Bug 2128872
| Summary: | [4.11]Can't restore cloned VM | ||
|---|---|---|---|
| Product: | Container Native Virtualization (CNV) | Reporter: | dalia <dafrank> |
| Component: | Storage | Assignee: | Álvaro Romero <alromero> |
| Status: | CLOSED ERRATA | QA Contact: | dalia <dafrank> |
| Severity: | unspecified | Docs Contact: | |
| Priority: | high | ||
| Version: | 4.11.0 | CC: | akalenyu, alitke, alromero, cnv-qe-bugs, j.thadden, mrashish, yadu |
| Target Milestone: | --- | ||
| Target Release: | 4.11.1 | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | virt-controller v4.11.1-11, CNV v4.11.1-92 | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2022-12-01 21:12:19 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
Note that VM clone isn't in 4.11 (the target release branch), but the fix might apply to this version anyway, what do you think alvaro? (In reply to Maya Rashish from comment #2) > Note that VM clone isn't in 4.11 (the target release branch), but the fix > might apply to this version anyway, what do you think alvaro? Thanks for noticing! Since the bug is in the restore controller and might still affect PVCs using dataSourceRef, I think it won't hurt to backport this. (In reply to Álvaro Romero from comment #3) > (In reply to Maya Rashish from comment #2) > > Note that VM clone isn't in 4.11 (the target release branch), but the fix > > might apply to this version anyway, what do you think alvaro? > > Thanks for noticing! Since the bug is in the restore controller and might > still affect PVCs using dataSourceRef, I think it won't hurt to backport > this. Now that I'm taking a closer look at the backport's code, I agree it might be unnecessary since we didn't create any restore PVC. I'll take a closer look but after seeing this I think it shouldn't be backported. (In reply to Álvaro Romero from comment #4) > (In reply to Álvaro Romero from comment #3) > > (In reply to Maya Rashish from comment #2) > > > Note that VM clone isn't in 4.11 (the target release branch), but the fix > > > might apply to this version anyway, what do you think alvaro? > > > > Thanks for noticing! Since the bug is in the restore controller and might > > still affect PVCs using dataSourceRef, I think it won't hurt to backport > > this. > > Now that I'm taking a closer look at the backport's code, I agree it might > be unnecessary since we didn't create any restore PVC. I'll take a closer > look but after seeing this I think it shouldn't be backported. Ignore my last comment, the PVC is created. I'll take a better look. Editing the target release was a mistake on my part. Reverting. I thought about doing it, but instead wanted to ask Alvaro first, and forgot to undo it. This error also occurs on any of the templates having their disks being derived from openshift-os-images in ODF. Workaround to be able to restore snapshots: - any time BEFORE you restore the snapshot edit the virtualmachinesnapshotcontent oc edit virtualmachinesnapshotcontent.snapshot.kubevirt.io/vmsnapshot-content-1f0fe2f2-324b-4104-af25-cd2ccf90cef7 - search for dataSourceRef: (spec.volumeBackups[persistentVolumeClaim].spec.dataSourceRef) - remove it Now you can restore the snapshot without any problem. For sure this have to be done for each and every snapshot you do on such a machine. For anyone stuck in restoring such a machine, here is the way to come out of this without removing the VM: - search for virtualmachinerestore.snapshot.kubevirt.io object that is stuck and shows "false" in COMPLETE column oc get virtualmachinerestore.snapshot.kubevirt.io - delete the object oc delete oc delete virtualmachinerestore.snapshot.kubevirt.io/resotre-snapshot-<whatevername> - extend oc/kubectl by installing krew (https://krew.sigs.k8s.io/docs/user-guide/setup/install/) and edit-status (https://github.com/ulucinar/kubectl-edit-status) - then remove the status.restoreInProgress line oc edit-status virtualmachine.kubevirt.io/<yourMachine> - now the VM can be started again virtctl start <yourMachine> verified on CNV 4.11.1 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 (Moderate: OpenShift Virtualization 4.11.1 security and bug fix update), 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/RHSA-2022:8750 |
Description of problem: Fail to restore a cloned VM. Version-Release number of selected component (if applicable): 4.12 How reproducible: 100% Steps to Reproduce: 1. Create a VM 2. Clone from it PVC to a new VM 3. Snapshot the cloned VM 4. Restore the VM Actual results: VM restored Expected results: restore failed Additional info: restore status: reason: 'PersistentVolumeClaim "restore-3327a680-b539-4604-8358-69f3a0b25d69-dv-disk" is invalid: spec: Invalid value: field.Path{name:"dataSource", index:"", parent:(*field.Path)(0xc0955e66c0)}: must match dataSourceRef' status: "False" ----------------------------- PVC spec: dataSource: apiGroup: snapshot.storage.k8s.io kind: VolumeSnapshot name: cloned-dv dataSourceRef: apiGroup: snapshot.storage.k8s.io kind: VolumeSnapshot name: cloned-dv