← Back to context

Comment by anonymousiam

12 hours ago

So the Linux qemu-img tool can still convert VM disk images from .vmdk into formats compatible with all the other hypervisors. Is there something more to the process of VM migration from VMWare than just using this tool (and manually creating the proper virtual hardware profile in the new hypervisor)?

Obviously the types of supported virtual hardware will be different between hypervisor types, but that's usually something that can be dealt with separately.

The problem is how you access the disks that are sitting on a VMware server. VDDK allows you to connect and read them (in raw format, in fact) at a reasonable speed, and also understands sparseness and change block tracking.

Exactly. Either everyone is confused, or something is missing from the article.

> Is there something more to the process of VM migration from VMWare than just using this tool (and manually creating the proper virtual hardware profile in the new hypervisor)?

Yes? Do that for the hundreds of VMs, with a minimal downtime, bonus points for a dumb C-level on your neck.

You would quickly learn what "just converting disk images" is not enough for moving between hypervisors.

Just as an example: last year I moved a company with two Proxmox servers to ESXi. It took a 3 months and another server as a staging point. A minuscule in modern times 1TB+ VM took a whole weekend to be moved. Of course it didn't work in that time.