docker - 为什么在我使用Kubernetes时docker重新下载相同的镜像

Question

我在没有Internet的隔离环境中使用kbuernetes和docker，我总是将镜像拉出并保存到其他计算机中的.tar文件中并加载到隔离环境中，但是有时kubernetes的Pod无法成功启动，并说Pod是拉镜像，但网络不正常。
但是我检查了docker的图像，该图像已加载，为什么我还需要再次拉相同的图像？

这是docker的镜像，已加载MYSQL的镜像:

[root@localhost kubecfg]# docker images
REPOSITORY                                                       TAG                 IMAGE ID            CREATED             SIZE
quay.io/kubernetes-ingress-controller/nginx-ingress-controller   0.12.0              4a9cd8a2008a        3 weeks ago         230.5 MB
docker.io/mysql                                                  latest              5195076672a7        3 weeks ago         371.4 MB
gcr.io/google_containers/kube-apiserver-amd64                    v1.9.2              7109112be2c7        11 weeks ago        210.4 MB

这是kubernetes的错误日志

[root@localhost kubecfg]# kubectl describe pod mysql-vmwdw
Name:           mysql-vmwdw
Namespace:      default
Node:           localhost.localdomain/192.168.88.129
Start Time:     Mon, 02 Apr 2018 14:14:07 +0800
Labels:         app=mysql
Annotations:    <none>
Status:         Running
IP:             192.168.0.61
Controlled By:  ReplicationController/mysql
Containers:
  mysql:
    Container ID:   docker://9aa3128eaa1f330dfd0d6ebf732dca5a99ad49d7d6d4002a2384bdb03e056d7d
    Image:          docker.io/mysql
    Image ID:       docker-pullable://docker.io/mysql@sha256:691c55aabb3c4e3b89b953dd2f022f7ea845e5443954767d321d5f5fa394e28c
    Port:           3306/TCP
    State:          Waiting
      Reason:       ImagePullBackOff
    Last State:     Terminated
      Reason:       Error
      Exit Code:    1
      Started:      Tue, 10 Apr 2018 14:56:04 +0800
      Finished:     Wed, 11 Apr 2018 08:56:04 +0800
    Ready:          False
    Restart Count:  3
    Environment:
      MYSQL_ROOT_PASSWORD:  123456
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-s7kq2 (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          False 
  PodScheduled   True 
Volumes:
  default-token-s7kq2:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-s7kq2
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                  Age                 From                            Message
  ----     ------                  ----                ----                            -------
  Normal   SuccessfulMountVolume   19m                 kubelet, localhost.localdomain  MountVolume.SetUp succeeded for volume "default-token-s7kq2"
  Normal   SandboxChanged          19m (x12 over 19m)  kubelet, localhost.localdomain  Pod sandbox changed, it will be killed and re-created.
  Warning  FailedCreatePodSandBox  19m (x12 over 19m)  kubelet, localhost.localdomain  Failed create pod sandbox.
  Warning  Failed                  9m (x6 over 18m)    kubelet, localhost.localdomain  Failed to pull image "docker.io/mysql": rpc error: code = Unknown desc = Network timed out while trying to connect to https://index.docker.io/v1/repositories/library/mysql/images. You may want to check your internet connection or if you are behind a proxy.
  Warning  Failed                  4m (x41 over 18m)   kubelet, localhost.localdomain  Error: ImagePullBackOff

它说: “无法提取图像” docker.io/mysql“:rpc错误:代码=未知desc =尝试连接到https://index.docker.io/v1/repositories/library/mysql/images时网络超时。您可能要检查Internet连接或是否在代理后面。”

我已经加载了图像，为什么 docker 会再次拉相同的图像？我该怎么做才能解决问题？

Answer 1

您正在使用:latest标签作为图像。当图像标签为latest时，kubernetes会将imagePullPolicy设置为Always。有关详细信息，请see the official doc。

您可以将标签更改为其他名称(例如docker.io/mysql:8.0)，或者只需将自己的imagePullPolicy指定为Never即可。