Free resiliency assessment

Expert guidance, applied to what you already own

← Field Notes index No. 05 Virtualization & DR Filed · July 2026 8 min read

Veeam just gave your VMware exit a safety net.

Backup & Replication 13.1 shipped on July 29 with the two Proxmox VE features everyone was waiting for: replication and Instant Recovery. Here is what works today, what remains on the roadmap, and how to leave VMware without betting the company on the move.

Backup & Replication 13.1 · shipped
01The renewal conversation

The two features everyone was holding out for

For two years, the same conversation has repeated in every VMware renewal review we run. The customer wants off VMware. The renewal quote made the decision for them. Proxmox VE is on the shortlist because the numbers work, and we walked through those numbers in Field Notes No. 01.

Then somebody asks the adult question: what happens to our disaster recovery?

Until this week, the answer was a compromise. Veeam has backed up Proxmox VE since 2024, but two capabilities that VMware shops treat as table stakes were missing: VM replication and Instant Recovery. Veeam's own forums filled up with IT teams and service providers saying they were holding migration projects on those two features alone.

On July 29, Veeam shipped Backup & Replication 13.1. Both features are in it.

02The release

What shipped

Straight from Veeam's What's New documentation, dated the same day.

Replication for Proxmox VE
VM-level replica jobs, including direct host-to-host replication without shared storage and cross-data center replication. Replicas can land on a different storage type than the source.
True replica semantics, rather than a backup copy with extra steps.
Instant Recovery to Proxmox
Boot a VM on Proxmox VE directly from a backup of nearly any platform Veeam protects, then migrate it to production storage. Sources include vSphere, Hyper-V, Nutanix AHV, physical servers via Agent, AWS, Azure and Google Cloud.
The source list is the whole point.
The boring plumbing
VirtIO drivers are injected automatically when you restore a non-Proxmox VE backup onto Proxmox VE, and restored VMs can carry VLAN tags. Veeam calls the VLAN feature out as being for migration scenarios.
That tells you how they expect this to be used.

Veeam's framing in the same document: workload migrations are accelerating, and Veeam is built to move with them. The subtext is loud. The same release added five more hypervisors to the supported list. Veeam is positioning itself as the portability layer for the post-Broadcom data center.

03Scope, plainly

The question everyone will ask

So can I replicate my VMware environment to a Proxmox VE target?

No. Replica jobs in 13.1 run Proxmox VE-to-Proxmox VE only. Cross-hypervisor replicas are absent, and continuous data protection for Proxmox VE is absent too. What you can do across hypervisors is restore-based: back up the VMware VM, then restore it, or Instant Recover it, onto Proxmox VE. The practical difference is the recovery point.

Replica-based DR
A warm standby VM and a recovery point measured in minutes. In 13.1 this requires Proxmox VE on both ends.
Tightest RPO, same-platform only.
Restore-based DR
A recovery point equal to your backup frequency, and a recovery time equal to a restore, or to a boot-from-backup with Instant Recovery.
Works VMware to Proxmox VE today.
The status caveat
Instant Recovery to Proxmox VE ships under Veeam's experimental support status. It works, and Veeam will take your ticket, but nobody should write it into a contractual RTO yet.
We time full restores for the numbers we commit to.
Our publishing rule

The fine print sits next to the promise. Instant Recovery to Proxmox VE is real and it is useful, and it is also labeled experimental in this release. We treat it as the bonus path until it graduates, and we quote recovery times from timed full restores, because that is the number a business can hold us to.

04The play

The DR-first migration

Here is the play this release enables, and it is the safest VMware exit pattern we know how to run.

DR-first migrationVMware today · Proxmox VE on your schedule
01 Now Keep VMware. Point your DR at Proxmox VE.
Production unchanged · recovery restore-based

Production stays exactly where it is. Your Veeam backups gain a second life: on a schedule, restore the important VMs onto a Proxmox VE environment and prove they boot, network, and serve. Recovery stays restore-based, with the recovery points you already run. The part I like most is that every DR test is now a migration rehearsal. Instead of wondering whether your workloads could run on Proxmox VE someday, you watch them run on Proxmox VE this quarter, with a stopwatch out, while VMware still carries production.

No production changeTimed restoresRehearsal data
02 At renewal Cut over on your schedule, not Broadcom's
Same procedure you already rehearsed

When the renewal numbers say go, the cutover is the procedure you have already run: final backup, restore to Proxmox VE, validate, move the network. If an application misbehaves on the new platform, restores run in both directions, so the retreat path is the same tool as the advance path. That symmetry is what makes this a safety net instead of a trapeze.

Rehearsed cutoverTwo-way restoresDocumented rollback
03 After Real replication, Proxmox VE on both sides
13.1's headline feature takes over

After the cutover, your production Proxmox VE cluster replicates, VM by VM, to a second Proxmox VE site, including a different data center on different storage. Now you have warm-standby DR with tight recovery points, on the platform you just moved to, from the backup vendor you already own.

Warm standbyCross-siteMinutes, not hours

If the ransomware recovery playbook taught one lesson, it is that recovery is a sequence you rehearse rather than a product you buy. That lesson stands. This release just makes the rehearsal cheaper and the sequence portable.

05Known limits

The fine print that belongs in your plan

We verified these against the release notes and Veeam's forum statements this week. Plan around them.

The fine printVerified July 2026 · re-check before you build
01 Missing Cloud Connect support at initial release
Confirmed by Veeam product management on the R&D forums

The classic multi-tenant model, replicate to your service provider, has yet to arrive for Proxmox VE. Provider-side replication has to be built co-managed, with one backup server that can see both sites. Watch this one: when Cloud Connect lands, Proxmox VE DRaaS becomes turnkey.

Co-managed workaroundOne backup serverRoadmap item
02 Status Experimental means experimental
Instant Recovery and granular RBAC for Proxmox

Both carry experimental support status in 13.1. Useful in a rehearsal, though never something to write into a contractual recovery time until Veeam graduates them.

Instant RecoveryProxmox VE RBAC
03 Edges Replica jobs have rough spots
From the release notes, known issues

A replication job can fail when a source disk's size sits off the target storage's block-size boundary, and TPM devices ignore per-VM storage mapping for now.

Block-size boundaryTPM mapping
04 Gotcha RHEL 10 family restores need one setting
RHEL, CentOS and Rocky 10

Restored onto Proxmox VE from another hypervisor, these may refuse to boot until you change the VM's processor type from the default x86-64-v2-AES to x86-64-v3. One line in VM settings, easy to script, brutal to discover at 2 a.m.

x86-64-v3ScriptableTest first
05 Floors Version requirements
Plan the upgrade path before the migration path

You need Backup & Replication 13.1 with the version 13.4 Proxmox plug-in, and Proxmox VE 8.2 through 9.2. Keep in mind that v13 dropped vSphere 6.x support entirely, so the longer a legacy environment waits, the fewer doors stay open.

B&R 13.1Plug-in 13.4Proxmox VE 8.2 to 9.2
06 Always Backups are still the foundation
A replica will never be a backup

It happily replicates your ransomware along with your data. Everything in Field Notes No. 04 still applies: at least one copy immutable, isolated from production credentials, and restore-tested. Replication raises availability. Immutability is what gets you back when an attacker is inside.

Immutable copyCredential isolationRestore-tested
A word on Nakivo

Veeam is the tool this piece covers because it is what most of our customers already own, and this release means they can keep it through a Proxmox VE move. If you are shopping rather than standing pat, Nakivo deserves a serious look. It shipped agentless Proxmox VE backup, replication and instant recovery ahead of Veeam, it is an official Proxmox VE solution provider, and in the quotes that cross our desk it tends to land at roughly half the cost of comparable Veeam licensing. It is a good product, and we say that as a shop that quotes both. The takeaway is the same either way: the Proxmox VE DR tooling grew up, and whichever backup vendor you pay probably supports it now.

06Operating model

How we run this at ModernOps

We built our managed recovery practice in two forms, and this release slots into both. In our co-managed model, we operate the customer's existing Veeam estate: we add a Proxmox VE recovery environment, schedule the restore tests, run the stopwatch, and keep the runbook current. In our hosted model, replication and restores flow into Proxmox infrastructure we operate in our Pennsylvania and Arizona data center regions, pre-staged and isolated from the customer's production credentials, with Backup as a Service underneath it on enterprise storage with immutability built in.

The new 13.1 capabilities change our answer to one question we hear weekly. When a mid-market IT director asks whether they can get off VMware without gambling their disaster recovery, the answer is now: keep VMware, DR onto Proxmox VE with us, rehearse until the timings bore you, cut over when your renewal calendar says so, and fail back if you ever need to.

We will run the stopwatch either way.
Disclosure

This section describes ModernOps services. The product capabilities and limitations above are drawn from Veeam's public documentation and forum statements listed in Sources. Verify them against your own versions before you build on them, and expect the details to shift as 13.1 matures.

07Assessment

The 60-second exit-readiness check

The 60-second check
0/5Not yet measured
A "no" is simply the next item on the rehearsal schedule. If the renewal clock is already ticking, this is exactly the work our free Rapid Infrastructure Resiliency Assessment scores.
08Primary reading

Sources

Ryan Beglau
Filed from Conshohocken, PA · July 2026
09Rehearse it

Score the exit before you commit to it.

The free Rapid Infrastructure Resiliency Assessment scores backup, recoverability, immutability and failure tolerance. It is remote-first, carries no obligation, and ends with a prioritized list of what to fix before a renewal deadline makes the decision for you.

Or call 484-429-9328. You'll talk to the engineer who would run the migration.

← All field notes Field Notes · No. 05 · Ryan Beglau · ModernOps, LLC