RE: Migrating a 10TB VM from ESXi to PVE/Ceph

[Lund Svein-Erik <sel@selund.se>]

NFS is the easiest way to solve this. 
1) Connect nfs storage to vmware (/<export>/vmware)
2) Move vm in vmware to nfs storage
3) Remove any vmware snapshots
4) Mount nfs storage on proxmox (/<export>/proxmox)
5) Create a identical vm on proxmox with the same sized vmdk files as storage, do not start vm
6) remove vmdk files in /<export>/proxmox/
7) hardlink vmdk files from /<export>/vmware/ to corresponding location in /<export>/proxmox
8) Stop vm on vmware
9) Start vm on proxmox
10) MIgrate vm to permanent storage online. 

The downtime in this move will be the time you stop the vm in vmware to you start the vm back up in proxmox. 
Where and what type of nfs storage you use isn't really important as long as both proxmox and vmware can access the same export, and that the storage supports hardlinks. You might need to create the hardlinks on the nfs server it self. 

--
Svein-Erik 

 ---------------------------

From: "Fabien Sirjean" <fsirjean@eddie.fdn.fr>
To: "undefined" <pve-user@lists.proxmox.com>
Sent: Monday, September 7, 2026 9:08 AM
Subject: Migrating a 10TB VM from ESXi to PVE/Ceph

Dear all,
<br/>
<br/>After successfully migrating most of our infrastructure (300+ VMs) from 
<br/>VMware ESXi to PVE, we're struggling with one last critical VM: our 
<br/>BlueMind/Cyrus mail server (~800 users), with a 10TB virtual disk.
<br/>
<br/>Our setup is:
<br/>
<br/>Source: ESXi → Dell EMC VPLEX (FC) → Dell Unity
<br/>Target: Proxmox VE → Ceph RBD (dedicated cluster)
<br/>
<br/>We tried the Proxmox ESXi importer, but I/O performance degraded so much 
<br/>during the live import that the mail service could not sustain its 
<br/>normal workload.
<br/>We then stopped the service to complete the migration, but the import 
<br/>eventually failed after almost 3 days, forcing us to start over.
<br/>
<br/>Application-level migration/synchronization is not an option, so we're 
<br/>looking for a block/storage-level approach with minimal downtime.
<br/>
<br/>Since the datastore ultimately resides on our FC SAN, could we bypass 
<br/>ESXi entirely and temporarily present the relevant VPLEX/Unity LUN to a 
<br/>PVE or migration host, mount/access the VMFS datastore read-only, and 
<br/>copy/convert the VMDK directly to Ceph RBD?
<br/>Even better, is there a practical way to perform the bulk copy while the 
<br/>VM is running, then only synchronize the changed blocks during the final 
<br/>cutover?
<br/>
<br/>Any advice or experience with direct VMFS access, VPLEX/Unity 
<br/>snapshots/clones, or other approaches for this kind of migration would 
<br/>be greatly appreciated.
<br/>
<br/>Thanks!
<br/>