某些情况下，您可能想要回滚（rollback）Deployment，例如：Deployment 不稳定（可能是不断地崩溃）。默认情况下，kubernetes 将保存 Deployment 的所有更新（rollout）历史。您可以设定 revision history limit 来确定保存的历史版本数量。

TIP 当且仅当 Deployment 的 .spec.template 字段被修改时（例如，您修改了容器的镜像），kubernetes 将为其创建一个 Deployment revision（版本）。Deployment 的其他更新（例如：修改 .spec.replicas 字段）将不会创建新的 Deployment reviesion（版本）。

本文提供了两种途径对 Deployment 执行回滚操作：

  • 使用 kubectl 回滚 Deployment
  • 使用 Kuboard 回滚 Deployment

使用kubectl回滚Deployment

模拟更新错误

  • 假设您在更新 Deployment 的时候，犯了一个拼写错误，将 nginx:1.9.1 写成了 nginx:1.91 ```bash kubectl set image deployment.v1.apps/nginx-deployment nginx=nginx:1.91 —record=true
  1. <br />    输出结果如下所示：
  2. ```bash
  3. deployment.apps/nginx-deployment image updated
  • 该更新将卡住，您可以执行命令 kubectl rollout status deployment.v1.apps/nginx-deployment 检查其状态，输出结果如下所示： ```bash Waiting for rollout to finish: 1 out of 3 new replicas have been updated…
  2. - <br />
  3. - 执行命令 kubectl get rs 您将看到两个旧的 ReplicaSet（nginx-deployment-1564180365 and nginx-deployment-2035384211）和一个新的 ReplicaSet (nginx-deployment-3066724191) 。如下所示：
  4. ```bash
  5. NAME                          DESIRED   CURRENT   READY   AGE
  6. nginx-deployment-1564180365   3         3         3       25s
  7. nginx-deployment-2035384211   0         0         0       36s
  8. nginx-deployment-3066724191   1         1         0       6s

  • 执行命令 kubectl get pods，您将看到 1 个由新 ReplicaSet 创建的 Pod 卡在抓取 image 的死循环里： ```bash NAME READY STATUS RESTARTS AGE nginx-deployment-1564180365-70iae 1/1 Running 0 25s nginx-deployment-1564180365-jbqqo 1/1 Running 0 25s nginx-deployment-1564180365-hysrc 1/1 Running 0 25s nginx-deployment-3066724191-08mng 0/1 ImagePullBackOff 0 6s
  1. > <br />**TIP**Deployment Controller 将自动停止有问题的更新（rollout），不会继续 scale up 新的 ReplicaSet。maxUnavailable 参数指定了最多会有几个 Pod 副本卡住，该参数的默认值是 25%。
  3. - 执行命令 kubectl describe deployment 查看 Deployment 的详情，输出结果如下所示：
  4. ```bash
  5. Name:           nginx-deployment
  6. Namespace:      default
  7. CreationTimestamp:  Tue, 15 Mar 2016 14:48:04 -0700
  8. Labels:         app=nginx
  9. Selector:       app=nginx
  10. Replicas:       3 desired | 1 updated | 4 total | 3 available | 1 unavailable
  11. StrategyType:       RollingUpdate
  12. MinReadySeconds:    0
  13. RollingUpdateStrategy:  25% max unavailable, 25% max surge
  14. Pod Template:
  15.   Labels:  app=nginx
  16.   Containers:
  17.   nginx:
  18.     Image:        nginx:1.91
  19.     Port:         80/TCP
  20.     Host Port:    0/TCP
  21.     Environment:  <none>
  22.     Mounts:       <none>
  23.   Volumes:        <none>
  24. Conditions:
  25.   Type           Status  Reason
  26.   ----           ------  ------
  27.   Available      True    MinimumReplicasAvailable
  28.   Progressing    True    ReplicaSetUpdated
  29. OldReplicaSets:     nginx-deployment-1564180365 (3/3 replicas created)
  30. NewReplicaSet:      nginx-deployment-3066724191 (1/1 replicas created)
  31. Events:
  32.   FirstSeen LastSeen    Count   From                    SubobjectPath   Type        Reason              Message
  33.   --------- --------    -----   ----                    -------------   --------    ------              -------
  34.   1m        1m          1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled up replica set nginx-deployment-2035384211 to 3
  35.   22s       22s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled up replica set nginx-deployment-1564180365 to 1
  36.   22s       22s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled down replica set nginx-deployment-2035384211 to 2
  37.   22s       22s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled up replica set nginx-deployment-1564180365 to 2
  38.   21s       21s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled down replica set nginx-deployment-2035384211 to 1
  39.   21s       21s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled up replica set nginx-deployment-1564180365 to 3
  40.   13s       13s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled down replica set nginx-deployment-2035384211 to 0
  41.   13s       13s         1       {deployment-controller }                Normal      ScalingReplicaSet   Scaled up replica set nginx-deployment-3066724191 to 1


为了修复这个问题，您需要将其回滚到之前的一个稳定版本。

检查 Deployment 的更新历史

  1. 执行命令 kubectl rollout history deployment.v1.apps/nginx-deployment 检查 Deployment 的历史版本，输出结果如下所示： ```bash deployments “nginx-deployment” REVISION CHANGE-CAUSE 1 kubectl apply —filename=https://k8s.io/examples/controllers/nginx-deployment.yaml —record=true 2 kubectl set image deployment.v1.apps/nginx-deployment nginx=nginx:1.9.1 —record=true 3 kubectl set image deployment.v1.apps/nginx-deployment nginx=nginx:1.91 —record=true
  1. <br />**CHANGE-CAUSE** 是该 revision（版本）创建时从 Deployment 的 annotation kubernetes.io/change-cause 拷贝而来。
  3. 您可以通过如下方式制定 **CHANGE-CAUSE** 信息：
  5.    - 为 Deployment 增加注解，kubectl annotate deployment.v1.apps/nginx-deployment kubernetes.io/change-cause="image updated to 1.9.1"
  6.    - 执行 kubectl apply 命令时，增加 --record 选项
  7.    - 手动编辑 Deployment 的 .metadata.annotation 信息
  9. 2. 执行命令 kubectl rollout history deployment.v1.apps/nginx-deployment --revision=2，查看 revision（版本）的详细信息，输出结果如下所示：
  10. ```yaml
  11. deployments "nginx-deployment" revision 2
  12.   Labels:       app=nginx
  13.           pod-template-hash=1159050644
  14.   Annotations:  kubernetes.io/change-cause=kubectl set image deployment.v1.apps/nginx-deployment nginx=nginx:1.9.1 --record=true
  15.   Containers:
  16.   nginx:
  17.     Image:      nginx:1.9.1
  18.     Port:       80/TCP
  19.     QoS Tier:
  20.         cpu:      BestEffort
  21.         memory:   BestEffort
  22.     Environment Variables:      <none>
  23.   No volumes.

回滚到前一个 revision（版本）

下面的步骤可将 Deployment 从当前版本回滚到前一个版本（version 2）

  1. 执行命令 kubectl rollout undo deployment.v1.apps/nginx-deployment 将当前版本回滚到前一个版本，输出结果如下所示： ```bash deployment.apps/nginx-deployment
  1. <br />    或者，您也可以使用 --to-revision 选项回滚到前面的某一个指定版本，执行如下命令：
  2. ```bash
  3. kubectl rollout undo deployment.v1.apps/nginx-deployment --to-revision=2


输出结果如下：

  1. deployment.apps/nginx-deployment


更详细的信息，请参考 kubectl rollout(opens new window)此时，Deployment 已经被回滚到前一个稳定版本。您可以看到 Deployment Controller 为该 Deployment 产生了 DeploymentRollback event。

  1. 执行命令 kubectl get deployment nginx-deployment，检查该回滚是否成功，Deployment 是否按预期的运行，输出结果如下： ```bash NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE nginx-deployment 3 3 3 3 30m
  2. 3. 执行命令 kubectl describe deployment nginx-deployment 查看 Deployment 的详情，输出结果如下：
  3. ```bash
  4. Name:                   nginx-deployment
  5. Namespace:              default
  6. CreationTimestamp:      Sun, 02 Sep 2018 18:17:55 -0500
  7. Labels:                 app=nginx
  8. Annotations:            deployment.kubernetes.io/revision=4
  9.                         kubernetes.io/change-cause=kubectl set image deployment.v1.apps/nginx-deployment nginx=nginx:1.9.1 --record=true
  10. Selector:               app=nginx
  11. Replicas:               3 desired | 3 updated | 3 total | 3 available | 0 unavailable
  12. StrategyType:           RollingUpdate
  13. MinReadySeconds:        0
  14. RollingUpdateStrategy:  25% max unavailable, 25% max surge
  15. Pod Template:
  16.   Labels:  app=nginx
  17.   Containers:
  18.   nginx:
  19.     Image:        nginx:1.9.1
  20.     Port:         80/TCP
  21.     Host Port:    0/TCP
  22.     Environment:  <none>
  23.     Mounts:       <none>
  24.   Volumes:        <none>
  25. Conditions:
  26.   Type           Status  Reason
  27.   ----           ------  ------
  28.   Available      True    MinimumReplicasAvailable
  29.   Progressing    True    NewReplicaSetAvailable
  30. OldReplicaSets:  <none>
  31. NewReplicaSet:   nginx-deployment-c4747d96c (3/3 replicas created)
  32. Events:
  33.   Type    Reason              Age   From                   Message
  34.   ----    ------              ----  ----                   -------
  35.   Normal  ScalingReplicaSet   12m   deployment-controller  Scaled up replica set nginx-deployment-75675f5897 to 3
  36.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled up replica set nginx-deployment-c4747d96c to 1
  37.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled down replica set nginx-deployment-75675f5897 to 2
  38.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled up replica set nginx-deployment-c4747d96c to 2
  39.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled down replica set nginx-deployment-75675f5897 to 1
  40.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled up replica set nginx-deployment-c4747d96c to 3
  41.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled down replica set nginx-deployment-75675f5897 to 0
  42.   Normal  ScalingReplicaSet   11m   deployment-controller  Scaled up replica set nginx-deployment-595696685f to 1
  43.   Normal  DeploymentRollback  15s   deployment-controller  Rolled back deployment "nginx-deployment" to revision 2
  44.   Normal  ScalingReplicaSet   15s   deployment-controller  Scaled down replica set nginx-deployment-595696685f to 0

使用Kuboard回滚Deployment

模拟更新错误

  • 假设您更新 Deployment 的时候，犯了一个拼写错误，将 1.9.1 写成了 1.91，如下图所示：image.png

  • 该更新将卡住，新的副本集中的 Pod 将因为抓取不到镜像而不能启动，并陷入不断抓取镜像的死循环当中，如下图所示：image.png

    检查 Deployment 的更新历史，并回滚

  • 上图中，显示了该 Deployment 有三个副本集，分别对应 Deployment 的版本 1、2、3。点击某一个副本集上的 回滚到 按钮，Kuboard 将为您对比两个版本之间的差异，如下图所示：image.png

  • 点击 确认 按钮，将回滚到选定的版本