Backup and Restore

Policy-driven backups

VMease manages backup and restore through central policies rather than ad hoc per-VM configuration. Operators protect virtual machines across a cluster through one workflow while still using Proxmox Backup Server reliability and efficiency.

This model avoids maintaining separate backup jobs in Proxmox and keeps scheduling, queueing, execution, verification, retention, and restore controls in one place.

Integrated with Proxmox Backup Server

VMease stores backups for protected virtual machines in Proxmox Backup Server (PBS). PBS provides incremental backups, chunk-based deduplication, compression, verification, and retention support.

VMease orchestrates delivery to PBS by selecting backup-capable storage, starting backup tasks, tracking progress, recording restore points, and validating restore points before they are exposed for recovery.

Each restore point is treated as a complete recovery option. Users can select any available restore point and restore the VM to that point in time.

After the first backup, only changed or newly created chunks are transferred, while each restore point still behaves as a full recovery point.

Compression and deduplication improve storage efficiency and make longer retention windows more practical.

Encrypted backup capability

Proxmox Backup Server supports authenticated encryption, including AES-256-GCM. This protects backup data when repositories are sensitive or located outside the primary compute cluster.

Within VMease, encryption operates alongside central policies, controlled backup workflows, restore-point verification, and auditable task processing. This keeps data flow and storage consistent with standard Proxmox VE and PBS operating models.

Centralised backup policy and scheduling

Backup protection in VMease is policy-based. Schedules are validated before execution, next run times are calculated centrally, and orchestration is controlled through one platform workflow.

VMease uses cron-style schedules and does not depend on separately maintained Proxmox backup jobs. This allows VMease to apply platform-level queueing, prioritisation, and concurrency controls.

• Which virtual machines are protected
• Which backup-capable storage pool is used
• The backup schedule
• Retention quantities for intraday, daily, weekly, monthly, and yearly restore points

Queue-based orchestration across the cluster

Backups are queue-driven rather than started simultaneously. This helps keep cluster and PBS performance stable and reduces operational risk during busy periods.

Scheduled backups are queued at normal priority, while manually requested backups are queued with higher priority and still respect safety controls.

• Manual versus scheduled backup priority
• VM importance
• Queued time
• Maximum concurrent backups per cluster
• Maximum concurrent backups per host
• Maximum concurrent backups per PBS server
• Retry limits and retry delay

Backup execution lifecycle

After a queued backup starts, VMease monitors the task to completion. On success, VMease inspects task output, identifies the backup snapshot, triggers verification, and records backup and restore-point metadata.

Restore points are tracked after completion and validated against PBS verification status. Customer-facing backup job records are only published once restore points are verified.

VMease and Proxmox do not require a long-lived VM disk snapshot chain with post-backup consolidation for this workflow. The backup process uses QEMU-level changed-block detection with chunked incremental transfer and deduplicated storage.

This avoids merge-back behaviour often associated with snapshot-chain consolidation and helps reduce performance disruption on busy workloads.

For Windows VMs, VSS (Volume Shadow Copy Service) is used for application-consistent backups. Brief consistency pauses can still occur, but there is no long snapshot chain to merge afterward.

Retention and pruning

Retention is enforced through VMease-managed PBS pruning operations. As restore points age out, VMease marks them and removes related backup job records so the interface only shows recoverable points.

VMease includes a backup policy visualiser that shows a restore-point timeline preview before a policy is put into operation. Operators can validate short-term granularity and long-term retention balance before applying changes.

The visual model makes over-retention and under-retention easier to identify and supports clearer planning conversations between administrators and customers.

• Intraday restore points
• Daily restore points
• Weekly restore points
• Monthly restore points
• Yearly restore points
• Operators can validate short-term recovery granularity
• Customers can quickly see how far back restore is available
• Retention intent can be reviewed visually before policy application

Tenant-aware restore

Restoring with VMease is more than a direct restore call. The workflow verifies restore-point validity, account ownership, compute quota availability, compute-to-storage placement alignment, and that the restore point has not been pruned.

VLAN-backed and VNET-backed interfaces are resolved to the correct bridge and network parameters before restore submission. Restored virtual machines can be created powered on or powered off based on request settings.

What this means for customers

• Policy-based VM protection with restore-point pattern visualisation
• Manual backup on demand
• Controlled scheduling across the cluster
• PBS-backed incremental and deduplicated storage
• Restore-point verification
• Retention and pruning controls
• Full VM restore into selected compute and storage context
• Cleanup handling for partial failures
• Reduced operational risk compared with ad hoc backup jobs

The result is a backup system designed for multi-tenant private cloud operation. It is straightforward for customers to use and built on a policy-driven VMease workflow with Proxmox VE and Proxmox Backup Server infrastructure components.