今天忽然发件worker不正常：

[root@node-1-bak mycluster]# kubectl get nodes
NAME         STATUS     ROLES    AGE   VERSION
node-1-bak   Ready      master   24h   v1.19.7
node-2-bak   Ready      master   24h   v1.19.7
node-3-bak   NotReady   <none>   24h   v1.19.7

查看节点描述和所有pod状态都是正常运行中，后面直接去node-3-bak查看kubelet的状态，kubelet是正常运行的

Sep 06 01:21:18 node-3-bak kubelet[10631]: E0906 01:21:18.048708   10631 kubelet.go:2186] node "node-3-bak" not found

Sep 06 01:21:18 node-3-bak kubelet[10631]: I0906 01:21:18.102505   10631 csi_plugin.go:994] Failed to contact API server when waiting for CSINode publishing: Get "https://l...tion refused

Sep 06 01:38:47 node-3-bak kubelet[10631]: E0906 01:38:47.081227   10631 kubelet.go:2186] node "node-3-bak" not found
Sep 06 01:38:47 node-3-bak kubelet[10631]: I0906 01:38:47.101517   10631 csi_plugin.go:994] Failed to contact API server when waiting for CSINode publishing: Get "https://localhost:6443/apis/storage.k8s.io/v1/csinodes/node-3-bak": dial tcp 127.0.0.1:6443: connect: connection refused
..........
..........
Sep 06 01:38:48 node-3-bak kubelet[10631]: E0906 01:38:48.503435   10631 remote_runtime.go:113] RunPodSandbox from runtime service failed: rpc error: code = Unknown desc = failed to get sandbox image "k8s.gcr.io/pause:3.3": failed to pull image "k8s.gcr.io/pause:3.3": failed to pull and unpack image "k8s.gcr.io/pause:3.3": failed to resolve reference "k8s.gcr.io/pause:3.3": failed to do request: Head "https://k8s.gcr.io/v2/pause/manifests/3.3": dial tcp 74.125.204.82:443: i/o timeout
Sep 06 01:38:48 node-3-bak kubelet[10631]: E0906 01:38:48.503495   10631 kuberuntime_sandbox.go:69] CreatePodSandbox for pod "nginx-proxy-node-3-bak_kube-system(4068a5c786f6c77faa6538742baa1d44)" failed: rpc error: code = Unknown desc = failed to get sandbox image "k8s.gcr.io/pause:3.3": failed to pull image "k8s.gcr.io/pause:3.3": failed to pull and unpack image "k8s.gcr.io/pause:3.3": failed to resolve reference "k8s.gcr.io/pause:3.3": failed to do request: Head "https://k8s.gcr.io/v2/pause/manifests/3.3": dial tcp 74.125.204.82:443: i/o timeout
Sep 06 01:38:48 node-3-bak kubelet[10631]: E0906 01:38:48.503519   10631 kuberuntime_manager.go:741] createPodSandbox for pod "nginx-proxy-node-3-bak_kube-system(4068a5c786f6c77faa6538742baa1d44)" failed: rpc error: code = Unknown desc = failed to get sandbox image "k8s.gcr.io/pause:3.3": failed to pull image "k8s.gcr.io/pause:3.3": failed to pull and unpack image "k8s.gcr.io/pause:3.3": failed to resolve reference "k8s.gcr.io/pause:3.3": failed to do request: Head "https://k8s.gcr.io/v2/pause/manifests/3.3": dial tcp 74.125.204.82:443: i/o timeout
Sep 06 01:38:48 node-3-bak kubelet[10631]: E0906 01:38:48.503591   10631 pod_workers.go:191] Error syncing pod 4068a5c786f6c77faa6538742baa1d44 ("nginx-proxy-node-3-bak_kube-system(4068a5c786f6c77faa6538742baa1d44)"), skipping: failed to "CreatePodSandbox" for "nginx-proxy-node-3-bak_kube-system(4068a5c786f6c77faa6538742baa1d44)" with CreatePodSandboxError: "CreatePodSandbox for pod \"nginx-proxy-node-3-bak_kube-system(4068a5c786f6c77faa6538742baa1d44)\" failed: rpc error: code = Unknown desc = failed to get sandbox image \"k8s.gcr.io/pause:3.3\": failed to pull image \"k8s.gcr.io/pause:3.3\": failed to pull and unpack image \"k8s.gcr.io/pause:3.3\": failed to resolve reference \"k8s.gcr.io/pause:3.3\": failed to do request: Head \"https://k8s.gcr.io/v2/pause/manifests/3.3\": dial tcp 74.125.204.82:443: i/o timeout

发现k8s.gcr.io/pause:3.3镜像被删掉了，重新pull问题也决解了。但是一开始一直被 Failed to contact API server when waiting for CSINode publishing: Get “https://localhost:6443/apis/storage.k8s.io/v1/csinodes/node-3-bak”: dial tcp 127.0.0.1:6443: connect: connection refused这个报错信息误导了很久，6443是master节点的kube-apiserver在监听，本地是worker子节点本来就没有这个服务。
我主从节点配置都是正确的，node-3-bak只是worker节点，pause:3.3拉取不了和本地为啥要请求127.0.0.1:6443而不是master的节点的呢？