Description of problem: In volume staging the driver scans the volume and if it doesn't have a file system on it, it will create one. When /run/udev is missing lsblk will not have any information on the filesystem, and with not failure reported. in other words this will overwrite the volume filesystem. Version-Release number of selected component (if applicable): How reproducible: 100%
Tested on: openshift-4.4.0-0.nightly-2020-06-14-013357 (the ovirt-csi operator functionality doesn't depend on the ocp version itself, the bz is marked as 4.5, because the operator is being added in this version) ovirt-engine-4.4.1.2-0.10.el8ev.noarch Steps: - Manually installed the operator following: https://github.com/openshift/ovirt-csi-driver-operator - Created and deployed a test pod that uses csi pvc disk # oc create -f testpodwithcsi.yaml - Once installed and in operation, created a dummy file in the mounted pv directory (my case /opt): # oc rsh testpodwithcsi # cd /opt && touch dummy.file - Deleted the pod, without touching the pvc, and deployed a new one: # oc delete pod testpodwithcsi && oc create -f testpodwithcsi.yaml - Checked if the created file still there (thus, no FS rewritten) Results: Failed, the FS was rewritten, because the yaml file that got the fix is not being used by the operator
Hey Guilherme, Could you check bug on 4.5 version? The bug is added in 4.5 released.
Retargeting because our operator will be available only in 4.6
Will test it on 4.6
Verified on: openshift-4.6.0-0.nightly-2020-07-22-074636 ovirt-engine-4.3.11.2-0.1.el7.noarch Steps: - same as comment #5 Results: created file stills there, FS not rewritten
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (OpenShift Container Platform 4.6 GA Images), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:4196