Kubernetes 無法啟動：無法建構初始容器的映射

從今天開始，k3s 無法啟動並出現以下錯誤："Failed to start ContainerManager" err="failed to build map of initial containers from runtime: no PodsandBox found with Id '9f141a500138e081ae1a641d7d4c00c3029ecce87da6e2fc80f4a14bd0a965fd'. 在此日誌行之後，它崩潰了。

我在網際網路上找不到任何東西，所以這裡有人知道如何解決這個問題嗎？

我正在執行 k3s 版本：k3s version v1.21.5+k3s2 (724ef700)

如果我需要提供更多詳細資訊，請告訴我。

日誌：

...
I1021 12:04:55.508161   78816 kuberuntime_manager.go:222] "Container runtime initialized" containerRuntime="containerd" version="v1.4.11-k3s1" apiVersion="v1alpha2"
I1021 12:04:55.508361   78816 server.go:1191] "Started kubelet"
E1021 12:04:55.509247   78816 cri_stats_provider.go:369] "Failed to get the info of the filesystem with mountpoint" err="unable to find data in memory cache" mountpoint="/var/lib/rancher/k3s/agent/containerd/io.containerd.snapshotter.v1.overlayfs"
E1021 12:04:55.509273   78816 kubelet.go:1306] "Image garbage collection failed once. Stats initialization may not have completed yet" err="invalid capacity 0 on image filesystem"
I1021 12:04:55.509255   78816 server.go:149] "Starting to listen" address="0.0.0.0" port=10250
I1021 12:04:55.509887   78816 server.go:409] "Adding debug handlers to kubelet server"
I1021 12:04:55.510952   78816 fs_resource_analyzer.go:67] "Starting FS ResourceAnalyzer"
I1021 12:04:55.512769   78816 scope.go:111] "RemoveContainer" containerID="e5ce1c151a24558e69f544794a15bb6d1238139439a0c6174acf720a4f531a7c"
I1021 12:04:55.512865   78816 volume_manager.go:271] "Starting Kubelet Volume Manager"
I1021 12:04:55.512923   78816 desired_state_of_world_populator.go:141] "Desired state populator starts to run"
INFO[2021-10-21T12:04:55.516702675+02:00] RemoveContainer for "e5ce1c151a24558e69f544794a15bb6d1238139439a0c6174acf720a4f531a7c" 
DEBU[2021-10-21T12:04:55.527595023+02:00] openat2 not available, falling back to securejoin 
I1021 12:04:55.538886   78816 controller.go:611] quota admission added evaluator for: leases.coordination.k8s.io
I1021 12:04:55.545188   78816 kubelet_network_linux.go:56] "Initialized protocol iptables rules." protocol=IPv4
I1021 12:04:55.561242   78816 kubelet_network_linux.go:56] "Initialized protocol iptables rules." protocol=IPv6
I1021 12:04:55.561266   78816 status_manager.go:157] "Starting to sync pod status with apiserver"
I1021 12:04:55.561282   78816 kubelet.go:1846] "Starting kubelet main sync loop"
E1021 12:04:55.561318   78816 kubelet.go:1870] "Skipping pod synchronization" err="[container runtime status check may not have completed yet, PLEG is not healthy: pleg has yet to be successful]"
I1021 12:04:55.571567   78816 shared_informer.go:247] Caches are synced for endpoint slice config 
I1021 12:04:55.571570   78816 shared_informer.go:247] Caches are synced for service config 
INFO[2021-10-21T12:04:55.604476442+02:00] RemoveContainer for "e5ce1c151a24558e69f544794a15bb6d1238139439a0c6174acf720a4f531a7c" returns successfully 
I1021 12:04:55.604584   78816 scope.go:111] "RemoveContainer" containerID="4d7578dd7f7574fd5deeae1ed53cf67d0a2fe64aa1d1214b1ba865622c05b4cd"
INFO[2021-10-21T12:04:55.604877204+02:00] labels have been set successfully on node: <node name>
INFO[2021-10-21T12:04:55.604936435+02:00] RemoveContainer for "4d7578dd7f7574fd5deeae1ed53cf67d0a2fe64aa1d1214b1ba865622c05b4cd" 
I1021 12:04:55.612875   78816 kuberuntime_manager.go:1044] "Updating runtime config through cri with podcidr" CIDR="10.42.0.0/24"
INFO[2021-10-21T12:04:55.612967745+02:00] No cni config template is specified, wait for other system components to drop the config. 
I1021 12:04:55.613044   78816 kubelet_network.go:76] "Updating Pod CIDR" originalPodCIDR="" newPodCIDR="10.42.0.0/24"
I1021 12:04:55.623215   78816 kubelet_node_status.go:71] "Attempting to register node" node="<node name>"
E1021 12:04:55.645403   78816 kubelet.go:1384] "Failed to start ContainerManager" err="failed to build map of initial containers from runtime: no PodsandBox found with Id '9f141a500138e081ae1a641d7d4c00c3029ecce87da6e2fc80f4a14bd0a965fd'"

在https://github.com/kubernetes/kubelet/issues/21的幫助下，我終於弄明白了。使用以下命令手動啟動 containerd 後：（containerd -c /var/lib/rancher/k3s/agent/etc/containerd/config.toml -a /run/k3s/containerd/containerd.sock --state /run/k3s/containerd --root /var/lib/rancher/k3s/agent/containerd我在 k3s 日誌中找到），我可以使用 crictl 搜尋容器：k3s crictl ps -a | grep 9f141a它給了我一個容器 ID。然後我刪除了 podk3s crictl rm <id>並重新啟動了 k3s，現在它又可以工作了。

引用自：https://serverfault.com/questions/1081253