An Azure service that is used to provision Windows and Linux virtual machines.
Hi LuisRP-2369,
Based on your description, the two Azure Managed Disks were deleted through Kubernetes, and no Azure Backup recovery points or snapshots are available.
Although Azure Managed Disk deletion is normally permanent, there is a limited possibility of recovery through Microsoft Support. I recommend investigating this immediately.
Here is documented specific conditions under which the Azure engineering team may be able to recover deleted managed disks:
- The disk was created more than 10 days before deletion.
- The disk was deleted less than 96 hours (4 days) ago.
- Ultra Disks and Premium SSD v2 are excluded from the described recovery scenario.
These are reported eligibility conditions, not a guaranteed recovery service or a customer-accessible soft-delete feature. Only Microsoft Support can confirm whether your disks remain recoverable.
However there is Microsoft Q&A case from June 2026 documents a successful restoration by the Managed Disk Recovery Engineering Team, even though no backup or snapshot was available.
Open an Azure Support request for Azure Managed Disks and explicitly request an urgent assessment of backend recovery eligibility.
Provide the following information for both disks:
- Subscription ID and original managed disk resource IDs.
- Disk names, SKUs, and region (East US).
- Disk creation and deletion timestamps, preferably in UTC.
- Activity Log operation IDs/correlation IDs.
- Confirmation that the deletion originated from your AKS/KubeVirt environment.
Since the exact deletion time is unknown, use Azure Activity Logs to retrieve the operation details:
az monitor activity-log list \
--resource-id "<deleted-disk-resource-id>" \
--offset 90d \
--query "[].{Time:eventTimestamp,Operation:operationName.value,Status:status.value,CorrelationId:correlationId}" \
-o table
Also check whether any managed disks or snapshots remain in the AKS node resource group or other resource groups:
az disk list -o table
az snapshot list -o table
For Kubernetes, inspect any surviving PVs and their reclaim policies:
kubectl get pv -o yaml
kubectl get storageclass
If the affected PVs used persistentVolumeReclaimPolicy: Delete, deleting their associated PVCs could have triggered deletion of the underlying Azure disks.
Additionally, check for existing Kubernetes VolumeSnapshots, KubeVirt VM snapshots, CDI DataVolume clones, and backups stored outside Azure Backup.
If the disks have been permanently deleted and no recoverable copies exist, there is no supported self-service method to reconstruct the original disk contents.
Your remaining options would involve rebuilding the KubeVirt VM from its original image and restoring data from external backups, database replicas, exported VM images, or other application-level data sources.
For critical AKS/KubeVirt workloads, consider:
- Using a dedicated StorageClass with
reclaimPolicy: Retain. - Configuring Azure Backup for AKS for supported CSI-based persistent volumes.
- Maintaining independent, application-consistent backups of the Linux guest and its databases.
- Regularly testing restoration procedures.
References
- Azure Managed Disk deletion and recovery limitations
- Microsoft Q&A – Managed disk recovery eligibility conditions
- Microsoft Q&A – Successfully recovered deleted managed disk (June 2026)
- AKS storage and reclaim policies
Since these disks contain critical data, contact Microsoft Support immediately to determine whether backend recovery is still possible. If the recovery window has expired or the disks are ineligible, restoration will require an existing copy of the data.