K8S编排 之 Deployment

发表于 更新于 分类于 阅读次数：

目录

引入

  • K8S的Pod解决了容器的”超亲密关系”设计

  • K8S的Deployment解决了Pod的”水平扩展/伸缩”设计

使用

1
vim deploy.yml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    app: nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.20.1
        ports:
        - containerPort: 80
1
2
3
kubectl apply -f deploy.yml

kubectl get deployments
1
2
NAME               READY   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment   3/3     3            3           55s
1
kubectl get pods
1
2
3
4
NAME                                READY   STATUS    RESTARTS   AGE
nginx-deployment-58b9b8ff79-bbjgq   1/1     Running   0          64s
nginx-deployment-58b9b8ff79-ghmww   1/1     Running   0          64s
nginx-deployment-58b9b8ff79-lxcqf   1/1     Running   0          64s

k8s-deployment-01.png

特性

水平扩展

如何水平扩展服务的能力?

1
2
kubectl scale deployment nginx-deployment --replicas=4
# deployment.apps/nginx-deployment scaled
1
kubectl get deployments
1
2
NAME               READY   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment   4/4     4            4           87s
1
kubectl get pods
1
2
3
4
5
NAME                                READY   STATUS    RESTARTS   AGE
nginx-deployment-58b9b8ff79-bbjgq   1/1     Running   0          95s
nginx-deployment-58b9b8ff79-ghmww   1/1     Running   0          95s
nginx-deployment-58b9b8ff79-lxcqf   1/1     Running   0          95s
nginx-deployment-58b9b8ff79-nmv2h   1/1     Running   0          19s

上述kubectl apply命令是一种声明式API 即描述”我所期望的[最终]结果” 而命令式操作描述的是”我所想要做的操作”

滚动更新

如何实现服务的平滑升级?

1
2
3
4
5
kubectl delete -f deploy.yml

kubectl apply -f deploy.yml

kubectl get pods
1
2
3
4
NAME                                READY   STATUS    RESTARTS   AGE
nginx-deployment-58b9b8ff79-dzfnj   1/1     Running   0          4s
nginx-deployment-58b9b8ff79-p56jb   1/1     Running   0          4s
nginx-deployment-58b9b8ff79-wlwq4   1/1     Running   0          4s
1
2
vim deploy.yml
# image: nginx:1.21.0
1
2
3
kubectl apply -f deploy.yml

kubectl get pods
1
2
3
4
5
6
NAME                                READY   STATUS              RESTARTS   AGE
nginx-deployment-58b9b8ff79-dzfnj   1/1     Running             0          55s
nginx-deployment-58b9b8ff79-p56jb   1/1     Running             0          55s
nginx-deployment-58b9b8ff79-wlwq4   1/1     Terminating         0          55s
nginx-deployment-6c4ccd94bc-4fz95   1/1     Running             0          3s
nginx-deployment-6c4ccd94bc-5r8wb   0/1     ContainerCreating   0          1s
1
2
# 稍等一会后
kubectl get pods
1
2
3
4
NAME                                READY   STATUS    RESTARTS   AGE
nginx-deployment-6c4ccd94bc-4fz95   1/1     Running   0          30s
nginx-deployment-6c4ccd94bc-5r8wb   1/1     Running   0          28s
nginx-deployment-6c4ccd94bc-qdjxp   1/1     Running   0          32s
1
kubectl describe deployment nginx-deployment
1
2
3
4
5
6
7
8
9
10
Events:
  Type    Reason             Age    From                   Message
  ----    ------             ----   ----                   -------
  Normal  ScalingReplicaSet  2m11s  deployment-controller  Scaled up replica set nginx-deployment-58b9b8ff79 to 3
  Normal  ScalingReplicaSet  79s    deployment-controller  Scaled up replica set nginx-deployment-6c4ccd94bc to 1
  Normal  ScalingReplicaSet  77s    deployment-controller  Scaled down replica set nginx-deployment-58b9b8ff79 to 2
  Normal  ScalingReplicaSet  77s    deployment-controller  Scaled up replica set nginx-deployment-6c4ccd94bc to 2
  Normal  ScalingReplicaSet  75s    deployment-controller  Scaled down replica set nginx-deployment-58b9b8ff79 to 1
  Normal  ScalingReplicaSet  75s    deployment-controller  Scaled up replica set nginx-deployment-6c4ccd94bc to 3
  Normal  ScalingReplicaSet  73s    deployment-controller  Scaled down replica set nginx-deployment-58b9b8ff79 to 0

k8s-deployment-02.png

上述新旧replica set中的Pod交替升级的过程 就是滚动更新

回滚更新

服务更新出错时如何回滚?

1
2
vim deploy.yml
# image: nginx:1.121.0 => 错误的镜像名称
1
2
3
kubectl apply -f deploy.yml

kubectl get pods
1
2
3
4
5
NAME                                READY   STATUS             RESTARTS   AGE
nginx-deployment-6c4ccd94bc-4fz95   1/1     Running            0          94m
nginx-deployment-6c4ccd94bc-5r8wb   1/1     Running            0          94m
nginx-deployment-6c4ccd94bc-qdjxp   1/1     Running            0          94m
nginx-deployment-7964764c96-drvjx   0/1     ImagePullBackOff   0          85s
1
2
3
4
5
6
kubectl rollout history deployment/nginx-deployment

kubectl rollout history deployment/nginx-deployment --revision=3
# Image:	nginx:1.121.0

kubectl rollout undo deployment/nginx-deployment --to-revision=2
1
kubectl get pods
1
2
3
4
NAME                                READY   STATUS    RESTARTS   AGE
nginx-deployment-6c4ccd94bc-4fz95   1/1     Running   0          97m
nginx-deployment-6c4ccd94bc-5r8wb   1/1     Running   0          97m
nginx-deployment-6c4ccd94bc-qdjxp   1/1     Running   0          96m

为了便于查看更新记录 可以添加参数”–record” 形如”kubectl apply -f deploy.yml –record”