeClaimTemplate: spec: storageClassName: dynamic-ceph-rbd resources: requests: storage: 5Gi

manifests/grafana/grafana-deployment.yaml

...
      serviceAccountName: grafana volumes: # - emptyDir: {} # name: grafana-storage - name: grafana-storage persistentVolumeClaim: claimName: grafana-data

新增grafana的pvc，manifests/other/grafana-pvc.yaml

kind: PersistentVolumeClaim
apiVersion: v1 metadata: name: grafana-data namespace: monitoring annotations: volume.beta.kubernetes.io/storage-class: "dynamic-ceph-rbd" spec: accessModes: - ReadWriteMany resources: requests: storage: 5Gi

4、開始部署

部署清單

➜  kubectl create -f other/grafana-pvc.yaml ➜ kubectl create -f operator/ ➜ kubectl create -f adapter/ -f alertmanager/ -f grafana/ -f kube-state-metrics/ -f node-exporter/ -f prometheus/ -f serviceMonitor/ 

查看狀態

➜  kubectl get po,svc -n monitoring NAME READY STATUS RESTARTS AGE pod/alertmanager-main-0 2/2 Running 0 15m pod/alertmanager-main-1 2/2 Running 0 10m pod/alertmanager-main-2 2/2 Running 0 15m pod/grafana-d69dcf947-wnspk 1/1 Running 0 22m pod/kube-state-metrics-587bfd4f97-bffqv 3/3 Running 0 22m pod/node-exporter-2vvhv 2/2 Running 0 22m pod/node-exporter-7nsz5 2/2 Running 0 22m pod/node-exporter-wggpp 2/2 Running 0 22m pod/prometheus-adapter-69b8496df6-cjw6w 1/1 Running 0 23m pod/prometheus-k8s-0 2/2 Running 1 75s pod/prometheus-k8s-1 2/2 Running 0 9m33s pod/prometheus-operator-7649c7454f-nhl72 2/2 Running 0 28m NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE service/alertmanager-main ClusterIP 10.1.189.238 <none> 9093/TCP 23m service/alertmanager-operated ClusterIP None <none> 9093/TCP,9094/TCP,9094/UDP 23m service/grafana ClusterIP 10.1.29.30 <none> 3000/TCP 23m service/kube-state-metrics ClusterIP None <none> 8443/TCP,9443/TCP 23m service/node-exporter ClusterIP None <none> 9100/TCP 23m service/prometheus-adapter ClusterIP 10.1.75.64 <none> 443/TCP 23m service/prometheus-k8s ClusterIP 10.1.111.121 <none> 9090/TCP 23m service/prometheus-operated ClusterIP None <none> 9090/TCP 14m service/prometheus-operator ClusterIP None <none> 8443/TCP 28m

為prometheus、grafana、alertmanager創建ingress

manifests/other/ingress.yaml

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: prom-ingresses
  namespace: monitoring
spec:
  rules:
  - host: alert.test.com
    http:
      paths:
      - backend:
          serviceName: alertmanager-main
          servicePort: 9093
        path: /
  - host: grafana.test.com
    http:
      paths:
      - backend:
          serviceName: grafana
          servicePort: 3000
        path: /
  - host: prom.test.com
    http:
      paths:
      - backend:
          serviceName: prometheus-k8s
          servicePort: 9090
        path: /

5、解決ControllerManager、Scheduler監控問題

默認安裝后訪問prometheus，會發現有以下有三個報警：

Watchdog、KubeControllerManagerDown、KubeSchedulerDown

Watchdog是一個正常的報警，這個告警的作用是：如果alermanger或者prometheus本身掛掉了就發不出告警了，因此一般會采用另一個監控來監控prometheus，或者自定義一個持續不斷的告警通知，哪一天這個告警通知不發了，說明監控出現問題了。prometheus operator已經考慮了這一點，本身攜帶一個watchdog，作為對自身的監控。

如果需要關閉，刪除或注釋掉Watchdog部分

prometheus-rules.yaml

...
  - name: general.rules rules: - alert: TargetDown annotations: message: 'xxx' expr: 100 * (count(up == 0) BY (job, namespace, service) / count(up) BY (job, namespace, service)) > 10 for: 10m labels: severity: warning # - alert: Watchdog # annotations: # message: | # This is an alert meant to ensure that the entire alerting pipeline is functional. # This alert is always firing, therefore it should always be firing in Alertmanager # and always fire against a receiver. There are integrations with various notification # mechanisms that send a notification when this alert is not firing. For example the # "DeadMansSnitch" integration in PagerDuty. # expr: vector(1) # labels: # severity: none

KubeControllerManagerDown、KubeSchedulerDown的解決

原因是因為在prometheus-serviceMonitorKubeControllerManager.yaml中有如下內容，但默認安裝的集群並沒有給系統kube-controller-manager組件創建svc

  selector:
    matchLabels: k8s-app: kube-controller-manager

修改kube-controller-manager的監聽地址

# vim /etc/kubernetes/manifests/kube-controller-manager.yaml ... spec: containers: - command: - kube-controller-manager - --allocate-node-cidrs=true - --authentication-kubeconfig=/etc/kubernetes/controller-manager.conf - --authorization-kubeconfig=/etc/kubernetes/controller-manager.conf - --bind-address=0.0.0.0 # netstat -lntup|grep kube-contro tcp6 0 0 :::10257 :::* LISTEN 38818/kube-controll

創建一個service和endpoint，以便serviceMonitor監聽

other/kube-controller-namager-svc-ep.yaml

apiVersion: v1
kind: Service metadata: name: kube-controller-manager namespace: kube-system labels: k8s-app: kube-controller-manager spec: type: ClusterIP clusterIP: None ports: - name: https-metrics port: 10257 targetPort: 10257 protocol: TCP --- apiVersion: v1 kind: Endpoints metadata: name: kube-controller-manager namespace: kube-system labels: k8s-app: kube-controller-manager subsets: - addresses: - ip: 172.16.1.71 ports: - name: https-metrics port: 10257 protocol: TCP

kube-scheduler同理，修改kube-scheduler的監聽地址

# vim /etc/kubernetes/manifests/kube-scheduler.yaml ... spec: containers: - command: - kube-scheduler - --authentication-kubeconfig=/etc/kubernetes/scheduler.conf - --authorization-kubeconfig=/etc/kubernetes/scheduler.conf - --bind-address=0.0.0.0 # netstat -lntup|grep kube-sched tcp6 0 0 :::10259 :::* LISTEN 100095/kube-schedul

創建一個service和endpoint，以便serviceMonitor監聽

kube-scheduler-svc-ep.yaml

apiVersion: v1
kind: Service metadata: name: kube-scheduler namespace: kube-system labels: k8s-app: kube-scheduler spec: type: ClusterIP clusterIP: None ports: - name: https-metrics port: 10259 targetPort: 10259 protocol: TCP --- apiVersion: v1 kind: Endpoints metadata: name: kube-scheduler namespace: kube-system labels: k8s-app: kube-scheduler subsets: - addresses: - ip: 172.16.1.71 ports: - name: https-metrics port: 10259 protocol: TCP

再次查看prometheus的alert界面，全部恢復正常

登錄到grafana，查看相關圖像展示

至此，通過kube-prometheus部署k8s監控已經基本完成了，后面再分享自定義監控和告警、告警通知、高可用、規模化部署等相關內容

參考：https://github.com/prometheus-operator/kube-prometheus