k8s (3)k8s的Pod管理

标签

w1n0

584人浏览 · 2020-12-27 18:53:59

w1n0 · 2020-12-27 18:53:59 发布

文章目录

Pod简介

什么是Pod？

kubernetes中的一切都可以理解为是一种资源对象，pod,rs,svc,都可以理解是一种资源对象。
在这里插入图片描述

Pod分类

Pod分为两类：

自主式Pod：这种Pod本身是不能自我修复的。当Pod被创建后（不论是由你直接创建还是被其它controller创建），都会被k8s调度到集群的Node上。直到pod的进程终止、被删掉。因为缺少资源而被驱逐、或者Node故障之前这个Pod都会一直保持在个Node上。Pod不会自愈。如果Pod运行的Node故障，或者是调度器本身故障，这个Pod就会被删除。同样的，如果Pod所在Node缺少资源或者Pod处于维护状态，Pod也会被驱逐。
控制器管理的Pod：k8s使用更高级的称为Controller的抽象层，来管理Pod实例。Controller可以创建和管理多个Pod，提供副本管理、滚动升级和集群级别的自愈能力。例如。如果一个Node故障，Controller就能自动将该节点上的Pod调度到其他健康的Node上，虽然可以直接使用Pod，但是在K8s中通常是使用Controller来管理Pod的。

Pod特性

Pod标签

apiVersion: apps/v1
kind: Deployment
metadata: 
  name: deployment-example
spec:
  replicas: 4
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: IfNotPresent

运行：

[k8s@server1 ~]$ kubectl apply -f nginx.yml 
deployment.apps/deployment-example created
[k8s@server1 ~]$ kubectl get pod 
NAME                                  READY   STATUS    RESTARTS   AGE
deployment-example-7cf7d6dbc8-dzpgg   1/1     Running   0          27s
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          28s
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          27s
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          27s

查看pod的标签：

[k8s@server1 ~]$ kubectl get pod --show-labels 
NAME                                  READY   STATUS    RESTARTS   AGE   LABELS
deployment-example-7cf7d6dbc8-dzpgg   1/1     Running   0          13m   app=nginx,pod-template-hash=7cf7d6dbc8
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          13m   app=nginx,pod-template-hash=7cf7d6dbc8
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          13m   app=nginx,pod-template-hash=7cf7d6dbc8
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          13m   app=nginx,pod-template-hash=7cf7d6dbc8

过滤标签

[k8s@server1 ~]$ kubectl get pod -l app
NAME                                  READY   STATUS    RESTARTS   AGE
deployment-example-7cf7d6dbc8-dzpgg   1/1     Running   0          16m
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          16m
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          16m
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          16m
[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS    RESTARTS   AGE   APP
deployment-example-7cf7d6dbc8-dzpgg   1/1     Running   0          17m   nginx
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          17m   nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          17m   nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          17m   nginx

修改标签：

[k8s@server1 ~]$ kubectl label pod deployment-example-7cf7d6dbc8-dzpgg app=apache --overwrite 
pod/deployment-example-7cf7d6dbc8-dzpgg labeled
[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS    RESTARTS   AGE   APP
deployment-example-7cf7d6dbc8-b26t7   1/1     Running   0          5s    nginx
deployment-example-7cf7d6dbc8-dzpgg   1/1     Running   0          18m   apache
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          18m   nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          18m   nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          18m   nginx
[k8s@server1 ~]$ kubectl delete pod deployment-example-7cf7d6dbc8-dzpgg 
pod "deployment-example-7cf7d6dbc8-dzpgg" deleted
[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS    RESTARTS   AGE     APP
deployment-example-7cf7d6dbc8-b26t7   1/1     Running   0          5m39s   nginx
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          24m     nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          24m     nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          24m     nginx
[k8s@server1 ~]$ kubectl delete pod deployment-example-7cf7d6dbc8-b26t7 
pod "deployment-example-7cf7d6dbc8-b26t7" deleted
[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS    RESTARTS   AGE   APP
deployment-example-7cf7d6dbc8-dx7qd   1/1     Running   0          77s   nginx
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          25m   nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          25m   nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          25m   nginx
[k8s@server1 ~]$

因为控制器的筛选条件是标签为nginx的容器，当其中一个容器标签被更改以后，脱离了控制器的控制，又因为控制器设置副本数为4，所以控制器会直接创建一个容器，所以标签为nginx的pod被删以后会自动重建，而标签不为nginx的pod不会自动重建，这就是有没有控制器的区别。如果将apache的标签改回nginx,此时标签为nginx的容器为5个，控制器则会删除多余的容器。

[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS    RESTARTS   AGE     APP
deployment-example-7cf7d6dbc8-9ktfc   1/1     Running   0          4s      nginx
deployment-example-7cf7d6dbc8-dx7qd   1/1     Running   0          8m31s   apache
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          32m     nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          32m     nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          32m     nginx
[k8s@server1 ~]$ kubectl label pod deployment-example-7cf7d6dbc8-dx7qd app=nginx --overwrite 
pod/deployment-example-7cf7d6dbc8-dx7qd labeled
[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS        RESTARTS   AGE     APP
deployment-example-7cf7d6dbc8-9ktfc   1/1     Running       0          19s     nginx
deployment-example-7cf7d6dbc8-dx7qd   1/1     Terminating   0          8m46s   nginx
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running       0          33m     nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running       0          33m     nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running       0          33m     nginx
[k8s@server1 ~]$ kubectl get pod -L app
NAME                                  READY   STATUS    RESTARTS   AGE     APP
deployment-example-7cf7d6dbc8-9ktfc   1/1     Running   0          3m25s   nginx
deployment-example-7cf7d6dbc8-r2qz8   1/1     Running   0          36m     nginx
deployment-example-7cf7d6dbc8-sx4mv   1/1     Running   0          36m     nginx
deployment-example-7cf7d6dbc8-v9s2d   1/1     Running   0          36m     nginx

Node标签

设置server2 node的标签 disktype=ssd

[k8s@server1 ~]$ kubectl label node server2 
.bash_history      .bash_profile      .config/           kube-flannel.yaml  nginx.yml          .viminfo           
.bash_logout       .bashrc            .kube/             .local/            pod.yml            
[k8s@server1 ~]$ kubectl label node server2 disktype=ssd
node/server2 labeled
[k8s@server1 ~]$ kubectl get nodes --show-labels 
NAME      STATUS   ROLES                  AGE     VERSION   LABELS
server1   Ready    control-plane,master   2d21h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server1,kubernetes.io/os=linux,node-role.kubernetes.io/control-plane=,node-role.kubernetes.io/master=
server2   Ready    <none>                 2d20h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,disktype=ssd,kubernetes.io/arch=amd64,kubernetes.io/hostname=server2,kubernetes.io/os=linux
server3   Ready    <none>                 2d9h    v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server3,kubernetes.io/os=linux

设置控制器把pod调度到带有disktype=ssd的node上

apiVersion: apps/v1
kind: Deployment
metadata: 
  name: deployment-example
spec:
  replicas: 4
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: IfNotPresent
      nodeSelector:
        disktype: ssd

都调度到了server2上

[k8s@server1 ~]$ kubectl get pod -o wide
NAME                                  READY   STATUS    RESTARTS   AGE     IP            NODE      NOMINATED NODE   READINESS GATES
deployment-example-54b8d4d9f6-757ln   1/1     Running   0          3m14s   10.244.1.24   server2   <none>           <none>
deployment-example-54b8d4d9f6-8rq5r   1/1     Running   0          3m14s   10.244.1.23   server2   <none>           <none>
deployment-example-54b8d4d9f6-qbwlq   1/1     Running   0          3m8s    10.244.1.25   server2   <none>           <none>
deployment-example-54b8d4d9f6-vlqgc   1/1     Running   0          3m6s    10.244.1.26   server2   <none>           <none>

将server2标签去掉

[k8s@server1 ~]$ kubectl label nodes server2 disktype-
node/server2 labeled
[k8s@server1 ~]$ kubectl get nodes --show-labels                        
NAME      STATUS   ROLES                  AGE     VERSION   LABELS
server1   Ready    control-plane,master   2d21h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server1,kubernetes.io/os=linux,node-role.kubernetes.io/control-plane=,node-role.kubernetes.io/master=
server2   Ready    <none>                 2d20h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server2,kubernetes.io/os=linux
server3   Ready    <none>                 2d10h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server3,kubernetes.io/os=linux

此时将副本数改为8个

[k8s@server1 ~]$ kubectl apply -f nginx.yml 
deployment.apps/deployment-example configured
[k8s@server1 ~]$ kubectl get pod -o wide
NAME                                  READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
deployment-example-54b8d4d9f6-757ln   1/1     Running   0          10m   10.244.1.24   server2   <none>           <none>
deployment-example-54b8d4d9f6-8rq5r   1/1     Running   0          10m   10.244.1.23   server2   <none>           <none>
deployment-example-54b8d4d9f6-bs5w7   0/1     Pending   0          5s    <none>        <none>    <none>           <none>
deployment-example-54b8d4d9f6-hqvmf   0/1     Pending   0          5s    <none>        <none>    <none>           <none>
deployment-example-54b8d4d9f6-p25cc   0/1     Pending   0          5s    <none>        <none>    <none>           <none>
deployment-example-54b8d4d9f6-qbwlq   1/1     Running   0          10m   10.244.1.25   server2   <none>           <none>
deployment-example-54b8d4d9f6-vlqgc   1/1     Running   0          10m   10.244.1.26   server2   <none>           <none>
deployment-example-54b8d4d9f6-w9llw   0/1     Pending   0          5s    <none>        <none>    <none>           <none>
[k8s@server1 ~]$ kubectl describe pod deployment-example-54b8d4d9f6-bs5w7 
Events:
  Type     Reason            Age                From               Message
  ----     ------            ----               ----               -------
  Warning  FailedScheduling  34s (x2 over 34s)  default-scheduler  0/3 nodes are available: 1 node(s) had taint {node-role.kubernetes.io/master: }, that the pod didn't tolerate, 2 node(s) didn't match Pod's node affinity.

再给server3加上标签

[k8s@server1 ~]$ kubectl label nodes server3 disktype=ssd
node/server3 labeled
[k8s@server1 ~]$ kubectl get nodes --show-labels 
NAME      STATUS   ROLES                  AGE     VERSION   LABELS
server1   Ready    control-plane,master   2d21h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server1,kubernetes.io/os=linux,node-role.kubernetes.io/control-plane=,node-role.kubernetes.io/master=
server2   Ready    <none>                 2d21h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=server2,kubernetes.io/os=linux
server3   Ready    <none>                 2d10h   v1.20.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,disktype=ssd,kubernetes.io/arch=amd64,kubernetes.io/hostname=server3,kubernetes.io/os=linux

[k8s@server1 ~]$ kubectl get pod -o wide
NAME                                  READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
deployment-example-54b8d4d9f6-757ln   1/1     Running   0          36m   10.244.1.24   server2   <none>           <none>
deployment-example-54b8d4d9f6-8rq5r   1/1     Running   0          36m   10.244.1.23   server2   <none>           <none>
deployment-example-54b8d4d9f6-bs5w7   1/1     Running   0          26m   10.244.3.18   server3   <none>           <none>
deployment-example-54b8d4d9f6-hqvmf   1/1     Running   0          26m   10.244.3.19   server3   <none>           <none>
deployment-example-54b8d4d9f6-p25cc   1/1     Running   0          26m   10.244.3.20   server3   <none>           <none>
deployment-example-54b8d4d9f6-qbwlq   1/1     Running   0          36m   10.244.1.25   server2   <none>           <none>
deployment-example-54b8d4d9f6-vlqgc   1/1     Running   0          36m   10.244.1.26   server2   <none>           <none>
deployment-example-54b8d4d9f6-w9llw   1/1     Running   0          26m   10.244.3.21   server3   <none>           <none>

可以发现k8s node上的标签只会影响未调度的pod，对已经调度的pod不会生效
Pod 在其生命周期中只会被调度一次。一旦 Pod 被调度（分派）到某个节点，Pod 会一直在该节点运行，直到 Pod 停止或者被终止。

Pod的状态及生命周期

在这里插入图片描述

Init容器

官方文档
每个 Pod 中可以包含多个容器，应用运行在这些容器里面，同时 Pod 也可以有一个或多个先于应用容器启动的 Init 容器。
Init 容器与普通的容器非常像，除了如下两点：

它们总是运行到完成。
每个都必须在下一个启动之前成功完成。

如果 Pod 的 Init 容器失败，kubelet 会不断地重启该 Init 容器直到该容器成功为止。然而，如果 Pod 对应的 restartPolicy 值为 “Never”，Kubernetes 不会重新启动 Pod。

apiVersion: v1
Kind: Pod
metadata:
  name: myapp-pod
  labels:
    app: myapp
spec:
  containers:
  - name: myapp-container
    image: nginx
    imagePullPolicy: IfNotPresent
    initContainers:
      - name: init-myservices
        image: busyboxplus
        imagePullPolicy: IfNotPresent
        command: ['sh', '-c', "until nslookup myservice.default.svc.cluster.local; do echo waiting for myservices; sleep 2; done"]

启动后 kubectl describe pod

init-myservices:
    Container ID:  docker://fe8985e5895ea8e7557b74c86f298e4a3405396b0b98cf58a86ff8f7397683d0
    Image:         busyboxplus
    Image ID:      docker-pullable://busyboxplus@sha256:9d1c242c1fd588a1b8ec4461d33a9ba08071f0cc5bb2d50d4ca49e430014ab06
    Port:          <none>
    Host Port:     <none>
    Command:
      sh
      -c
      until nslookup myservice.default.svc.cluster.local; do echo waiting for myservices; sleep 2; done
    State:          Running
      Started:      Sun, 27 Dec 2020 14:22:42 +0800
    Ready:          False
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-4dd4l (ro)

在这里插入图片描述
创建一个service

kind: Service
apiVersion: v1
metadata:
  name: myservice
spec:
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80

[k8s@server1 ~]$ kubectl get svc
NAME         TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)   AGE
kubernetes   ClusterIP   10.96.0.1       <none>        443/TCP   14d
myservice    ClusterIP   10.104.20.158   <none>        80/TCP    33s

解析到域名，容器启动

Init Containers:
  init-myservices:
    Container ID:  docker://fe8985e5895ea8e7557b74c86f298e4a3405396b0b98cf58a86ff8f7397683d0
    Image:         busyboxplus
    Image ID:      docker-pullable://busyboxplus@sha256:9d1c242c1fd588a1b8ec4461d33a9ba08071f0cc5bb2d50d4ca49e430014ab06
    Port:          <none>
    Host Port:     <none>
    Command:
      sh
      -c
      until nslookup myservice.default.svc.cluster.local; do echo waiting for myservices; sleep 2; done
    State:          Terminated
      Reason:       Completed
      Exit Code:    0
      Started:      Sun, 27 Dec 2020 14:22:42 +0800
      Finished:     Sun, 27 Dec 2020 14:31:34 +0800
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-4dd4l (ro)

在这里插入图片描述

配置探针

官方文档

存活探针

apiVersion: apps/v1
kind: Deployment
metadata: 
  name: deployment-example
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: IfNotPresent
        livenessProbe:
          tcpSocket:
            port: 80
          initialDelaySeconds: 3
          periodSeconds: 2
          timeoutSeconds: 3
      initContainers:
      - name: init-myservices
        image: busyboxplus
        imagePullPolicy: IfNotPresent
        command: ['sh', '-c', "until nslookup myservice.default.svc.cluster.local; do echo waiting for myservices; sleep 2; done"]

kubectl describe pod

Liveness:       tcp-socket :80 delay=3s timeout=3s period=2s #success=1 #failure=3

停止后容器直接退出
在这里插入图片描述

就绪探针

apiVersion: apps/v1
kind: Deployment
metadata: 
  name: deployment-example
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: IfNotPresent
        livenessProbe:
          tcpSocket:
            port: 80
          initialDelaySeconds: 3
          periodSeconds: 2
          timeoutSeconds: 3
        readinessProbe:
          httpGet:
            path: /test.html
            port: 80
          initialDelaySeconds: 1
          periodSeconds: 3
          timeoutSeconds: 1
      initContainers:
      - name: init-myservices
        image: busyboxplus
        imagePullPolicy: IfNotPresent
        command: ['sh', '-c', "until nslookup myservice.default.svc.cluster.local; do echo waiting for myservices; sleep 2; done"]

虽然容器running但没有ready

[k8s@server1 ~]$ kubectl get pod
NAME                                  READY   STATUS    RESTARTS   AGE
deployment-example-557997bffc-5f2js   0/1     Running   0          70s
deployment-example-557997bffc-tjrgx   0/1     Running   0          70s

kubectl describe pod

 Warning  Unhealthy  2m34s (x19 over 3m28s)  kubelet            Readiness probe failed: HTTP probe failed with statuscode: 404

创建test.html

[k8s@server1 ~]$ kubectl exec deployment-example-557997bffc-5f2js -it -- bash
root@deployment-example-557997bffc-5f2js:/# echo hello world > /usr/share/nginx/html/test.html
root@deployment-example-557997bffc-5f2js:/# curl localhost/test.html
hello world
root@deployment-example-557997bffc-5f2js:/# exit
[k8s@server1 ~]$ kubectl get pod
NAME                                  READY   STATUS    RESTARTS   AGE
deployment-example-557997bffc-5f2js   1/1     Running   0          6m2s
deployment-example-557997bffc-tjrgx   0/1     Running   0          6m2s

可以发现在node1上ready了
在这里插入图片描述

K8S/Kubernetes

K8S/Kubernetes社区为您提供最前沿的新闻资讯和知识内容

更多推荐

【深度】阿里巴巴万级规模 K8s 集群全局高可用体系之美

作者 | 韩堂、柘远、沉醉来源 | 阿里巴巴云原生公众号前言台湾作家林清玄在接受记者采访的时候，如此评价自己 30 多年写作生涯：“第一个十年我才华横溢，‘贼光闪现’，令周边黯然失色；第二个十年，我终于‘宝光现形’，不再去抢风头，反而与身边的美丽相得益彰；进入第三个十年，繁华落尽见真醇，我进入了‘醇光初现’的阶段，真正体味到了境界之美”。长夜有穷，真水无香。领略过了 K8s“身在江

K8S/Kubernetes

如何基于 K8s 构建下一代 DevOps 平台？

作者 | 孙健波（天元）导读：当前云原生 DevOps 体系现状如何？面临哪些挑战？如何通过 OAM 解决云原生 DevOps 场景下的诸多问题？云原生开发应用模型 OAM(Open Application Model) 社区核心成员孙健波将为大家一一解答，并分享如何基于 OAM 和 Kubernetes 打造无限能力的下一代 DevOps 平台。什么是 DevOps？为什么基于 Kub