晒晒狗 vs 练练手

0%

K8s学习笔记——作业副本与水平扩展

发表于 分类于 Disqus：
本文字数： 15k 阅读时长 ≈ 14 分钟

学习极客时间上的《深入剖析Kubernetes》

秉持眼过千遍不如手过一遍的原则。动手实践并记录结果

对应章节：17 | 经典PaaS的记忆：作业副本与水平扩展

ReplicaSet

书接上文中，创建了一个nginx-dp的deployment

1
2
3
4
$ kubectl get pods
NAME                        READY   STATUS    RESTARTS   AGE
nginx-dp-67f857c57f-4wptz   1/1     Running   0          7m34s
nginx-dp-67f857c57f-qx7l7   1/1     Running   0          25m

先来查看下ReplicaSet

1
2
3
$ kubectl get replicasets
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-67f857c57f   2         2         2       39m

由此可见，实际上当我们创建一个deployment的时候，k8s为我们创建了一个名为nginx-dp-67f857c57f的ReplicaSet。是不是觉得这个名字中的ID这么熟悉。正是Pod的labels字段中的那个ID

创建一个ReplicaSet

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: test-rs
  labels:
    app: nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: Never
1
2
3
4
5
6
7
8
9
10
11
12
13
$ kubectl apply -f rs-test.yaml
replicaset.apps/rs-test created
$ kubectl get pods
NAME                        READY   STATUS        RESTARTS   AGE
nginx-dp-67f857c57f-4wptz   1/1     Running       0          33m
nginx-dp-67f857c57f-qx7l7   1/1     Running       0          51m
rs-test-8zd6q               0/1     Terminating   0          4s
rs-test-f5vvr               0/1     Terminating   0          4s
rs-test-rkh7r               0/1     Terminating   0          4s
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-67f857c57f   2         2         2       51m
rs-test               0         0         0       14s

WHAT?!

居然创建失败，所谓在作死的道路上不停试探。谁让我用了和上节课中的Deployment一样的labels呢

再来一遍：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: test-rs
  labels:
    app: nginx-new
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx-new
  template:
    metadata:
      labels:
        app: nginx-new
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: Never
1
2
3
4
5
6
7
8
9
10
11
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-67f857c57f   2         2         2       54m
test-rs               3         3         3       7m30s
$ kubectl get pods
NAME                        READY   STATUS    RESTARTS   AGE
nginx-dp-67f857c57f-4wptz   1/1     Running   0          37m
nginx-dp-67f857c57f-qx7l7   1/1     Running   0          54m
test-rs-9wfx8               1/1     Running   0          7m33s
test-rs-qbckm               1/1     Running   0          7m33s
test-rs-rxjbv               1/1     Running   0          7m33s

现在看起来，创建出了一个名为test-rs的ReplicaSet，而test-rs的rs又创建出了三个pod

水平收缩

修改test-rs的RS：replicas: 3为replicas: 4

1
2
3
4
5
6
7
8
$ kubectl apply -f rs-test.yaml
replicaset.apps/test-rs configured
$ kubectl get pods
NAME                        READY   STATUS    RESTARTS   AGE
test-rs-9wfx8               1/1     Running   0          37m
test-rs-dwm9j               1/1     Running   0          4s
test-rs-qbckm               1/1     Running   0          37m
test-rs-rxjbv               1/1     Running   0          37m

修改为4个后，自动创建了一个新的test-rs-dwm9j的pod

再次修改为replicas: 2

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
$ kubectl get pods
NAME                        READY   STATUS    RESTARTS   AGE
test-rs-9wfx8               1/1     Running   0          41m
test-rs-rxjbv               1/1     Running   0          41m
$ kubectl describe rs test-rs
Name:         test-rs
Namespace:    default
Selector:     app=nginx-new
Labels:       app=nginx-new
Annotations:  Replicas:  2 current / 2 desired
Pods Status:  2 Running / 0 Waiting / 0 Succeeded / 0 Failed
...
Events:
  Type    Reason            Age    From                   Message
  ----    ------            ----   ----                   -------
  Normal  SuccessfulCreate  41m    replicaset-controller  Created pod: test-rs-qbckm
  Normal  SuccessfulCreate  41m    replicaset-controller  Created pod: test-rs-9wfx8
  Normal  SuccessfulCreate  41m    replicaset-controller  Created pod: test-rs-rxjbv
  Normal  SuccessfulCreate  4m41s  replicaset-controller  Created pod: test-rs-dwm9j
  Normal  SuccessfulDelete  16s    replicaset-controller  Deleted pod: test-rs-dwm9j
  Normal  SuccessfulDelete  16s    replicaset-controller  Deleted pod: test-rs-qbckm

查看ReplicaSet的Events可以看出来：

  1. 修改为4的时候，创建了一个新的pod
  2. 修改为2的时候，两个pod被删除

滚动更新

创建Deployment

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-dp
  labels:
    app: nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        ports:
        - containerPort: 80
1
2
3
4
5
6
7
8
9
10
11
12
13
14
$ kubectl apply -f nginx-dp.yaml --record
deployment.apps/nginx-dp created
# 注：执行完立即执行rollout才能看到
$ kubectl rollout status deployment/nginx-dp
Waiting for deployment "nginx-dp" rollout to finish: 0 of 3 updated replicas are available...
Waiting for deployment "nginx-dp" rollout to finish: 1 of 3 updated replicas are available...
Waiting for deployment "nginx-dp" rollout to finish: 2 of 3 updated replicas are available...
deployment "nginx-dp" successfully rolled out
$ kubectl get rs
NAME                 DESIRED   CURRENT   READY   AGE
nginx-dp-d46f5678b   3         3         3       3m41s
$ kubectl get deploy
NAME       READY   UP-TO-DATE   AVAILABLE   AGE
nginx-dp   3/3     3            3           3m45s

可以看到:

  1. ReplicaSet的状态有：Desired, Current, Ready
  2. Deployment的有： Ready, Available, UP-to-date

注：与课程中的不一样

修改image版本

1
2
3
4
5
6
7
8
9
10
$ kubectl edit deployment/nginx-dp
....
    spec:
      containers:
      - image: nginx:stable
        imagePullPolicy: Always
        name: nginx
        ports:
        - containerPort: 80
          protocol: TCP
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
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-7fb9ff5685   1         1         0       41s
nginx-dp-d46f5678b    3         3         3       8m2s
$ kubectl rollout status deployment/nginx-dp
Waiting for deployment "nginx-dp" rollout to finish: 1 out of 3 new replicas have been updated...
Waiting for deployment "nginx-dp" rollout to finish: 1 out of 3 new replicas have been updated...
Waiting for deployment "nginx-dp" rollout to finish: 1 out of 3 new replicas have been updated...
Waiting for deployment "nginx-dp" rollout to finish: 2 out of 3 new replicas have been updated...
Waiting for deployment "nginx-dp" rollout to finish: 2 out of 3 new replicas have been updated...
Waiting for deployment "nginx-dp" rollout to finish: 2 out of 3 new replicas have been updated...
Waiting for deployment "nginx-dp" rollout to finish: 1 old replicas are pending termination...
Waiting for deployment "nginx-dp" rollout to finish: 1 old replicas are pending termination...
deployment "nginx-dp" successfully rolled out
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-7fb9ff5685   3         3         3       10m
nginx-dp-d46f5678b    0         0         0       17m
$ kubectl describe deploy nginx-dp
...
StrategyType:           RollingUpdate
MinReadySeconds:        0
RollingUpdateStrategy:  25% max unavailable, 25% max surge
...
OldReplicaSets:  <none>
NewReplicaSet:   nginx-dp-7fb9ff5685 (3/3 replicas created)
Events:
  Type    Reason             Age    From                   Message
  ----    ------             ----   ----                   -------
  Normal  ScalingReplicaSet  18m    deployment-controller  Scaled up replica set nginx-dp-d46f5678b to 3
  Normal  ScalingReplicaSet  11m    deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 1
  Normal  ScalingReplicaSet  7m22s  deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 2
  Normal  ScalingReplicaSet  7m22s  deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 2
  Normal  ScalingReplicaSet  3m34s  deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 1
  Normal  ScalingReplicaSet  3m34s  deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 3
  Normal  ScalingReplicaSet  3m17s  deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 0

从上面的各种结果可以看出，当修改了deployment中的image后：

  1. k8s创建了一个新的ReplicaSet: nginx-dp-7fb9ff5685
  2. 逐个增加新的ReplicaSet，然后逐个减少旧的ReplicaSet
  3. 当滚动更新完毕后，旧的RS nginx-dp-d46f5678b数量变为0，但仍旧保留，没有自动删除
  4. Deployment的NewReplicaSet字段变成了nginx-dp-7fb9ff5685

修改为一个错误的image

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
$ kubectl set image deployment/nginx-dp nginx=nginx:1.91
deployment.apps/nginx-dp image updated
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   1         1         0       6s
nginx-dp-7fb9ff5685   3         3         3       24m
nginx-dp-d46f5678b    0         0         0       31m
$ kubectl get pods
NAME                        READY   STATUS         RESTARTS   AGE
nginx-dp-6d6678fb55-4xxpv   0/1     ErrImagePull   0          72s
nginx-dp-7fb9ff5685-25xlv   1/1     Running        0          25m
nginx-dp-7fb9ff5685-jxvt8   1/1     Running        0          17m
nginx-dp-7fb9ff5685-ld64t   1/1     Running        0          21m
$ kubectl describe deployment nginx-dp
...
OldReplicaSets:  nginx-dp-7fb9ff5685 (3/3 replicas created)
NewReplicaSet:   nginx-dp-6d6678fb55 (1/1 replicas created)
Events:
  Type    Reason             Age   From                   Message
  ----    ------             ----  ----                   -------
  Normal  ScalingReplicaSet  33m   deployment-controller  Scaled up replica set nginx-dp-d46f5678b to 3
  Normal  ScalingReplicaSet  25m   deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 1
  Normal  ScalingReplicaSet  21m   deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 2
  Normal  ScalingReplicaSet  21m   deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 2
  Normal  ScalingReplicaSet  18m   deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 1
  Normal  ScalingReplicaSet  18m   deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 3
  Normal  ScalingReplicaSet  17m   deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 0
  Normal  ScalingReplicaSet  98s   deployment-controller  Scaled up replica set nginx-dp-6d6678fb55 to 1

可以看出：

  1. 再次创建了一个新的RS: nginx-dp-6d6678fb55
  2. 创建了一个名为nginx-dp-6d6678fb55-4xxpv的pod，但状态为ErrImagePull
  3. Deployment的OldReplicaSets为原来的nginx-dp-7fb9ff5685，NewReplicaSet为新的nginx-dp-6d6678fb55

Rollout

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
$ kubectl rollout undo deployment/nginx-dp
deployment.apps/nginx-dp rolled back
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       6m49s
nginx-dp-7fb9ff5685   3         3         3       30m
nginx-dp-d46f5678b    0         0         0       38m
$ kubectl describe deployment/nginx-dp
...
OldReplicaSets:  <none>
NewReplicaSet:   nginx-dp-7fb9ff5685 (3/3 replicas created)
Events:
  Type    Reason             Age   From                   Message
  ----    ------             ----  ----                   -------
  Normal  ScalingReplicaSet  38m   deployment-controller  Scaled up replica set nginx-dp-d46f5678b to 3
  Normal  ScalingReplicaSet  31m   deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 1
  Normal  ScalingReplicaSet  27m   deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 2
  Normal  ScalingReplicaSet  27m   deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 2
  Normal  ScalingReplicaSet  23m   deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 1
  Normal  ScalingReplicaSet  23m   deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 3
  Normal  ScalingReplicaSet  23m   deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 0
  Normal  ScalingReplicaSet  7m2s  deployment-controller  Scaled up replica set nginx-dp-6d6678fb55 to 1
  Normal  ScalingReplicaSet  19s   deployment-controller  Scaled down replica set nginx-dp-6d6678fb55 to 0
$ kubectl get pods
NAME                        READY   STATUS    RESTARTS   AGE
nginx-dp-7fb9ff5685-25xlv   1/1     Running   0          40m
nginx-dp-7fb9ff5685-jxvt8   1/1     Running   0          33m
nginx-dp-7fb9ff5685-ld64t   1/1     Running   0          36m

可以看出：

  1. nginx-dp-6d6678fb55的RS的期望值变为了0
  2. 对应的pod被自动删除

history

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
$ kubectl rollout history deployment/nginx-dp
deployment.apps/nginx-dp
REVISION  CHANGE-CAUSE
1         kubectl apply --filename=nginx-dp.yaml --record=true
3         kubectl apply --filename=nginx-dp.yaml --record=true
4         kubectl apply --filename=nginx-dp.yaml --record=true

$ kubectl rollout history deployment/nginx-dp --revision=4
deployment.apps/nginx-dp with revision #4
Pod Template:
  Labels:   app=nginx
    pod-template-hash=7fb9ff5685
  Annotations:  kubernetes.io/change-cause: kubectl apply --filename=nginx-dp.yaml --record=true
  Containers:
   nginx:
    Image:  nginx:stable
    Port:   80/TCP
    Host Port:  0/TCP
    Environment:    <none>
    Mounts: <none>
  Volumes:  <none>

注：新版本的命令输出已经和课程中的不一样了

现在可以使用revision字段回滚到镜像为nginx:latest

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
$ kubectl rollout history deployment/nginx-dp --revision=1
deployment.apps/nginx-dp with revision #1
Pod Template:
  Labels:   app=nginx
    pod-template-hash=d46f5678b
  Annotations:  kubernetes.io/change-cause: kubectl apply --filename=nginx-dp.yaml --record=true
  Containers:
   nginx:
    Image:  nginx
    Port:   80/TCP
    Host Port:  0/TCP
    Environment:    <none>
    Mounts: <none>
  Volumes:  <none>
$ kubectl rollout undo deployment/nginx-dp --to-revision=1
deployment.apps/nginx-dp rolled back
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       29m
nginx-dp-7fb9ff5685   3         3         3       53m
nginx-dp-d46f5678b    1         1         0       60m
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       29m
nginx-dp-7fb9ff5685   1         1         1       53m
nginx-dp-d46f5678b    3         3         2       60m
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       29m
nginx-dp-7fb9ff5685   0         0         0       53m
nginx-dp-d46f5678b    3         3         3       60m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
$ kubectl rollout history deployment/nginx-dp
deployment.apps/nginx-dp
REVISION  CHANGE-CAUSE
3         kubectl apply --filename=nginx-dp.yaml --record=true
4         kubectl apply --filename=nginx-dp.yaml --record=true
5         kubectl apply --filename=nginx-dp.yaml --record=true

$ kubectl rollout history deployment/nginx-dp --revision=5
deployment.apps/nginx-dp with revision #5
Pod Template:
  Labels:   app=nginx
    pod-template-hash=d46f5678b
  Annotations:  kubernetes.io/change-cause: kubectl apply --filename=nginx-dp.yaml --record=true
  Containers:
   nginx:
    Image:  nginx
    Port:   80/TCP
    Host Port:  0/TCP
    Environment:    <none>
    Mounts: <none>
  Volumes:  <none>

可以看到，回滚之后：

  1. 原来的revision 1已经不存在了
  2. 出现了一个revision = 5
  3. revision 5和原来的revision 1的内容一致

所以，虽然是回滚到某个版本，但revision的号依然是增加了的。是通过滚动升级的方式实现了回滚的操作

暂停与恢复

  • 现在，先暂停
1
2
$ kubectl rollout pause deployment/nginx-dp
deployment.apps/nginx-dp paused
  • 修改image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
24
25
26
27
$ kubectl set image deployment/nginx-dp nginx=nginx:1.18
deployment.apps/nginx-dp image updated
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       37m
nginx-dp-7fb9ff5685   0         0         0       61m
nginx-dp-d46f5678b    3         3         3       68m
$ kubectl describe deployment/nginx-dp
...
OldReplicaSets:  nginx-dp-d46f5678b (3/3 replicas created)
NewReplicaSet:   <none>
Events:
  Type    Reason             Age                  From                   Message
  ----    ------             ----                 ----                   -------
  Normal  ScalingReplicaSet  57m                  deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 2
  Normal  ScalingReplicaSet  57m                  deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 2
  Normal  ScalingReplicaSet  54m                  deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 1
  Normal  ScalingReplicaSet  54m                  deployment-controller  Scaled up replica set nginx-dp-7fb9ff5685 to 3
  Normal  ScalingReplicaSet  53m                  deployment-controller  Scaled down replica set nginx-dp-d46f5678b to 0
  Normal  ScalingReplicaSet  37m                  deployment-controller  Scaled up replica set nginx-dp-6d6678fb55 to 1
  Normal  ScalingReplicaSet  30m                  deployment-controller  Scaled down replica set nginx-dp-6d6678fb55 to 0
  Normal  ScalingReplicaSet  8m32s                deployment-controller  Scaled up replica set nginx-dp-d46f5678b to 1
  Normal  ScalingReplicaSet  8m26s                deployment-controller  Scaled up replica set nginx-dp-d46f5678b to 2
  Normal  ScalingReplicaSet  8m26s                deployment-controller  Scaled down replica set nginx-dp-7fb9ff5685 to 2
  Normal  ScalingReplicaSet  8m22s (x2 over 69m)  deployment-controller  Scaled up replica set nginx-dp-d46f5678b to 3
  Normal  ScalingReplicaSet  8m22s                deployment-controller  Scaled down replica set nginx-dp-7fb9ff5685 to 1
  Normal  ScalingReplicaSet  8m17s                deployment-controller  Scaled down replica set nginx-dp-7fb9ff5685 to 0

可以看到：

  1. 修改了image为nginx:1.18后，没有创建出新的RS
  2. 但：Deployment的OldReplicaSet变成了nginx-dp-d46f5678b，而NewReplicaSet变成了<none>，这与最早之前看到的不一样
  • 如果在暂停状态下将image再设置回nginx:latest会怎样呢
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
$ kubectl set image deployment/nginx-dp nginx=nginx
deployment.apps/nginx-dp image updated
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       44m
nginx-dp-7fb9ff5685   0         0         0       68m
nginx-dp-d46f5678b    3         3         3       76m
# 取消暂停
$ kubectl rollout resume deployment/nginx-dp
deployment.apps/nginx-dp resumed
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       46m
nginx-dp-7fb9ff5685   0         0         0       70m
nginx-dp-d46f5678b    3         3         3       77m
$ kubectl rollout history deployment/nginx-dp
deployment.apps/nginx-dp
REVISION  CHANGE-CAUSE
3         kubectl apply --filename=nginx-dp.yaml --record=true
4         kubectl apply --filename=nginx-dp.yaml --record=true
5         kubectl apply --filename=nginx-dp.yaml --record=true
$ kubectl describe deployment/nginx-dp
...
OldReplicaSets:  <none>
NewReplicaSet:   nginx-dp-d46f5678b (3/3 replicas created)
...

可以看到，修改回暂停前的镜像后，没有生成新的revision

  • 继续暂停的实验
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
# 暂停
$ kubectl rollout pause deployment/nginx-dp
deployment.apps/nginx-dp paused
# 修改
$ kubectl set image deployment/nginx-dp nginx=nginx:1.18
deployment.apps/nginx-dp image updated
# 恢复
$ kubectl rollout resume deployment nginx-dp
deployment.apps/nginx-dp resumed
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       53m
nginx-dp-7f6cd547bd   1         1         0       3s
nginx-dp-7fb9ff5685   0         0         0       77m
nginx-dp-d46f5678b    3         3         3       85m
$ kubectl rollout history deployment/nginx-dp
deployment.apps/nginx-dp
REVISION  CHANGE-CAUSE
3         kubectl apply --filename=nginx-dp.yaml --record=true
4         kubectl apply --filename=nginx-dp.yaml --record=true
5         kubectl apply --filename=nginx-dp.yaml --record=true
6         kubectl apply --filename=nginx-dp.yaml --record=true
$ kubectl get rs
NAME                  DESIRED   CURRENT   READY   AGE
nginx-dp-6d6678fb55   0         0         0       55m
nginx-dp-7f6cd547bd   3         3         3       90s
nginx-dp-7fb9ff5685   0         0         0       79m
nginx-dp-d46f5678b    0         0         0       86m
$ kubectl rollout history deployment/nginx-dp --revision=6
deployment.apps/nginx-dp with revision #6
Pod Template:
  Labels:   app=nginx
    pod-template-hash=7f6cd547bd
  Annotations:  kubernetes.io/change-cause: kubectl apply --filename=nginx-dp.yaml --record=true
  Containers:
   nginx:
    Image:  nginx:1.18
    Port:   80/TCP
    Host Port:  0/TCP
    Environment:    <none>
    Mounts: <none>
  Volumes:  <none>

可以看出，当取消暂停后，对应的动作按照滚动更新的方式开始执行。

小结

本节课程主要学习了：

  1. 水平扩展时，实际上是修改了ReplicaSet的期望值，并按照期望值去逐步拉起pod
  2. 而滚动更新时，会创建新的ReplicaSet，并通过逐个增加新的，而逐个减少旧的的方式来逐步替换pod

最后，一个对原文的引用：

Deployment 实际上是一个两层控制器。首先，它通过 ReplicaSet 的个数来描述应用的版本；然后，它再通过 ReplicaSet 的属性（比如 replicas 的值），来保证 Pod 的副本数量。

相关文章