Open naama-connecteam opened 1 year ago
@naama-connecteam Did you use CSI or Azure plugin to do the volume snapshot?
if so, the providerSnapshotID filed should be generated by Azure volume snapshot, which format would be like this /subscriptions/2261f3e7-d159-48fe-95a3-0e6a96exxx/resourceGroups/MC_xxx-test_eastus/providers/Microsoft.Compute/snapshots/restore-82ca08c2-3972-4511-acdc-b0e74559c24-8cd76dd9-95d2-4353-b78e-xxxx"
, the providerSnapshotID is generated by the provider's snapshot service, Which is a random character string and Velero retrieved it, I'm not sure about the format you combined with is correct
Is it any option to use Velero and backup volume snapshots into two different Azure resource groups with two different locations?
Once I've used Azure CSI instead of Velero CSI plugin, it is possible to create snapshots into two different Azure resource groups but location of snapshots are always the same as AKS location. If it is not supported by Velero (and Azure maybe) and Azure snapshots are copied manually to different Azure location, what should be changed in Velero stored backups to use copied Azure snapshots?
Any news about this topic?
We’re trying to use Velero to restore a backup across different regions in Azure. We copied the snapshots manually to a new storage account (in a new resource group) we created in the region where we want to restore, and manually changed the references to the snapshots location in
<BACKUP_NAME>-volumesnapshots.json
file, specifically the field “providerSnapshotID”. The change in the json is fromMC_k8s-env-<SRC_ENV>-production_env-<SRC_ENV>-production_<SRC_REGION>
toMC_k8s-env-<DEST_ENV>-production_env-<DEST_ENV>-production_<DEST_REGION>
.When we created the restore, we got this error for all the volumes:
The resource group name in the error doesn’t exist. It is a combination of the new resourcegroup name and the old region:
MC_k8s-env-<DEST_ENV>-production_env-<DEST_ENV>-production_<SRC_REGION>
which we cannot find in any of the json files and we don’t know how it is generated.Also I can't find any error when I run
velero restore logs backup-mongo-XXX
, or look at the pod's logs.