ambiknai
commented
2 years ago Actual error when mount point doesnt exist
I1021 04:55:24.204564 1 nodeserver.go:264] Output: Output string error mountpoint: /var/data/kubelet/pods/a3c4dee5-ab87-4abf-93f7-db3802fa0ce1/volumes/kubernetes.io~csi/pvc-917efa96-ed15-475f-bbe6-48578720a70c/mount: No such file or directory
E1021 04:55:24.204597 1 nodeserver.go:126] Can not validate target mount point: /var/data/kubelet/pods/a3c4dee5-ab87-4abf-93f7-db3802fa0ce1/volumes/kubernetes.io~csi/pvc-917efa96-ed15-475f-bbe6-48578720a70c/mount exit status 1
E1021 04:55:24.204609 1 server.go:155] GRPC error: rpc error: code = Internal desc = exit status 1Test Results
➜ csi-object oc get pods
NAME READY STATUS RESTARTS AGE
cos-csi-app 1/1 Running 0 6m33s
cos-s3-csi-attacher-0 1/1 Running 0 7m42s
cos-s3-csi-controller-0 2/2 Running 0 7m41s
cos-s3-csi-driver-bcl6m 2/2 Running 0 8m15s
cos-s3-csi-driver-rdqqr 2/2 Running 0 8m16s
cos-s3-csi-driver-wxldt 2/2 Running 0 94s
➜ csi-object oc exec -it cos-csi-app /bin/bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
root@cos-csi-app:/# mount | grep s3fs
s3fs on /data type fuse.s3fs (rw,nosuid,nodev,relatime,user_id=0,group_id=0,allow_other)
root@cos-csi-app:/# ls /data
root@cos-csi-app:/# cd /data
root@cos-csi-app:/data# echo "test mount" > new.txt
root@cos-csi-app:/data# exit
exit
➜ csi-object oc get pods
oc delete pods NAME READY STATUS RESTARTS AGE
cos-csi-app 1/1 Running 0 10m
cos-s3-csi-attacher-0 1/1 Running 0 11m
cos-s3-csi-controller-0 2/2 Running 0 11m
cos-s3-csi-driver-bcl6m 2/2 Running 0 11m
cos-s3-csi-driver-rdqqr 2/2 Running 0 11m
cos-s3-csi-driver-wxldt 2/2 Running 0 5m11s
➜ csi-object oc delete pods cos-csi-app
pod "cos-csi-app" deleted
➜ csi-object oc get pods
NAME READY STATUS RESTARTS AGE
cos-s3-csi-attacher-0 1/1 Running 0 14m
cos-s3-csi-controller-0 2/2 Running 0 14m
cos-s3-csi-driver-bcl6m 2/2 Running 0 14m
cos-s3-csi-driver-rdqqr 2/2 Running 0 14m
cos-s3-csi-driver-wxldt 2/2 Running 0 8m17s
➜ csi-object oc create -f cos-csi-app.yaml
Warning: would violate PodSecurity "restricted:latest": allowPrivilegeEscalation != false (container "app-frontend" must set securityContext.allowPrivilegeEscalation=false), unrestricted capabilities (container "app-frontend" must set securityContext.capabilities.drop=["ALL"]), runAsNonRoot != true (pod or container "app-frontend" must set securityContext.runAsNonRoot=true), seccompProfile (pod or container "app-frontend" must set securityContext.seccompProfile.type to "RuntimeDefault" or "Localhost")
pod/cos-csi-app created
➜ csi-object oc get pods
NAME READY STATUS RESTARTS AGE
cos-csi-app 1/1 Running 0 17s
cos-s3-csi-attacher-0 1/1 Running 0 14m
cos-s3-csi-controller-0 2/2 Running 0 14m
cos-s3-csi-driver-bcl6m 2/2 Running 0 15m
cos-s3-csi-driver-rdqqr 2/2 Running 0 15m
cos-s3-csi-driver-wxldt 2/2 Running 0 8m48s
➜ csi-object oc exec -it cos-csi-app /bin/bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
root@cos-csi-app:/# ls /data
new.txt
root@cos-csi-app:/# exit
exit
➜ csi-object oc logs cos-s3-csi-driver-wxldt -c cos-s3-csi-driver > cos-s3-csi-driver.log
➜ csi-object cat cos-s3-csi-driver.log | grep "System has not been"
➜ csi-object
driver fails producing only a known no-harm error from linux-mounter pckeage if waiting for mount is enabled
temp fix is to comment out call to the method https://github.com/IBM/satellite-object-storage-plugin/blob/main/pkg/mounter/mounter.go#L53