当前位置：首页 > article >正文

kubernetes学习-应用程序的生命周期管理

article 2025/1/16 12:36:58

kubernetes学习-应用程序的生命周期管理

- 一、应用的部署
- - 1.1 通过命令进行部署
  - 1.2 通过yaml进行部署
  - 1.3 部署示例
  - 1.3.1 部署deployment
  - 1.3.2 部署service
  - 1.3.3 访问nginx页面
- 二、应用的升级
- - 2.1 通过set 升级
  - 2.2 通过yaml升级
  - 2.3 通过编辑升级
- 三、应用的回滚
- 四、应用的扩容与缩容
- - 4.1 通过命令扩容和缩容
  - 4.2 通过yaml扩容和缩容
- 五、应用的下线

一、应用的部署

应用程序的部署方式有两种，一种是通过命令直接进行部署，另一种是通过yaml文件进行部署。

1.1 通过命令进行部署

示例，具体参数和说明可自行查看官方说明或百度。

 ~]# kubectl create deployment {pod名称} --image=镜像名:版本 --replicas=副本数

1.2 通过yaml进行部署

可使用官方yaml或命令生成yaml。
生成yaml示例：

 ~]# kubectl create deployment {pod名称} --image={镜像名}:{镜像版本} -o yaml --dry-run=client > xxx.yaml
# --dry-run=client    表示本地尝试运行pod，主要用于测试是否正常
 ~]# kubectl get deployment {pod名称} -o yaml > xxxx.yaml   # 导出现有pod配置到yaml

部署命令示例

~]# kubectl apply -f xxxx.ayml

1.3 部署示例

1.3.1 部署deployment

deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-nginx
spec:
  replicas: 3  # 设置Pod副本数量
  selector:
    matchLabels:
      app: my-nginx
  template:
    metadata:
      labels:
        app: my-nginx
    spec:
      containers:
      - name: my-nginx
        image: nginx:1.16  # 你的容器镜像
        ports:
        - containerPort: 80

~]# kubectl apply -f deployment.yaml
~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-657d778b65-2v745   1/1     Running   0          27m
my-nginx-657d778b65-5bfc6   1/1     Running   0          27m
my-nginx-657d778b65-z7n6l   1/1     Running   0          27m

1.3.2 部署service

service.yaml

apiVersion: v1
kind: Service
metadata:
  name: my-service
spec:
  selector:
    app: my-nginx
  ports:
  - protocol: TCP
    port: 8081
    targetPort: 80
    nodePort: 32222
  type: NodePort

~]# kubectl apply -f service.yaml
~]# kubectl get svc
NAME           TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)          AGE
kubernetes     ClusterIP      10.96.0.1       <none>        443/TCP          8h
my-service     NodePort       10.97.148.243   <none>        8081:32222/TCP   28m

1.3.3 访问nginx页面

访问k8s集群内任意一主机的32222端口
在这里插入图片描述

二、应用的升级

升级可以分为三种方式进行升级，在升级过程以3个副本数为例。当更新 Deployment 时，系统创建了一个新的 ReplicaSet，并将其副本数量扩展到 1，然后将旧的 ReplicaSet 缩减为 2。之后，系统继续按照相同的更新策略对新旧两个ReplicaSet 进行逐个调整。最后，新的ReplicaSet 运行了 3 个新版本 Pod 副本，旧的 ReplicaSet 副本数量则缩减为0。
在这里插入图片描述

2.1 通过set 升级

示例

~]# kubectl set image deployment {pod名称} {pod名称}=镜像名:版本

查看当前镜像版本，可以看到镜像当前版本为nginx:1.16

~]# kubectl describe pod {pod名称} | grep image
~]# kubectl describe pod my-nginx | grep image
  Normal  Pulled     44m   kubelet            Container image "nginx:1.16" already present on machine
  Normal  Pulled     44m   kubelet            Container image "nginx:1.16" already present on machine
  Normal  Pulled     44m   kubelet            Container image "nginx:1.16" already present on machine

可以看到当前镜像为nginx:1.16，通过set升级到1.17版本。

~]# kubectl set image deployment my-nginx my-nginx=nginx:1.17
~]# kubectl describe pod my-nginx | grep image                  # 查看镜像版本，升级成功
  Normal  Pulled     76s   kubelet            Container image "nginx:1.17" already present on machine
  Normal  Pulled     2m9s  kubelet            Container image "nginx:1.17" already present on machine
  Normal  Pulled     83s   kubelet            Container image "nginx:1.17" already present on machine

2.2 通过yaml升级

~]# kubectl apply -f xxxx.ayml

deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-nginx
spec:
  replicas: 3  # 设置Pod副本数量
  selector:
    matchLabels:
      app: my-nginx
  template:
    metadata:
      labels:
        app: my-nginx
    spec:
      containers:
      - name: my-nginx
        image: nginx:1.18  # 你的容器镜像
        ports:
        - containerPort: 80

~]# kubectl apply -f deployment.yaml
deployment.apps/my-nginx configured
~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-657d778b65-2v745   1/1     Running   0          27m
my-nginx-657d778b65-5bfc6   1/1     Running   0          27m
my-nginx-657d778b65-z7n6l   1/1     Running   0          27m
~]# kubectl describe pod my-nginx | grep image
  Normal  Pulled     <invalid>  kubelet            Container image "nginx:1.18" already present on machine
  Normal  Pulled     5s         kubelet            Container image "nginx:1.18" already present on machine
  Normal  Pulled     12s        kubelet            Container image "nginx:1.18" already present on machine

2.3 通过编辑升级

kubectl edit deployment {pod名称}           #使用系统编辑器打开

修改镜像版本为1.16，编辑方式同vim文件，可使用:wq保存退出，保存以后将会立即生效执行。
在这里插入图片描述

~]# kubectl edit deployment my-nginx
deployment.apps/my-nginx edited
~]# kubectl describe pod my-nginx | grep image
  Normal  Pulled     8s         kubelet            Container image "nginx:1.16" already present on machine
  Normal  Pulled     <invalid>  kubelet            Container image "nginx:1.16" already present on machine
  Normal  Pulled     5s         kubelet            Container image "nginx:1.16" already present on machine

三、应用的回滚

应用回滚需要掌握的几个命令

~]# kubectl rollout history deployment {pod名称}       # 查看历史发布版本 
~]# kubectl get rs                                    # 查看RS记录
~]# kubectl describe rs                               # 查看rs的详细信息
~]# kubectl describe rs | grep -E "revision:|Image"   # 查看发布记录对应的镜像版本

可以看到，版本4对应1.16，版本3对应1.18，版本2对应1.17

~]# kubectl rollout history deployment my-nginx
deployment.apps/my-nginx 
REVISION  CHANGE-CAUSE
2         <none>
3         <none>
4         <none>
~]# kubectl describe rs | grep -E "revision:|Image"    
                deployment.kubernetes.io/revision: 4
    Image:        nginx:1.16
                deployment.kubernetes.io/revision: 3
    Image:        nginx:1.18
                deployment.kubernetes.io/revision: 2
    Image:        nginx:1.17

可以看到通过rollout undo可以回滚到指定的版本，以回滚到版本2为例。

 ~]# kubectl rollout undo deployment my-nginx --to-revision=2
deployment.apps/my-nginx rolled back
]# kubectl describe pod my-nginx | grep image
  Normal  Pulled     36s        kubelet            Container image "nginx:1.17" already present on machine
  Normal  Pulled     <invalid>  kubelet            Container image "nginx:1.17" already present on machine
  Normal  Pulled     35s        kubelet            Container image "nginx:1.17" already present on machine

四、应用的扩容与缩容

应用的扩容与缩容即pod副本数的增加或者减少，可以通过yaml和命令行方式进行操作，更新后的副本数大于当前副本数则为扩容，反之则为缩容。

4.1 通过命令扩容和缩容

可以通过scale命令指定–replicas副本数

~]# kubectl scale deployment {pod名称} --replicas={副本数}

~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-7c4f7df846-dcswr   1/1     Running   0          7m7s
my-nginx-7c4f7df846-ngdns   1/1     Running   0          7m8s
my-nginx-7c4f7df846-wn6zd   1/1     Running   0          7m5s
~]# kubectl scale deployment my-nginx --replicas=6		    # 增加副本数
deployment.apps/my-nginx scaled
[root@node-02 ~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-7c4f7df846-dcswr   1/1     Running   0          8m18s
my-nginx-7c4f7df846-g57g7   1/1     Running   0          3s
my-nginx-7c4f7df846-j2j25   1/1     Running   0          3s
my-nginx-7c4f7df846-ngdns   1/1     Running   0          8m19s
my-nginx-7c4f7df846-wn6zd   1/1     Running   0          8m16s
my-nginx-7c4f7df846-xh925   1/1     Running   0          3s
~]# kubectl scale deployment my-nginx --replicas=2			# 减少副本数
deployment.apps/my-nginx scaled
[root@node-02 ~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-7c4f7df846-ngdns   1/1     Running   0          9m16s
my-nginx-7c4f7df846-wn6zd   1/1     Running   0          9m13s

4.2 通过yaml扩容和缩容

deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-nginx
spec:
  replicas: 5  # 设置Pod副本数量
  selector:
    matchLabels:
      app: my-nginx
  template:
    metadata:
      labels:
        app: my-nginx
    spec:
      containers:
      - name: my-nginx
        image: nginx:1.17  # 你的容器镜像
        ports:
        - containerPort: 80

~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-7c4f7df846-ngdns   1/1     Running   0          9m16s
my-nginx-7c4f7df846-wn6zd   1/1     Running   0          9m13s
~]# kubectl apply -f deployment.yaml
deployment.apps/my-nginx configured
~]# kubectl get pod
NAME                        READY   STATUS    RESTARTS   AGE
my-nginx-7c4f7df846-c56b9   1/1     Running   0          5s
my-nginx-7c4f7df846-mbx92   1/1     Running   0          5s
my-nginx-7c4f7df846-mrbgl   1/1     Running   0          5s
my-nginx-7c4f7df846-ngdns   1/1     Running   0          13m
my-nginx-7c4f7df846-wn6zd   1/1     Running   0          13m

五、应用的下线

需要删除svc和deployment

~]# kubectl delete deployment {pod名称}
~]# kubectl delete svc {svc名称}

~]# kubectl delete deployment my-nginx
deployment.apps "my-nginx" deleted
~]# kubectl delete svc my-service
service "my-service" deleted
~]# kubectl get svc
NAME           TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes     ClusterIP      10.96.0.1       <none>        443/TCP        9h