Open survivant opened 3 years ago
Can we get output pod describe which is not able to mount? Is it still reproducible in the latest version of OpenEBS? I quick tried to reproduce but didn't hit the issue. Can you try in your setup and close the issue if it no longer reproducible:
Steps I tried in 2.12.0 version of OpenEBS
## PVC
[sai@centos-master ~]$ kubectl get pvc
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
csi-pvc Bound pvc-b29bf7ca-c569-4493-ac95-5cd6bb5fa2c3 10Gi RWO openebs-csi-sc 3h17m
## Pod consuming above PVC
[sai@centos-master ~]$ kubectl get po
NAME READY STATUS RESTARTS AGE
fio-2-deployment-6c948d6799-hknfd 1/1 Running 0 7m45s
# Expanded the volume and deleted the pod consuming the volume
[sai@centos-master ~]$ kubectl patch pvc csi-pvc --patch '{"spec": {"resources": {"requests": {"storage": "11Gi"}}}}' && kubectl delete po fio-2-deployment-6c948d6799-hknfd
persistentvolumeclaim/csi-pvc patched
pod "fio-2-deployment-6c948d6799-hknfd" deleted
^C
[sai@centos-master ~]$ kubectl get po
NAME READY STATUS RESTARTS AGE
fio-2-deployment-6c948d6799-87sk7 1/1 Running 0 17s
fio-2-deployment-6c948d6799-hknfd 1/1 Terminating 0 8m30s
# Again deleted the pod after coming to running state
[sai@centos-master ~]$ kubectl delete po fio-2-deployment-6c948d6799-87sk7
pod "fio-2-deployment-6c948d6799-87sk7" deleted
^C
[sai@centos-master ~]$ kubectl get po -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
fio-2-deployment-6c948d6799-87sk7 1/1 Terminating 0 35s 10.244.1.238 centos-worker-1 <none> <none>
fio-2-deployment-6c948d6799-fljbp 0/1 ContainerCreating 0 11s <none> centos-master <none> <none>
# pod came to running state
[sai@centos-master ~]$ kubectl get po
NAME READY STATUS RESTARTS AGE
fio-2-deployment-6c948d6799-fljbp 1/1 Running 0 52s
I followed this guide : https://github.com/openebs/cstor-operators/blob/master/docs/tutorial/volumes/resize.md
After that I deleted the pod. The pod was never able to mount the PV. I had to scale down to 0 the deployment to fix the problem