velero: Fail get DataUploadResult for restore - multiple DataUpload result cms found

What steps did you take and what happened: When trying to restore a backup I get the following error:

    rc-aimit-io:  error preparing persistentvolumeclaims/rc-aimit-io/datadir-rocketchat-mongodb-0: rpc error: code = Unknown desc = fail get DataUploadResult for restore: velero-imdc-202311
05205455-20231105210928: multiple DataUpload result cms found with labels velero.io/pvc-namespace-name=rc-aimit-io.datadir-rocketchat-mongodb-0,velero.io/restore-uid=630b736f-e04a-4fb7-b937
-7654feb4062e,velero.io/resource-usage=DataUpload
                  error preparing persistentvolumeclaims/rc-aimit-io/datadir-rocketchat-mongodb-1: rpc error: code = Unknown desc = fail get DataUploadResult for restore: velero-imdc-202311
05205455-20231105210928: multiple DataUpload result cms found with labels velero.io/pvc-namespace-name=rc-aimit-io.datadir-rocketchat-mongodb-1,velero.io/restore-uid=630b736f-e04a-4fb7-b937
-7654feb4062e,velero.io/resource-usage=DataUpload

The restore command velero restore create --from-backup velero-imdc-20231105205455 --include-namespaces rc-aimit-io --include-resources persistentvolumeclaims,persistentvolumes,dataupload

What did you expect to happen: The PV/PVCs would be restored.

The following information will help us better understand what’s going on: bundle-2023-11-05-23-09-42.tar.gz

Anything else you would like to add: I was able to confirm the data is there by manually mounting the bucket using Kopia tools and restoring the files from there.

Environment:

  • Velero version (use velero version):
Client:
        Version: v1.12.1
        Git commit: 5c4fdfe147357ec7b908339f4516cd96d6b97c61
Server:
        Version: v1.12.1
  • Velero features (use velero client config get features): features: <NOT SET>
  • Kubernetes version (use kubectl version):
Client Version: version.Info{Major:"1", Minor:"22", GitVersion:"v1.22.17+rke2r1", GitCommit:"a7736eaf34d823d7652415337ac0ad06db9167fc", GitTreeState:"clean", BuildDate:"2022-12-08T21:42:35Z", GoVersion:"go1.16.15b7", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"22", GitVersion:"v1.22.17+rke2r1", GitCommit:"a7736eaf34d823d7652415337ac0ad06db9167fc", GitTreeState:"clean", BuildDate:"2022-12-08T21:42:35Z", GoVersion:"go1.16.15b7", Compiler:"gc", Platform:"linux/amd64"}
  • Kubernetes installer & version:
  • RKE2
  • v1.22.17+rke2r1

About this issue

  • Original URL
  • State: open
  • Created 3 years ago
  • Reactions: 1
  • Comments: 41 (20 by maintainers)

Most upvoted comments

Facing same issue when trying to restore with velero version 1.14

For 1, see issue #8696 For 2, see issue #8663

I don’t think it has been included in the doc. It is an untested scenario, the current identified issue is this one but there may be more.

Could you share your case that Velero namespace backup is required? During our assessment, we don’t see it is an important requirement. We may change its priority if we see one or more valid cases.

The problem for me isn’t that the velero namespace needs to be included - it’s that the docs don’t say it must be excluded. I set up a Velero schedule designed to backup everything in the cluster (no exclude* settings used), so the velero namespace was automatically included. When I went to test those backups, I found the volumes couldn’t be restored.

At a minimum, the docs need to mention that including the velero namespace in a backup may cause issues.

For the doc requirement, a separate doc issue #8663 is opened.

If this is true, can we change the default to permanently exclude the velero namespace, or even better, maybe, always exclude velero resources by default?