Excited to share a new project with the Veeam community: Moov
If you run Veeam and you're staring down a move off VMware, you already know the awkward part: the backups are right there, your target hypervisor is ready, but getting the workloads across is still slow, manual, and one VM at a time.
Moov closes that gap. It uses your Veeam Backup & Replication restore points as the source and brings VMs up directly on KVM platforms, Proxmox VE, oVirt / OLVM / RHV, and HPE VM Essentials, without ever touching the source vSphere environment during the migration.
It's open to the community on GitHub (VeeamHub), currently in beta.
https://github.com/VeeamHub/moov
Why another migration tool?
Full-restore workflows work, but at scale is complex: extended downtime windows, manual VirtIO driver injection, and moving machines serially while you make some manual tasks. Moov was built for the fleet case, hundreds of VMs, in waves, with the boring parts automated.
Three ideas drive it:
- Veeam is the single source of truth. Every VM comes from a Veeam restore point. Moov never connects to vCenter or ESXi. Your production VMware environment stays untouched.
- The guest should just boot. Moov injects VirtIO drivers and installs the QEMU Guest Agent automatically, so the migrated VM comes up with working storage and networking and reports status to the hypervisor.
- The network should survive the move. Static IPs and multi-NIC layouts are carried over on both Linux and Windows or reassigned per NIC from the console.
How it works
Moov ships as appliance plus lightweight, disposable helper VMs you deploy on the target hypervisor:
1. The Pre-flight Inspector scans your Veeam backups and classifies each VM, OS, CPU, RAM, disks, network interfaces, and VirtIO readiness, so you know exactly what you're moving.
2. You plan a wave in the web console and pick a target and migration mode per VM.
3. The Core publishes the restore point from Veeam and hands the work to a helper, which serves the disk over NBD, injects drivers and the agent, and drives the boot.
4. Progress streams live to the console; the VM boots, the guest agent comes online, and (for Instant Migration) the disk pivots to local storage with no downtime.
Two migration modes
- Instant Migration (IVMR), in minutes. The VM boots directly from the backup over NBD, then pivots to local storage with no downtime. Best for critical workloads.
- Cold Migration (Convert and Boot), in hours. Full materialization before boot, for guests that need conversion.
Scale is just "add helpers"
Each concurrent VM needs roughly 1 vCPU and 2 GiB of RAM on the helper, so an 8 vCPU / 16 GiB helper moves 8 VMs at once. Total throughput is the sum across all helpers, deploy more helpers (they can even live on different target hypervisors) and you move more VMs in parallel. Moov also respects the real limits of your Veeam Mount Server so you don't saturate the source.
Getting started
1. Deploy the appliance image and complete the first-boot wizard.
2. Open the web console and sign in.
3. Connect your Veeam server and at least one target hypervisor (Proxmox can bootstrap its API token automatically).
4. Set the per-hypervisor defaults, deploy a helper, run a pre-flight scan.
5. Plan a wave and watch it run.
The appliance is self-contained and air-gap friendly, no internet access is needed during a migration.
Where to find it
Moov is on GitHub under VeeamHub: https://github.com/VeeamHub/moov with the appliance images. It's beta today, so if you're mid-migration, test it and tell us what's missing. That feedback is exactly what shapes the next release.
