18 Deployment 声明式地升级应用

发表于 | 分类于


使用 RC 实现自动滚动升级

先定义两个容器镜像 v1 和 v2 版本

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
# v1 版本
[root@k8s-master2 v1]# cat app.js 
const http = require('http');
const os = require('os');

console.log("Kubia server starting...");

var handler = function(request, response) {
  console.log("Received request from " + request.connection.remoteAddress);
  response.writeHead(200);
  response.end("This is v1 running in pod " + os.hostname() + "\n");
};

var www = http.createServer(handler);
www.listen(8080);

[root@k8s-master2 v1]# cat Dockerfile 
FROM node:7
ADD app.js /app.js
ENTRYPOINT ["node", "app.js"]

[root@k8s-master2 v1]# docker build -t 172.31.228.68/game/kubia:v1 .
[root@k8s-master2 v1]# docker push 172.31.228.68/game/kubia:v1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
# v2 版本 区别在于返回的是 v2的信息


[root@k8s-master2 v2]# docker build -t 172.31.228.68/game/kubia:v2 .
[root@k8s-master2 v2]# cat app.js 
const http = require('http');
const os = require('os');

console.log("Kubia server starting...");

var handler = function(request, response) {
  console.log("Received request from " + request.connection.remoteAddress);
  response.writeHead(200);
  response.end("This is v2 running in pod " + os.hostname() + "\n");
};

var www = http.createServer(handler);
www.listen(8080);

[root@k8s-master2 v2]# cat Dockerfile 
FROM node:7
ADD app.js /app.js
ENTRYPOINT ["node", "app.js"]

[root@k8s-master2 v2]# docker build -t 172.31.228.68/game/kubia:v2 .
[root@k8s-master2 v2]# docker push 172.31.228.68/game/kubia:v2

单个 yaml 运行 rc 和 service

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
# 多种资源定义通过 --- 分割

[root@k8s-master1 deployment]# vim kubia-rc-svc-v1.yaml 

apiVersion: v1
kind: ReplicationController
metadata:
  name: kubia-v1
spec:
  replicas: 3
  template:
    metadata:
      name: kubia
      labels:
        app: kubia
    spec:
      imagePullSecrets:
      - name: my-harbor-secret
      containers:
      - image: 172.31.228.68/game/kubia:v1
        name: nodejs
---
apiVersion: v1
kind: Service
metadata:
  name: kubia
spec:
  type: NodePort
  selector:
    app: kubia
  ports:
  - port: 80
    targetPort: 8080
    nodePort: 30123
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
[root@k8s-master1 deployment]# kubectl get pods,svc
NAME                                          READY   STATUS    RESTARTS   AGE
pod/kubia-v1-5cgjx                            1/1     Running   0          2m21s
pod/kubia-v1-tbhqr                            1/1     Running   0          2m21s
pod/kubia-v1-th6rr                            1/1     Running   0          2m21s
pod/nfs-client-provisioner-56f4b98d47-v4nf6   1/1     Running   8          3d14h

NAME                 TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)        AGE
service/kubernetes   ClusterIP   10.0.0.1     <none>        443/TCP        13d
service/kubia        NodePort    10.0.0.119   <none>        80:30123/TCP   2m21s

# 测试程序一直循环访问
[root@k8s-master1 demo]# while true;do curl http://10.0.0.119;sleep 1; done
This is v1 running in pod kubia-v1-tbhqr
This is v1 running in pod kubia-v1-th6rr
This is v1 running in pod kubia-v1-5cgjx

RC 的滚动升级 rolling-update

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
[root@k8s-master1 deployment]# kubectl rolling-update kubia-v1 kubia-v2 --image=172.31.228.68/game/kubia:v2
Command "rolling-update" is deprecated, use "rollout" instead
Created kubia-v2
Scaling up kubia-v2 from 0 to 3, scaling down kubia-v1 from 3 to 0 (keep 3 pods available, don't exceed 4 pods)
Scaling kubia-v2 up to 1
Scaling kubia-v1 down to 2
Scaling kubia-v2 up to 2
Scaling kubia-v1 down to 1
Scaling kubia-v2 up to 3
Scaling kubia-v1 down to 0
Update succeeded. Deleting kubia-v1
replicationcontroller/kubia-v2 rolling updated to "kubia-v2"


[root@k8s-master1 ~]# kubectl describe rc kubia-v2
Name:         kubia-v2
Namespace:    default
Selector:     app=kubia,deployment=1c9b94fec2662e272187373534aef5c6
Labels:       app=kubia
...

[root@k8s-master1 ~]# kubectl describe rc kubia-v1
Name:         kubia-v1
Namespace:    default
Selector:     app=kubia,deployment=9b82b6cdfb3cd18a5e53068dda002fe1-orig
Labels:       app=kubia
...

# 查看 Pod
[root@k8s-master1 ~]# kubectl get pods --show-labels
NAME                                      READY   STATUS    RESTARTS   AGE     LABELS
kubia-v1-tbhqr                            1/1     Running   0          18m     app=kubia,deployment=9b82b6cdfb3cd18a5e53068dda002fe1-orig
kubia-v2-p8vbl                            1/1     Running   0          3m15s   app=kubia,deployment=1c9b94fec2662e272187373534aef5c6
kubia-v2-x7g6g                            1/1     Running   0          2m8s    app=kubia,deployment=1c9b94fec2662e272187373534aef5c6
kubia-v2-xr8hz                            1/1     Running   0          62s     app=kubia,deployment=1c9b94fec2662e272187373534aef5c6
nfs-client-provisioner-56f4b98d47-v4nf6   1/1     Running   8          3d14h   app=nfs-client-provisioner,pod-template-hash=56f4b98d47

# v1的pod被滚动升级到v2版本 并且多了一个标签 deployment=9b82...

[root@k8s-master1 deployment]# kubectl get pods -o wide
NAME                                      READY   STATUS    RESTARTS   AGE     IP            NODE          NOMINATED NODE   READINESS GATES
kubia-v2-p8vbl                            1/1     Running   0          5m54s   10.244.1.75   k8s-node2     <none>           <none>
kubia-v2-x7g6g                            1/1     Running   0          4m47s   10.244.0.61   k8s-node1     <none>           <none>
kubia-v2-xr8hz                            1/1     Running   0          3m41s   10.244.2.62   k8s-master1   <none>           <none>
nfs-client-provisioner-56f4b98d47-v4nf6   1/1     Running   8          3d14h   10.244.0.59   k8s-node1     <none>           <none>

# 请求也都到了 v2 
This is v2 running in pod kubia-v2-x7g6g
This is v2 running in pod kubia-v2-p8vbl
This is v2 running in pod kubia-v2-xr8hz

# rolling-update 已过时,现在使用 deployment 
# deployment 是否可以通过修改pod中的镜像 升级应用 而不是通过删除手法的自动伸缩,应该通过期望副本数来做升级

使用 Deployment 声明式升级应用

1
2
3
1. Deployment -> rs -> pod
2. Deployment 是一种更高级资源 用于部署应用程序，并且用声明式方式更新应用
3. Deployment 被创建时也会创建 ReplicaSet ，在使用Deployment时实际上Pod是由ReplicaSet创建和管理的

创建 Deployment

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
[root@k8s-master1 deployment]# kubectl explain deployment
KIND:     Deployment
VERSION:  apps/v1

[root@k8s-master1 deployment]# vim kubia-deployment-v1.yaml 

apiVersion: apps/v1
kind: Deployment
metadata:
  name: kubia
spec:
  replicas: 3
  selector:
    matchLabels:
      app: kubia
  template:
    metadata:
      name: kubia
      labels:
        app: kubia
    spec:
      containers:
      - image: 172.31.228.68/game/kubia:v1
        name: nodejs


[root@k8s-master1 deployment]# kubectl delete rc --all
replicationcontroller "kubia-v2" deleted

# 创建deployment并记录版本号
[root@k8s-master1 deployment]# kubectl create -f kubia-deployment-v1.yaml --record
deployment.apps/kubia created

[root@k8s-master1 deployment]# kubectl get pods -o wide
NAME                                      READY   STATUS        RESTARTS   AGE     IP            NODE          NOMINATED NODE   READINESS GATES
kubia-dbd9664c7-4mggm                     1/1     Running       0          35s     10.244.2.64   k8s-master1   <none>           <none>
kubia-dbd9664c7-8b629                     1/1     Running       0          35s     10.244.1.77   k8s-node2     <none>           <none>
kubia-dbd9664c7-8npt4                     1/1     Running       0          35s     10.244.0.63   k8s-node1     <none>           <none>

[root@k8s-master1 deployment]# kubectl get rs
NAME                                DESIRED   CURRENT   READY   AGE
kubia-dbd9664c7                     3         3         3       93s

# 别忘记deployment是通过rs来创建pod 所有rs的名字中包含了pod模板的哈希值
# Deployment可以创建多个RS来对应管理一组Pod模板
# 可以通过RS这个值去区分 

[root@k8s-master1 demo]# kubectl get pod --show-labels
NAME                                      READY   STATUS    RESTARTS   AGE     LABELS
kubia-dbd9664c7-4mggm                     1/1     Running   0          5m36s   app=kubia,pod-template-hash=dbd9664c7
kubia-dbd9664c7-8b629                     1/1     Running   0          5m36s   app=kubia,pod-template-hash=dbd9664c7
kubia-dbd9664c7-8npt4                     1/1     Running   0          5m36s   app=kubia,pod-template-hash=dbd9664c7

升级 Deployment

1
2
3
Deployment 有两种升级策略
1. 默认滚动更新 渐进删除旧Pod,同时创建新Pod ，整个应用在升级中都处于可用状态  RollingUpdate
2. 一次删除所有旧Pod,然后创建新Pod 应用如果不支持多版本同时对外提供服务,就需要再启动新版本之前停止旧版本，会导致短期业务无不可用 Recre-ate

减慢滚动升级速度

1
2
3
4
5
6
7
8
9
10
# 用来测试查看观察升级过程 
# 设置为10秒
# kubectl patch 可以用来修改单个资源信息
[root@k8s-master1 demo]# kubectl patch deployment kubia -p '{"spec": {"minReadySeconds": 10}}'
deployment.apps/kubia patched

# 查看下
[root@k8s-master1 deployment]# kubectl edit deploy kubia
spec:
  minReadySeconds: 10

修改镜像 触发滚动升级

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
[root@k8s-master1 demo]# while true;do curl http://10.0.0.119;sleep 1; done
This is v1 running in pod kubia-dbd9664c7-8npt4
This is v1 running in pod kubia-dbd9664c7-4mggm
This is v1 running in pod kubia-dbd9664c7-8b629

[root@k8s-master1 deployment]# kubectl set image deployment kubia nodejs=172.31.228.68/game/kubia:v2
deployment.apps/kubia image updated

# 滚动升级过程 一个个的将v1版本pod 替换成 v2 pod 
This is v2 running in pod kubia-56fd945996-lz847
This is v2 running in pod kubia-56fd945996-glj6d
This is v2 running in pod kubia-56fd945996-ccg9c

[root@k8s-master1 deployment]# kubectl get pods -o wide
NAME                                      READY   STATUS        RESTARTS   AGE     IP            NODE          NOMINATED NODE   READINESS GATES
kubia-56fd945996-ccg9c                    1/1     Running       0          61s     10.244.1.78   k8s-node2     <none>           <none>
kubia-56fd945996-glj6d                    1/1     Running       0          49s     10.244.0.64   k8s-node1     <none>           <none>
kubia-56fd945996-lz847                    1/1     Running       0          38s     10.244.2.65   k8s-master1   <none>           <none>
kubia-dbd9664c7-8b629                     1/1     Terminating   0          20m     10.244.1.77   k8s-node2     <none>           <none>

# 升级过程
# 1. 创建新的的RS(v2) 并且慢慢扩容
# 2. 之前版本(v1) 会缩容至0

[root@k8s-master1 deployment]# kubectl get rs
NAME                                DESIRED   CURRENT   READY   AGE
kubia-56fd945996                    3         3         3       5m28s
kubia-dbd9664c7                     0         0         0       25m

# 旧版本的RS会被保留，可以用来做更多的操作 
# RC 无法实现这些 比如修改镜像版本触发更新,或者是回滚

更新 ConfigMap

1
2
3
# 如果 deployment 中的pod 引用了1个 ConfigMap(或Secret) 
# 1. 更改ConfigMap资源本身不会触发升级操作
# 2. 如果需要修改应用程序的配置并且想触发更新 需要创建一个新的ConfigMap 并且修改Pod模板引用新的CM

回滚 Deployment

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
# 先创建一个错误版本 v3 
# 5个请求后 会返回500错误

[root@k8s-master2 v3]# cat app.js 
const http = require('http');
const os = require('os');

var requestCount = 0;

console.log("Kubia server starting...");

var handler = function(request, response) {
  console.log("Received request from " + request.connection.remoteAddress);
  if (++requestCount >= 5) {
    response.writeHead(500);
    response.end("Some internal error has occurred! This is pod " + os.hostname() + "\n");
    return;
  }
  response.writeHead(200);
  response.end("This is v3 running in pod " + os.hostname() + "\n");
};

var www = http.createServer(handler);
www.listen(8080);


[root@k8s-master2 v3]# docker build -t 172.31.228.68/game/kubia:v3 .
[root@k8s-master2 v3]# docker push 172.31.228.68/game/kubia:v3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
# 升级 然后查看错误请求
[root@k8s-master1 deployment]# kubectl set image deployment kubia nodejs=172.31.228.68/game/kubia:v3
deployment.apps/kubia image updated

[root@k8s-master1 deployment]# kubectl set image deployment kubia nodejs=172.31.228.68/game/kubia:v3
deployment.apps/kubia image updated

[root@k8s-master1 ~]# while true;do curl http://10.0.0.119;sleep 1; done
This is v3 running in pod kubia-698c869b67-z7qpw
This is v3 running in pod kubia-698c869b67-228w5
Some internal error has occurred! This is pod kubia-698c869b67-vkvch
Some internal error has occurred! This is pod kubia-698c869b67-z7qpw
Some internal error has occurred! This is pod kubia-698c869b67-228w5
Some internal error has occurred! This is pod kubia-698c869b67-vkvch
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
# 回滚升级 到之前的版本
[root@k8s-master1 deployment]# kubectl rollout undo deployment kubia
deployment.apps/kubia rolled back


...
This is v2 running in pod kubia-56fd945996-t28ch
This is v2 running in pod kubia-56fd945996-n7cwd
This is v2 running in pod kubia-56fd945996-jzzl4

# 显示Deployment的滚动升级历史
[root@k8s-master1 deployment]# kubectl rollout history deployment kubia
deployment.apps/kubia 
REVISION  CHANGE-CAUSE
1         kubectl create --filename=kubia-deployment-v1.yaml --record=true
3         kubectl create --filename=kubia-deployment-v1.yaml --record=true
4         kubectl create --filename=kubia-deployment-v1.yaml --record=true



# 版本记录大小 
[root@k8s-master1 deployment]# kubectl edit deploy kubia
revisionHistoryLimit: 10

# 根据版本号回滚
[root@k8s-master1 deployment]# kubectl rollout undo deployment kubia --to-revision=1
deployment.apps/kubia rolled back
[root@k8s-master1 deployment]# kubectl get rs
NAME                                DESIRED   CURRENT   READY   AGE
kubia-56fd945996                    0         0         0       3m26s
kubia-698c869b67                    0         0         0       2m47s
kubia-dbd9664c7                     3         3         3       8m52s

# RS中记录着版本号
[root@k8s-master1 deployment]# kubectl edit rs kubia-dbd9664c7

apiVersion: apps/v1
kind: ReplicaSet
metadata:
  annotations:
    deployment.kubernetes.io/desired-replicas: "3"
    deployment.kubernetes.io/max-replicas: "4"
    deployment.kubernetes.io/revision: "5"
    deployment.kubernetes.io/revision-history: "1"
    kubernetes.io/change-cause: kubectl create --filename=kubia-deployment-v1.yaml
      --record=true

控制滚动升级速率

1
2
3
4
5
6
7
8
9
[root@k8s-master1 deployment]# kubectl edit deploy kubia
  strategy:
    rollingUpdate:
      maxSurge: 25%
      maxUnavailable: 25%
    type: RollingUpdate

# maxSurge 最多超出副本数之外 设置成1或者2 最多超出1到2个Pod 
# maxUnavailable 滚动升级期间,有多少个Pod成为不可用状态

暂停滚动升级 使用试探性发布

1
2
3
4
# 假金丝雀发布
# 让少数请求到升级后的服务
# 之前的继续使用旧版本
# 有问题回滚到旧版本,没问题全部升级
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
# 先做一个最新版本
[root@k8s-master2 v4]# cat app.js 
const http = require('http');
const os = require('os');

console.log("Kubia server starting...");

var handler = function(request, response) {
  console.log("Received request from " + request.connection.remoteAddress);
  response.writeHead(200);
  response.end("This is v4 running in pod " + os.hostname() + "\n");
};

var www = http.createServer(handler);
www.listen(8080);

# 暂停滚动升级
[root@k8s-master1 deployment]# kubectl set image deployment kubia nodejs=172.31.228.68/game/kubia:v4
deployment.apps/kubia image updated

[root@k8s-master1 deployment]# kubectl rollout pause deployment kubia
deployment.apps/kubia paused

[root@k8s-master1 ~]# kubectl get pods -o wide
NAME                                      READY   STATUS    RESTARTS   AGE     IP            NODE          NOMINATED NODE   READINESS GATES
kubia-557944968b-579xn                    1/1     Running   0          71s     10.244.1.86   k8s-node2     <none>           <none>
kubia-557944968b-vr6s5                    1/1     Running   0          59s     10.244.2.73   k8s-master1   <none>           <none>
kubia-dbd9664c7-bhpf5                     1/1     Running   0          26m     10.244.0.71   k8s-node1     <none>           <none>
kubia-dbd9664c7-hc4kj                     1/1     Running   0          27m     10.244.1.85   k8s-node2     <none>           <none>

[root@k8s-master1 ~]# while true;do curl http://10.0.0.119;sleep 1; done
This is v4 running in pod kubia-557944968b-579xn
This is v1 running in pod kubia-dbd9664c7-bhpf5
This is v1 running in pod kubia-dbd9664c7-hc4kj
This is v4 running in pod kubia-557944968b-vr6s5
This is v4 running in pod kubia-557944968b-579xn

# 正常 恢复滚动升级
[root@k8s-master1 deployment]# kubectl rollout resume deployment kubia
deployment.apps/kubia resumed

# 如果想完全的对应pod数量 可以考虑使用两个不同的Deployment 同时数量一致,同时上线,再调整下线的Deployment

[root@k8s-master1 deployment]# vim kubia-deployment-v2.yaml 

apiVersion: apps/v1
kind: Deployment
metadata:
  name: kubia-v2
spec:
  replicas: 3
  selector:
    matchLabels:
      app: kubia
  template:
    metadata:
      name: kubia
      labels:
        app: kubia
    spec:
      containers:
      - image: 172.31.228.68/game/kubia:v2
        name: nodejs

[root@k8s-master1 deployment]# kubectl create -f kubia-deployment-v2.yaml 
deployment.apps/kubia-v2 created

[root@k8s-master1 deployment]# kubectl get ep
NAME         ENDPOINTS                                                        AGE
kubernetes   172.31.228.67:6443                                               13d
kubia        10.244.0.72:8080,10.244.0.73:8080,10.244.1.86:8080 + 3 more...   128m

# 由于svc匹配两个Pod的标签,所以被加入服务
This is v4 running in pod kubia-557944968b-6bw9r
This is v4 running in pod kubia-557944968b-vr6s5
This is v4 running in pod kubia-557944968b-579xn
This is v2 running in pod kubia-v2-56fd945996-t9zdb
This is v2 running in pod kubia-v2-56fd945996-76g6l
This is v2 running in pod kubia-v2-56fd945996-fgn2k

增加 就绪探针

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
# 先全部回滚到v2版本

[root@k8s-master1 deployment]# vim kubia-deployment-v3-with-readinesscheck.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: kubia
spec:
  replicas: 3
  selector:
    matchLabels:
      app: kubia
  minReadySeconds: 10
  strategy:
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 0
    type: RollingUpdate
  template:
    metadata:
      name: kubia
      labels:
        app: kubia
    spec:
      containers:
      - image: 172.31.228.68/game/kubia:v3
        name: nodejs
        readinessProbe:
          periodSeconds: 1
          httpGet:
            path: /
            port: 8080

[root@k8s-master1 deployment]# kubectl apply -f kubia-deployment-v3-with-readinesscheck.yaml 

This is v2 running in pod kubia-56fd945996-8zr5j
This is v2 running in pod kubia-56fd945996-6g4zl
This is v2 running in pod kubia-56fd945996-hfvk5
This is v2 running in pod kubia-56fd945996-8zr5j
This is v2 running in pod kubia-56fd945996-6g4zl

# 没有出现v3版本
[root@k8s-master1 deployment]# kubectl get pods -o wide
NAME                                      READY   STATUS    RESTARTS   AGE     IP            NODE          NOMINATED NODE   READINESS GATES
kubia-56fd945996-6g4zl                    1/1     Running   0          3m59s   10.244.0.75   k8s-node1     <none>           <none>
kubia-56fd945996-8zr5j                    1/1     Running   0          4m22s   10.244.2.75   k8s-master1   <none>           <none>
kubia-56fd945996-hfvk5                    1/1     Running   0          4m10s   10.244.1.89   k8s-node2     <none>           <none>
kubia-58c74659b7-mf2bv                    0/1     Running   0          69s     10.244.1.90   k8s-node2     <none>           <none>

# kubia-58c74659b7-mf2bv  为处于就绪状态
# 当新的Pod启动时候,就绪探针每隔1秒去发起请求 在上面的pod里 第5个请求会出现500返回,就绪探针会报错
# 因为 pod会从Svc的ep中被剔除 所以curl也请求不到新的Pod
[root@k8s-master1 deployment]# kubectl get ep
NAME         ENDPOINTS                                            AGE
kubernetes   172.31.228.67:6443                                   13d
kubia        10.244.0.75:8080,10.244.1.89:8080,10.244.2.75:8080   143m

# 滚动升级不继续下去,新的Pod一直处于不可用状态，就绪10秒后才真可用
# maxUnavailable: 0 使滚动更新正好卡在这里 除非之前的Pod准备就绪才行1

滚动升级时限

1
2
3
4
5
[root@k8s-master1 deployment]# kubectl edit deploy kubia
spec:
progressDeadlineSeconds: 600

# 滚动超时会自动取消

取消出错版本的滚动升级

1
2
3
# 直接回滚到上一版本
[root@k8s-master1 deployment]#  kubectl rollout undo deployment kubia
deployment.apps/kubia rolled back