Il lavoratore inizia a fallire CSINodeIfo: errore durante l'aggiornamento dell'annotazione CSINode

10

Ho creato un cluster kubernetes 1 master e 2 nodi worker 2 mesi fa, oggi un nodo worker ha iniziato a fallire e non so perché. Penso che non sia successo niente di insolito al mio lavoratore.

Ho usato flanella e kubeadm per creare il cluster e funzionava molto bene.

Se descrivo il nodo:

tommy@bxybackend:~$ kubectl describe node bxybackend-node01
Name:               bxybackend-node01
Roles:              <none>
Labels:             beta.kubernetes.io/arch=amd64
                    beta.kubernetes.io/os=linux
                    kubernetes.io/arch=amd64
                    kubernetes.io/hostname=bxybackend-node01
                    kubernetes.io/os=linux
Annotations:        flannel.alpha.coreos.com/backend-data: {"VtepMAC":"06:ca:97:82:50:10"}
                    flannel.alpha.coreos.com/backend-type: vxlan
                    flannel.alpha.coreos.com/kube-subnet-manager: true
                    flannel.alpha.coreos.com/public-ip: 10.168.10.4
                    kubeadm.alpha.kubernetes.io/cri-socket: /var/run/dockershim.sock
                    node.alpha.kubernetes.io/ttl: 0
                    volumes.kubernetes.io/controller-managed-attach-detach: true
CreationTimestamp:  Sun, 03 Nov 2019 09:41:48 -0600
Taints:             node.kubernetes.io/not-ready:NoExecute
                    node.kubernetes.io/not-ready:NoSchedule
Unschedulable:      false
Conditions:
  Type             Status  LastHeartbeatTime                 LastTransitionTime                Reason                       Message
  ----             ------  -----------------                 ------------------                ------                       -------
  MemoryPressure   False   Wed, 11 Dec 2019 11:17:05 -0600   Wed, 11 Dec 2019 10:37:19 -0600   KubeletHasSufficientMemory   kubelet has sufficient memory available
  DiskPressure     False   Wed, 11 Dec 2019 11:17:05 -0600   Wed, 11 Dec 2019 10:37:19 -0600   KubeletHasNoDiskPressure     kubelet has no disk pressure
  PIDPressure      False   Wed, 11 Dec 2019 11:17:05 -0600   Wed, 11 Dec 2019 10:37:19 -0600   KubeletHasSufficientPID      kubelet has sufficient PID available
  Ready            False   Wed, 11 Dec 2019 11:17:05 -0600   Wed, 11 Dec 2019 10:37:19 -0600   KubeletNotReady              Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Addresses:
  InternalIP:  10.168.10.4
  Hostname:    bxybackend-node01
Capacity:
 cpu:                12
 ephemeral-storage:  102684600Ki
 hugepages-1Gi:      0
 hugepages-2Mi:      0
 memory:             14359964Ki
 pods:               110
Allocatable:
 cpu:                12
 ephemeral-storage:  94634127204
 hugepages-1Gi:      0
 hugepages-2Mi:      0
 memory:             14257564Ki
 pods:               110
System Info:
 Machine ID:                 3afa24bb05994ceaaf00e7f22b9322ab
 System UUID:                80951742-F69F-6487-F2F7-BE2FB7FEFBF8
 Boot ID:                    115fbacc-143d-4007-90e4-7fdcb5462680
 Kernel Version:             4.15.0-72-generic
 OS Image:                   Ubuntu 18.04.3 LTS
 Operating System:           linux
 Architecture:               amd64
 Container Runtime Version:  docker://18.9.7
 Kubelet Version:            v1.17.0
 Kube-Proxy Version:         v1.17.0
PodCIDR:                     10.244.1.0/24
PodCIDRs:                    10.244.1.0/24
Non-terminated Pods:         (2 in total)
  Namespace                  Name                           CPU Requests  CPU Limits  Memory Requests  Memory Limits  AGE
  ---------                  ----                           ------------  ----------  ---------------  -------------  ---
  kube-system                kube-flannel-ds-amd64-sslbg    100m (0%)     100m (0%)   50Mi (0%)        50Mi (0%)      8m31s
  kube-system                kube-proxy-c5gxc               0 (0%)        0 (0%)      0 (0%)           0 (0%)         8m52s
Allocated resources:
  (Total limits may be over 100 percent, i.e., overcommitted.)
  Resource           Requests   Limits
  --------           --------   ------
  cpu                100m (0%)  100m (0%)
  memory             50Mi (0%)  50Mi (0%)
  ephemeral-storage  0 (0%)     0 (0%)
Events:
  Type     Reason                   Age                  From                           Message
  ----     ------                   ----                 ----                           -------
  Warning  SystemOOM                52m                  kubelet, bxybackend-node01     System OOM encountered, victim process: dotnet, pid: 12170
  Normal   NodeHasNoDiskPressure    52m (x12 over 38d)   kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeHasNoDiskPressure
  Normal   NodeHasSufficientPID     52m (x12 over 38d)   kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeHasSufficientPID
  Normal   NodeNotReady             52m (x6 over 23d)    kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeNotReady
  Normal   NodeHasSufficientMemory  52m (x12 over 38d)   kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeHasSufficientMemory
  Warning  ContainerGCFailed        52m (x3 over 6d23h)  kubelet, bxybackend-node01     rpc error: code = DeadlineExceeded desc = context deadline exceeded
  Normal   NodeReady                52m (x13 over 38d)   kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeReady
  Normal   NodeAllocatableEnforced  43m                  kubelet, bxybackend-node01     Updated Node Allocatable limit across pods
  Warning  SystemOOM                43m                  kubelet, bxybackend-node01     System OOM encountered, victim process: dotnet, pid: 9699
  Warning  SystemOOM                43m                  kubelet, bxybackend-node01     System OOM encountered, victim process: dotnet, pid: 12639
  Warning  SystemOOM                43m                  kubelet, bxybackend-node01     System OOM encountered, victim process: dotnet, pid: 16194
  Warning  SystemOOM                43m                  kubelet, bxybackend-node01     System OOM encountered, victim process: dotnet, pid: 19618
  Warning  SystemOOM                43m                  kubelet, bxybackend-node01     System OOM encountered, victim process: dotnet, pid: 12170
  Normal   Starting                 43m                  kubelet, bxybackend-node01     Starting kubelet.
  Normal   NodeHasSufficientMemory  43m (x2 over 43m)    kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeHasSufficientMemory
  Normal   NodeHasSufficientPID     43m (x2 over 43m)    kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeHasSufficientPID
  Normal   NodeNotReady             43m                  kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeNotReady
  Normal   NodeHasNoDiskPressure    43m (x2 over 43m)    kubelet, bxybackend-node01     Node bxybackend-node01 status is now: NodeHasNoDiskPressure
  Normal   Starting                 42m                  kubelet, bxybackend-node01     Starting kubelet.

Se guardo syslog nel lavoratore: 

Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.552152   19331 kuberuntime_manager.go:981] updating runtime config through cri with podcidr 10.244.1.0/24
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.552162   19331 kubelet_node_status.go:294] Setting node annotation to enable volume controller attach/detach
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.552352   19331 docker_service.go:355] docker cri received runtime config &RuntimeConfig{NetworkConfig:&NetworkConfig{PodCidr:10.244.1.0/24,},}
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.552600   19331 kubelet_network.go:77] Setting Pod CIDR:  -> 10.244.1.0/24
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.555142   19331 kubelet_node_status.go:70] Attempting to register node bxybackend-node01
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.652843   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "kube-proxy" (UniqueName: "kubernetes.io/configmap/d6b534db-c32c-491b-a665-cf1ccd6cd089-kube-proxy") pod "kube-proxy-c5gxc" (UID: "d6b534db-c32c-491b-a665-cf1ccd6cd089")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753179   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "xtables-lock" (UniqueName: "kubernetes.io/host-path/d6b534db-c32c-491b-a665-cf1ccd6cd089-xtables-lock") pod "kube-proxy-c5gxc" (UID: "d6b534db-c32c-491b-a665-cf1ccd6cd089")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753249   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "lib-modules" (UniqueName: "kubernetes.io/host-path/d6b534db-c32c-491b-a665-cf1ccd6cd089-lib-modules") pod "kube-proxy-c5gxc" (UID: "d6b534db-c32c-491b-a665-cf1ccd6cd089")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753285   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "kube-proxy-token-ztrh4" (UniqueName: "kubernetes.io/secret/d6b534db-c32c-491b-a665-cf1ccd6cd089-kube-proxy-token-ztrh4") pod "kube-proxy-c5gxc" (UID: "d6b534db-c32c-491b-a665-cf1ccd6cd089")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753316   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "run" (UniqueName: "kubernetes.io/host-path/6a2299cf-63a4-4e96-8b3b-acd373de12c2-run") pod "kube-flannel-ds-amd64-sslbg" (UID: "6a2299cf-63a4-4e96-8b3b-acd373de12c2")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753342   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "cni" (UniqueName: "kubernetes.io/host-path/6a2299cf-63a4-4e96-8b3b-acd373de12c2-cni") pod "kube-flannel-ds-amd64-sslbg" (UID: "6a2299cf-63a4-4e96-8b3b-acd373de12c2")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753461   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "flannel-cfg" (UniqueName: "kubernetes.io/configmap/6a2299cf-63a4-4e96-8b3b-acd373de12c2-flannel-cfg") pod "kube-flannel-ds-amd64-sslbg" (UID: "6a2299cf-63a4-4e96-8b3b-acd373de12c2")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753516   19331 reconciler.go:209] operationExecutor.VerifyControllerAttachedVolume started for volume "flannel-token-ts2qt" (UniqueName: "kubernetes.io/secret/6a2299cf-63a4-4e96-8b3b-acd373de12c2-flannel-token-ts2qt") pod "kube-flannel-ds-amd64-sslbg" (UID: "6a2299cf-63a4-4e96-8b3b-acd373de12c2")
Dec 11 11:20:10 bxybackend-node01 kubelet[19331]: I1211 11:20:10.753531   19331 reconciler.go:156] Reconciler: start to sync state
Dec 11 11:20:12 bxybackend-node01 kubelet[19331]: I1211 11:20:12.052813   19331 kubelet_node_status.go:112] Node bxybackend-node01 was previously registered
Dec 11 11:20:12 bxybackend-node01 kubelet[19331]: I1211 11:20:12.052921   19331 kubelet_node_status.go:73] Successfully registered node bxybackend-node01
Dec 11 11:20:13 bxybackend-node01 kubelet[19331]: E1211 11:20:13.051159   19331 csi_plugin.go:267] Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Dec 11 11:20:16 bxybackend-node01 kubelet[19331]: E1211 11:20:16.051264   19331 csi_plugin.go:267] Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Dec 11 11:20:18 bxybackend-node01 kubelet[19331]: E1211 11:20:18.451166   19331 csi_plugin.go:267] Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Dec 11 11:20:21 bxybackend-node01 kubelet[19331]: E1211 11:20:21.251289   19331 csi_plugin.go:267] Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Dec 11 11:20:25 bxybackend-node01 kubelet[19331]: E1211 11:20:25.019276   19331 csi_plugin.go:267] Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Dec 11 11:20:46 bxybackend-node01 kubelet[19331]: E1211 11:20:46.772862   19331 csi_plugin.go:267] Failed to initialize CSINodeInfo: error updating CSINode annotation: timed out waiting for the condition; caused by: the server could not find the requested resource
Dec 11 11:20:46 bxybackend-node01 kubelet[19331]: F1211 11:20:46.772895   19331 csi_plugin.go:281] Failed to initialize CSINodeInfo after retrying
Dec 11 11:20:46 bxybackend-node01 systemd[1]: kubelet.service: Main process exited, code=exited, status=255/n/a
Dec 11 11:20:46 bxybackend-node01 systemd[1]: kubelet.service: Failed with result 'exit-code'.
kubernetes  kubeadm 
— Tommy
fonte
vedo solo che il nodo problematico è in un'altra versione NOME STATO RUOLI VERSIONE ETÀ bxybackend Ready master 38d v1.16.2 bxybackend-node01 NotReady <none> 38d v1.17.0 bxybackend-node02 Ready <none> 38d v1.16.2
— Tommy

Risposte:

11

Durante l'installazione di kubeadm dovresti eseguire il comando seguente per contenere i pacchetti kubelet, kubeadm e kubectl e impedire che vengano aggiornati erroneamente. 

$ sudo apt-mark hold kubelet kubeadm kubectl

Ho riprodotto il tuo scenario e quello che è successo al tuo cluster è che 3 giorni fa è stata rilasciata una nuova versione di Kubernetes (v 1.17.0) e il tuo kubelet è stato aggiornato accidentalmente.

Sul nuovo Kubernetes sono state apportate alcune modifiche al CSI ed è per questo che hai dei problemi in questo nodo.

Ti consiglio di svuotare questo nodo, installarne uno nuovo con Kubernetes 1.16.2 e unire quello nuovo al tuo cluster.

Per svuotare questo nodo è necessario eseguire:

$ kubectl drain bxybackend-node01 --delete-local-data --force --ignore-daemonsets

Opzionalmente puoi effettuare il downgrade di kubelet alla versione precedente usando il seguente comando: 

$ sudo apt-get install kubelet=1.16.2-00

Non dimenticare di contrassegnare il tuo kubelet per evitare che venga aggiornato di nuovo: 

$ sudo apt-mark hold kubelet

Puoi usare il comando apt-mark showholdper elencare tutti i pacchetti in attesa e assicurarti che kubelet, kubeadm e kubectl siano in attesa.

Per aggiornare da 1.16.xa 1.17.x segui questa guida dalla documentazione di Kubernetes. L'ho convalidato e funziona come previsto.

— mWatney
fonte
Sto riscontrando lo stesso problema in cui sembra che kubernetes sia stato aggiornato automaticamente alla 1.17.0. Nessuno dei miei nodi segnala pronto, incluso il master. Stai suggerendo di ripristinare 1.16.2?
— Codice Junkie,
1
Si è verificato lo stesso problema dopo una patch di routine dei server nell'ambiente di test.
— Bhesh Gurung,
1
Tornando alla 1.16.4 ha funzionato, ci sono consigli su come passare alla 1.17?
— Codice Junkie,
2
@CodeJunkie Ho aggiunto informazioni su come eseguire l'aggiornamento a 1.17.
— mWatney,
2
Grazie, ho seguito la stessa guida ieri e sono riuscito ad aggiornarmi con successo. È possibile creare script per gestire automaticamente questi aggiornamenti durante l'applicazione di patch?
— Codice Junkie,
0

Ho anche affrontato questo stesso problema oggi su CentOS Linux versione 7.7.1908. La mia versione di kubernetes era v1.16.3 e ho eseguito il comando "yum update" e la versione di kubernetes aggiornata alla v1.17.0. Dopo di che ho fatto "annulla cronologia" no "e poi sono tornato alla vecchia versione di kubernetes ed è stato ricominciato a funzionare. Successivamente ho seguito il metodo di aggiornamento ufficiale e ora kubernetes v1.17.0 funziona senza problemi.

root@kube-master1:/root>kubectl get no -o wide
NAME           STATUS   ROLES    AGE    VERSION   INTERNAL-IP       EXTERNAL-IP   OS-IMAGE                KERNEL-VERSION               CONTAINER-RUNTIME
kube-master1   Ready    master   7d9h   v1.17.0   192.168.159.135   <none>        CentOS Linux 7 (Core)   3.10.0-1062.9.1.el7.x86_64   docker://1.13.1
kube-worker1   Ready    worker   7d9h   v1.17.0   192.168.159.136   <none>        CentOS Linux 7 (Core)   3.10.0-1062.9.1.el7.x86_64   docker://1.13.1
kube-worker2   Ready    worker   7d9h   v1.17.0   192.168.159.137   <none>        CentOS Linux 7 (Core)   3.10.0-1062.9.1.el7.x86_64   docker://1.13.1
root@kube-master1:/root>
— Arunabha
fonte
Grazie per il tuo post! Sarebbe bello se potessi pubblicare un link al official upgrade method!
— Segna il
@Mark, nel mio post puoi trovare questo link alla documentazione ufficiale dove è descritto come farlo su Ubuntu, Debian, HypriotOS, CentOS, RHEL o Fedora.
— mWatney il
Utilizzando il nostro sito, riconosci di aver letto e compreso le nostre Informativa sui cookie e Informativa sulla privacy.
Licensed under cc by-sa 3.0 with attribution required.