Finish the VMware Exit
VergeOS 26.2 moves the workloads that usually stay behind on VMware. Migrations run faster, Microsoft Failover Clustering runs on VergeFS, and a remote site needs only one server with two network connections.
- Faster initial and final transfers shorten cutover
- Clustered SQL Server moves without a dedicated storage array
- Single-server sites run with no core network
Most VMware Exits Stop Short of the Finish
The bulk of the VMs move quickly. The clustered applications and the small remote sites keep a second hypervisor, and a second license bill, running long after the project should have closed.
The Last Workloads Hold the Renewal
Three holdouts usually keep VMware in the data center after the main migration finishes:
- Windows failover clusters that need a shared disk with SCSI-3 Persistent Reservations
- A storage array kept alive only to provide that shared disk
- Remote and retail sites that run on a single server
Each one means two platforms, two consoles, and a VMware renewal paid for a handful of workloads.
A Complete Exit, on a Shorter Schedule
The release goes after each of those holdouts directly.
- Microsoft Failover Clustering runs inside VergeFS
- Faster transfers shorten the cutover window
- One server and two network connections run a remote site
The second hypervisor retires with the rest of the VMware estate.
Three Changes That Close Out the Exit
Each one removes a reason to keep VMware running after the migration.
Faster VMware Migration
Changed block tracking already kept cutover short. VergeOS 26.2 makes both transfers faster.
- Initial full transfer and final transfer finish 120% faster
- Asynchronous I/O buffers speed NBD transfers
- Tags, categories, and a unique serial number arrive with each VM
Mission-Critical Windows Workloads
Microsoft Failover Clustering runs on VergeFS with shared disks.
- One shared disk attaches to every node in the guest cluster
- SCSI-3 Persistent Reservations for cluster arbitration
- VergeFS replaces the dedicated array behind the cluster
Single-Server Distributed Data Centers
Remote sites run the same VergeOS as the main data center.
- One server per site
- Two network connections and no core network
- Both ports serve external and workload traffic
How VergeOS 26.2 Brings the Holdouts Across
A shared disk inside VergeFS serves clustered Windows applications, and a smaller footprint serves the sites at the edge of the network.
Shared Disks Inside VergeFS
An administrator marks a virtual disk as shared and attaches it to each node in the guest cluster. The disk presents proper NAA and WWN identity and supports the SCSI-3 Persistent Reservations that Windows Failover Clustering checks before it builds a cluster. VergeFS aggregates the drives inside the servers into the vSAN behind it, so SQL Server failover instances, clustered file servers, and Cluster Shared Volumes run on the same storage as every other VM.
A Faster Migration Service
The VergeOS VMware migration service uses changed block tracking, so the full copy runs ahead of cutover and the final transfer moves only the blocks that changed. VergeOS 26.2 adds asynchronous I/O buffers to NBD transfers, the transport most mid-sized environments use, and pulls inventory from vCenter faster. Tags and categories arrive with each VM, and each VM receives its own SMBIOS serial number for asset tracking.
Four Steps from VMware to VergeOS 26.2
The migration runs one server at a time on the hardware you already own, and the clustered applications that used to stay behind move last.
Upgrade to 26.2
Systems on VergeOS 26.1.6 or later upgrade directly. VMs move to other nodes as each node updates, one node at a time, so the upgrade is non-disruptive.
Connect to vCenter
Point the migration service at vCenter and pull the inventory. VergeOS 26.2 retrieves it faster, which shortens planning for large batches.
Run the Initial Transfer
The full copy runs ahead of cutover and finishes faster in 26.2. Tags, categories, and a unique serial number come across with each VM.
Cut Over, Clusters Included
Changed block tracking limits the final transfer to changed blocks, which keeps the outage short. Clustered Windows workloads move in the same project, onto shared disks in VergeFS.
A Faster, More Complete VMware Exit
What VergeOS 26.2 changes, line by line.
Today’s release lets IT execute a faster, less disruptive, and more complete VMware exit.Parag Patel, CEO, VergeIO
| Capability | What VergeOS 26.2 does |
|---|---|
| Faster transfers | Initial and final VMware transfers finish 120% faster, with asynchronous I/O for NBD transfers |
| vCenter inventory | Faster retrieval, which shortens planning for large batches |
| Tags and categories | Imported with each VM, with descriptions and one-tag-per-object rules |
| Unique serial numbers | Each VM receives its own SMBIOS serial for asset and CMDB tools |
| Backup continuity | Restores through the oVirt-compatible path keep the original VM UUID, so existing Veeam backup jobs keep running |
| Shared disks | Multi-writer disks with NAA and WWN identity and SCSI-3 Persistent Reservations |
| Failover Clustering | Windows Failover Clustering and Cluster Shared Volumes on VergeFS |
| Single-server sites | One server, two network connections, no core network |
| Session security | Configurable idle timeout for shell and SSH sessions, supporting CIS, DISA STIG, and NIST 800-53 controls |
| Networking | VergeFabric tested at 600 networks, advanced BGP configuration, and Balance-SLB bonding |
| Processor support | Intel Diamond Rapids, Clearwater Forest, and Sierra Forest, and AMD EPYC Turin |
Five Questions Before You Plan the Last Mile
Answer these for your own environment. The list of holdouts usually shrinks to zero.
Which workloads are still on VMware after your main migration waves?
Most teams can name them without looking. Clustered SQL Server, clustered file servers, and a handful of remote sites make up the usual list, and each one is a reason the VMware renewal still shows up.
Does any application depend on a shared disk with SCSI-3 Persistent Reservations?
Windows Failover Clustering requires both before it builds a cluster. On VergeOS 26.2 the shared disk lives inside VergeFS, so the cluster moves with the rest of the estate.
Is a storage array in your data center there only to serve cluster disks?
That array is a refresh cycle and a support contract tied to a few VMs. VergeFS aggregates the drives already inside your servers and provides the shared disk itself.
How long can your largest VMs stay down at cutover?
The final transfer sets the length of the outage. Changed block tracking limits it to the blocks that changed, and VergeOS 26.2 adds asynchronous I/O to NBD transfers, so the final transfer finishes 120% faster and cutover gets even shorter.
How many of your sites run on a single server?
Retail, branch, and industrial locations often do. A VergeOS 26.2 site needs one server with two network connections and no core network, and IT manages it from the same platform as the main data center.
See VergeOS 26.2 Run Live on October 22
Aaron Richman and Dave Vincent explain and demonstrate the release, then take your hardest workload questions.
Bring the Workload You Think Has to Stay on VMware
The session walks through the migration service and its use of changed block tracking, shows a Windows failover cluster running on VergeFS with shared disks and SCSI-3 Persistent Reservations, and covers the single-server sites at the edge of the network. Dave runs a working VergeOS 26.2 environment live from his lab.
- Why VMware exits stall at the last workloads
- Faster initial and final transfers at cutover
- Microsoft Failover Clustering on VergeFS
- Single-server sites with two network connections
- A live VergeOS 26.2 lab demo, then Q&A
Go Deeper on the VMware Exit
VergeOS 26.2 White Paper
The full story behind the release and the workloads it brings across.
Get the white paperBeyond the Hypervisor Swap
The VergeOS 2026 architecture, from the single code base to virtual data centers.
Get the white paperVMware Exit ROI and TCO Analysis
A sample report that prices the exit, including the cost of running two platforms.
See the sample reportCommon Questions
What is new in VergeOS 26.2?
Faster VMware migration transfers, support for Microsoft Failover Clustering on VergeFS with shared disks and SCSI-3 Persistent Reservations, and single-server distributed data centers that need one server with two network connections and no core network.
How much faster is migration in 26.2?
Both the initial full backup and the final transfer finish 120% faster than on VergeOS 26.1, 2.2 times the transfer speed, in VergeIO lab testing over NBD. The gain comes from asynchronous I/O in the network block device (NBD) transfers, and changed block tracking still limits the final transfer to the blocks that changed.
Which clustered workloads does VergeOS 26.2 support?
Windows Failover Clustering and Cluster Shared Volumes, which covers SQL Server failover instances and clustered file servers. The shared disk presents NAA and WWN identity and supports SCSI-3 Persistent Reservations.
Where does the shared disk for a Windows cluster live?
Inside VergeFS, the VergeOS internal storage capability. VergeFS aggregates the commodity drives inside the servers into a vSAN, and the shared disk sits there with every other VM disk.
What does a single-server site need?
One server with two network connections. The site needs no core network, and both ports serve external and workload traffic.
Do VMware tags and categories come across?
Yes. Tags and categories import with each VM, along with their descriptions and one-tag-per-object rules. Each VM also receives its own SMBIOS serial number, so asset management tools track it as an individual device.
How do I upgrade to VergeOS 26.2?
Systems on VergeOS 26.1.6 or later upgrade directly. VMs move to other nodes as each node updates, one node at a time, so workloads keep running through the upgrade.
Retire the Second Platform
VergeOS 26.2 brings the clustered applications and the single-server sites across with the rest of the estate. Watch it run live on October 22, or book time with an engineer to walk through your own holdouts.