kubectl管理Pod-2: 命名空间，镜像拉取策略，健康检查-重启策略，资源限制，调度策略

阳光穿透心脏的1/2处 2024-04-18 23:00 2阅读 0赞

# 指定程序运行的命名空间 #

[root@master aa]# cat ns-deploy.yaml
    kind: Namespace
    apiVersion: v1
    metadata:
      name: test
      labels:
        name: test
    ---
    apiVersion: apps/v1
    kind: Deployment
    ....

# 1, 镜像拉取策略 #

imagePullPolicy：默认Always, ( Never, IfNotPresent ）

[root@master aa]# cat pullpolicy-deploy.yaml
    apiVersion: apps/v1
    kind: Deployment
    metadata:
      labels:
        app: nginx-test
      name: nginx-test
    spec:
      replicas: 1
      selector:
        matchLabels:
          app: nginx-test
      template:
        metadata:
          labels:
            app: nginx-test
        spec:
          containers:
          - name: nginx1
            image: nginx
            imagePullPolicy: Always
          - name: nginx2
            image: nginx
            imagePullPolicy: Always
          - name: nginx3
            image: nginx
            imagePullPolicy: IfNotPresent
            
    ###查看创建的pod日志
    Events:
      Type     Reason     Age                From               Message
      ----     ------     ----               ----               -------
      Normal   Scheduled  25s                default-scheduler  Successfully assigned default/nginx-test-c67597c44-chdw8 to node1
      Normal   Pulling    24s                kubelet, node1     Pulling image "nginx"
      Normal   Pulled     22s                kubelet, node1     Successfully pulled image "nginx" ##拉取
      Normal   Created    22s                kubelet, node1     Created container nginx1
      Normal   Started    21s                kubelet, node1     Started container nginx1
      Normal   Pulling    15s (x2 over 21s)  kubelet, node1     Pulling image "nginx"
      Normal   Pulled     14s (x2 over 19s)  kubelet, node1     Successfully pulled image "nginx" ##拉取
      Normal   Created    14s (x2 over 19s)  kubelet, node1     Created container nginx2
      Normal   Started    13s (x2 over 19s)  kubelet, node1     Started container nginx2
      Normal   Pulled     12s (x2 over 19s)  kubelet, node1     Container image "nginx" already present on machine ##存在就不拉取
      Normal   Created    12s (x2 over 19s)  kubelet, node1     Created container nginx3
      Normal   Started    11s (x2 over 18s)  kubelet, node1     Started container nginx3
      Warning  BackOff    8s (x3 over 10s)   kubelet, node1     Back-off restarting failed container
      Warning  BackOff    8s                 kubelet, node1     Back-off restarting failed container

# 2, 健康检查，重启策略 #

*  直接创建的pod对象，可以自定义重启策略:Always,OnFailure,Never
 *  通过deploy创建的pod对象，若有健康检查，默认重启策略为Always（且只支持Always）

[root@master live]# ls
    dep.yaml  pod.yaml
    [root@master live]# kubectl apply -f .
    
    ##查看pod状态： restart值递增中...
    [root@master live]# while true; do kubectl get deploy,pod; sleep 2 ; done
    NAME                                READY   UP-TO-DATE   AVAILABLE   AGE
    deployment.extensions/deploy-live   1/1     1            1           69s
    NAME                               READY   STATUS    RESTARTS   AGE
    pod/deploy-live-55595cf5c4-mxcx9   2/2     Running   5          69s
    pod/pod-live                       2/2     Running   3          69s
    
    ## 查看启动日志
    [root@master live]# kubectl describe default/deploy-live-55595cf5c4-f46xb 
    Events:
      Type     Reason     Age               From               Message
      ----     ------     ----              ----               -------
      Normal   Scheduled  28s               default-scheduler  Successfully assigned default/deploy-live-55595cf5c4-f46xb to node1
      Normal   Pulled     5s (x2 over 26s)  kubelet, node1     Container image "busybox" already present on machine
      Normal   Created    5s (x2 over 25s)  kubelet, node1     Created container live2
      Normal   Pulled     4s (x2 over 26s)  kubelet, node1     Container image "busybox" already present on machine
      Normal   Started    4s (x2 over 25s)  kubelet, node1     Started container live2
      Normal   Created    3s (x2 over 26s)  kubelet, node1     Created container live1
      Normal   Started    3s (x2 over 26s)  kubelet, node1     Started container live1
      Warning  Unhealthy  1s (x4 over 23s)  kubelet, node1     Liveness probe failed: cat: can't open '/a.txt': No such file or directory
      Warning  Unhealthy  0s (x6 over 23s)  kubelet, node1     Liveness probe failed: cat: can't open '/a.txt': No such file or directory
      Normal   Killing    0s (x2 over 21s)  kubelet, node1     Container live2 failed liveness probe, will be restarted
    
    [root@master live]# kubectl describe pod pod-live
    Events:
      Type     Reason     Age              From               Message
      ----     ------     ----             ----               -------
      Normal   Scheduled  10s              default-scheduler  Successfully assigned default/pod-live to node1
      Normal   Pulled     8s               kubelet, node1     Container image "busybox" already present on machine
      Normal   Created    8s               kubelet, node1     Created container pode-live1
      Normal   Started    7s               kubelet, node1     Started container pode-live1
      Normal   Pulled     7s               kubelet, node1     Container image "busybox" already present on machine
      Normal   Created    7s               kubelet, node1     Created container live2
      Normal   Started    7s               kubelet, node1     Started container live2
      Warning  Unhealthy  4s (x3 over 6s)  kubelet, node1     Liveness probe failed: cat: can't open '/tmp/healthy': No such file or directory
      Normal   Killing    4s               kubelet, node1     Container pode-live1 failed liveness probe, will be restarted
      Warning  Unhealthy  3s (x3 over 5s)  kubelet, node1     Liveness probe failed: cat: can't open '/tmp/healthy': No such file or directory

restart-pod.yaml

apiVersion: v1
    kind: Pod
    metadata:
      labels:
        test: pod-live
      name: pod-live
    spec:
      containers:
      - name: pode-live1
        image: busybox
        imagePullPolicy: Never
        args: ["sleep","20"]
        livenessProbe:
          exec:
            command:
            - cat
            - /tmp/healthy
          initialDelaySeconds: 1
          periodSeconds: 1
      - name: live2
        image: busybox
        imagePullPolicy: Never
        args:
        - /bin/sh
        - -c
        - sleep 20
        livenessProbe:
          exec:
            command:
            - cat
            - /tmp/healthy
          initialDelaySeconds: 1
          periodSeconds: 1
      restartPolicy: OnFailure

restart-deploy.yaml

apiVersion: apps/v1
    kind: Deployment
    metadata:
      labels:
        app: busyxob-live
      name: deploy-live
    spec:
      replicas: 1
      selector:
        matchLabels:
          app: deploy-live
      template:
        metadata:
          labels:
            app: deploy-live
        spec:
          containers:
          - name: live1
            image: busybox
            imagePullPolicy: Never
            command: ["sleep","20"]
            livenessProbe:
              exec:
                command: ["cat","/a.txt"]
              initialDelaySeconds: 2
              periodSeconds: 1
    
          - name: live2
            image: busybox
            imagePullPolicy: IfNotPresent
            args:
              - /bin/sh
              - -c
              - sleep 20
            livenessProbe:
              exec:
                command: ["cat","/a.txt"]
              initialDelaySeconds: 2
              periodSeconds: 1
          restartPolicy: Always

# 3, 资源限制 #

*  1核2线程CPU= 2个逻辑CPU ，1个逻辑CPU= 1000个millicore ,即1000m
 *  1M=1000, 1Mi=1024

[root@master aa]# cat limit-deploy.yaml 
    apiVersion: apps/v1
    kind: Deployment
    metadata:
      labels:
        app: nginx-1
      name: nginx-1
    spec:
      replicas: 1
      selector:
        matchLabels:
          app: nginx-1
      template:
        metadata:
          labels:
            app: nginx-1
        spec:
          containers:
          - name: nginx-1
            image: nginx
            resources:
              requests:
                 memory: "64Mi"
                 cpu: "250m"
              limits:
                 memory: "128Mi"
                 cpu: "500m"

# 4, 调度策略 #

*  nodeName: 指定pod运行的节点-使用节点名匹配
 *  nodeSelector: 指定pod运行的节点-使用标签匹配

[root@master live]# cat dep.yaml 
    apiVersion: apps/v1
    kind: Deployment
    metadata:
      labels:
        app: busyxob-live
      name: busybox-test
    spec:
      replicas: 1
      selector:
        matchLabels:
          app: busybox-test
      template:
        metadata:
          labels:
            app: busybox-test
        spec:
          nodeName: node2
          #nodeSelector: 
            #group: dev2
            #location: shenzhen
          containers:
          - name: busybox-test
            image: busybox
            imagePullPolicy: Never
            command: ["sleep","20"]
    
    [root@master live]# kubectl apply -f dep.yaml 
    deployment.apps/busybox-test created
    [root@master live]# kubectl get pod -o wide
    NAME                            READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
    busybox-test-6bddc5b945-jt9cl   1/1     Running   0          10s   10.244.2.80   node1   <none>           <none>
    
    ##改变调度的节点名：
    [root@master live]# sed -i "s/nodeName: node1/nodeName: node2/"  dep.yaml 
    [root@master live]# kubectl apply -f dep.yaml 
    deployment.apps/busybox-test configured
    [root@master live]# kubectl get pod -o wide
    NAME                            READY   STATUS        RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
    busybox-test-57b8497f89-b442v   0/1     Terminating   1          48s   10.244.2.82   node1   <none>           <none>
    busybox-test-798ffc744-spzhn    1/1     Running       0          3s    10.244.1.75   node2   <none>           <none>
    
    ### 通过nodeSelector标签：为pod匹配合适的运行节点
    [root@master live]# kubectl label node node1 group=dev1 location=beijing
    node/node1 labeled
    [root@master live]# kubectl label node node2 group=dev2 location=shenzhen
    node/node2 labeled
    [root@master live]# kubectl get node --show-labels=true
    NAME     STATUS   ROLES    AGE   VERSION   LABELS
    master   Ready    master   19d   v1.15.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=master,kubernetes.io/os=linux,node-role.kubernetes.io/master=
    node1    Ready    <none>   19d   v1.15.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,group=dev1,kubernetes.io/arch=amd64,kubernetes.io/hostname=node1,kubernetes.io/os=linux,location=beijing
    node2    Ready    <none>   19d   v1.15.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,group=dev2,isok=yes,kubernetes.io/arch=amd64,kubernetes.io/hostname=node2,kubernetes.io/os=linux,location=shenzhen