kubernetesv1.16監控系列——kube-prometheus安裝

下載源碼

我這直接通過kube-prometheus的源碼進行安裝，當讓也可以通helm 安裝。

採用代碼如下：

<code>git clone https://github.com/coreos/kube-prometheus.git
cd manifests
alertmanager-alertmanager.yaml                              prometheus-adapter-clusterRole.yaml
alertmanager-secret.yaml                                    prometheus-adapter-configMap.yaml
alertmanager-serviceAccount.yaml                            prometheus-adapter-deployment.yaml
alertmanager-serviceMonitor.yaml                            prometheus-adapter-roleBindingAuthReader.yaml
alertmanager-service.yaml                                   prometheus-adapter-serviceAccount.yaml
grafana-dashboardDatasources.yaml                           prometheus-adapter-service.yaml
grafana-dashboardDefinitions.yaml                           prometheus-clusterRoleBinding.yaml
grafana-dashboardSources.yaml                               prometheus-clusterRole.yaml
grafana-deployment.yaml                                     prometheus-kubeControllerManagerService.yaml
grafana-serviceAccount.yaml                                 
grafana-serviceMonitor.yaml                                 prometheus-kubeSchedulerService.yaml
grafana-service.yaml                                        prometheus-operator-serviceMonitor.yaml
kube-state-metrics-clusterRoleBinding.yaml                  prometheus-prometheus.yaml
kube-state-metrics-clusterRole.yaml                         prometheus-roleBindingConfig.yaml
kube-state-metrics-deployment.yaml                          prometheus-roleBindingSpecificNamespaces.yaml
kube-state-metrics-serviceAccount.yaml                      prometheus-roleConfig.yaml
kube-state-metrics-serviceMonitor.yaml                      prometheus-roleSpecificNamespaces.yaml
kube-state-metrics-service.yaml                             prometheus-rules.yaml
node-exporter-clusterRoleBinding.yaml                       prometheus-serviceAccount.yaml
node-exporter-clusterRole.yaml                              prometheus-serviceMonitorApiserver.yaml
node-exporter-daemonset.yaml                                prometheus-serviceMonitorCoreDNS.yaml
node-exporter-serviceAccount.yaml                          
node-exporter-serviceMonitor.yaml                           prometheus-serviceMonitorKubeControllerManager.yaml
node-exporter-service.yaml                                  prometheus-serviceMonitorKubelet.yaml
prometheus-adapter-apiService.yaml                          prometheus-serviceMonitorKubeScheduler.yaml
prometheus-adapter-clusterRoleAggregatedMetricsReader.yaml  prometheus-serviceMonitor.yaml
prometheus-adapter-clusterRoleBindingDelegator.yaml         prometheus-service.yaml
prometheus-adapter-clusterRoleBinding.yaml                  setup
prometheus-adapter-clusterRoleServerResources.yaml          /<code>

<code>需要修改以下文件：
grafana-service.yaml 
/<code>

為了能直接訪問grafana、Prometheus、alertmanager修改以下文件

修改grafana-service.yaml文件，修改如下：

<code>修改grafana-service.yaml文件，修改如下：
grafana-service.yaml /<code> 

修改alertmanager-secret.yaml文件，修改如下：

<code>apiVersion: v1
kind: Service
metadata:
  labels:
    alertmanager: main
  name: alertmanager-main
  namespace: monitoring
spec:
  ports:
  - name: web
    port: 9093
    targetPort: web
    nodePort: 30093
  type: NodePort
  selector:
    alertmanager: main
    app: alertmanager
  sessionAffinity: ClientIP/<code>

修改prometheus-service.yaml文件，修改如下：

<code>apiVersion: v1
kind: Service
metadata:
  labels:
    prometheus: k8s
  name: prometheus-k8s
  namespace: monitoring
spec:
  ports:
  - name: web
    port: 9090
    targetPort: web
    nodePort: 30090
  type: NodePort
  selector:
    app: prometheus
    prometheus: k8s
  sessionAffinity: ClientIP/<code>

如果是二進制安裝kubernetes的集群，ControllerManager和Scheduler就是監控不到的，所以需要新建這個service

prometheus-kubeControllerManagerService.yaml

<code>apiVersion: v1
kind: Service
metadata:
  namespace: kube-system
  name: kube-controller-manager
  labels:
    k8s-app: kube-controller-manager
spec:
  type: ClusterIP
  clusterIP: None
  ports:
  - name: http-metrics
    port: 10252
    targetPort: 10252
    protocol: TCP

---
apiVersion: v1
kind: Endpoints
metadata:
  labels:
    k8s-app: kube-controller-manager
  name: kube-controller-manager
  namespace: kube-system
subsets:
- addresses:
  - ip: 10.6.2.121
  ports:
  - name: http-metrics
    port: 10252
    protocol: TCP/<code>

prometheus-kubeSchedulerService.yaml

<code>apiVersion: v1
kind: Service
metadata:
  namespace: kube-system
  name: kube-scheduler
  labels:
    k8s-app: kube-scheduler
spec:
  ports:
  - name: http-metrics
    port: 10251
    targetPort: 10251
    protocol: TCP
  clusterIP: None
  type: ClusterIP 

---
apiVersion: v1
kind: Endpoints
metadata:
  name: kube-scheduler
  namespace: kube-system
  labels:
    k8s-app: kube-scheduler
subsets:
- addresses:
  - ip: 10.6.2.121
    nodeName: kube-scheduler
  ports:
  - name: http-metrics
    port: 10251
    protocol: TCP/<code>

執行命令：

<code>kubectl apply -f setup/
kubectl apply -f ./<code>

驗證

部署完成後，會創建一個名為monitoring的 namespace，所以資源對象對將部署在改命名空間下面，此外 kube-prometheus會自動創建6個 CRD 資源對象：

<code>[root@master-121 manifests]# kubectl get crd |grep coreos
alertmanagers.monitoring.coreos.com     2020-02-22T15:19:18Z
podmonitors.monitoring.coreos.com       2020-02-22T15:19:18Z
prometheuses.monitoring.coreos.com      2020-02-22T15:19:18Z
prometheusrules.monitoring.coreos.com   2020-02-22T15:19:18Z
servicemonitors.monitoring.coreos.com   2020-02-22T15:19:18Z
thanosrulers.monitoring.coreos.com      2020-02-22T15:19:19Z

[root@master-121 manifests]# kubectl get pods -n monitoring
NAME                                   READY   STATUS    RESTARTS   AGE
alertmanager-main-0                    2/2     Running   0          3d21h
alertmanager-main-1                    2/2     Running   0          3d21h
alertmanager-main-2                    2/2     Running   0          3d21h
grafana-849658db4d-cp5h5               1/1     Running   0          3d20h
kube-state-metrics-5859ffdc64-hknmj    1/1     Running   0          3d21h
node-exporter-gb2x4                    2/2     Running   0          3d21h
node-exporter-mjllw                    2/2     Running   0          3d21h
node-exporter-v8vwd                    2/2     Running   0          3d21h
node-exporter-xtwzj                    2/2     Running   0          3d21h 

prometheus-adapter-5cd5798d96-k9hqm    1/1     Running   0          3d21h
prometheus-k8s-0                       3/3     Running   1          88m
prometheus-k8s-1                       3/3     Running   1          88m
prometheus-operator-5d94cdc9bf-6flp9   1/1     Running   0          3d13h

[root@master-121 manifests]# kubectl get svc -n monitoring
NAME                    TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)                      AGE
alertmanager-main       NodePort    10.16.101.12   <none>        9093:30093/TCP               3d21h
alertmanager-operated   ClusterIP   None           <none>        9093/TCP,9094/TCP,9094/UDP   3d21h
blackbox-exporter       ClusterIP   None           <none>        9115/TCP                     18h
grafana                 NodePort    10.16.77.172   <none>        3000:32000/TCP               3d21h
kube-state-metrics      ClusterIP   None           <none>        8080/TCP,8081/TCP            3d21h
node-exporter           ClusterIP   None           <none>        9100/TCP                     3d21h
prometheus-adapter      ClusterIP   10.16.39.20    <none>        443/TCP                      3d21h
prometheus-k8s          NodePort    10.16.230.93   <none>        9090:30090/TCP               3d21h
prometheus-operated     ClusterIP   None           <none>        9090/TCP                     3d21h
prometheus-operator     ClusterIP   None           <none>        8080/TCP                     3d13h/<none>/<none>/<none>/<none>/<none>/<none>/<none>/<none>/<none>/<none>/<code>

檢查無失敗，訪問Prometheus和grafana查查看

http://10.6.2.121:30090/targets

http://10.6.2.121:32000/

<code>注：如果還是沒有ControllerManager和Scheduler的監控，查看集群安裝是否把監聽地址改為--address=0.0.0.0/<code>

閱讀更多 上海IT故事 的文章

關鍵字: 系列 安裝 CoreOS