Multi-Tenant Backup

By George Crump

VergeOS multi-tenant backup: Veeam Backup and Replication protecting isolated Virtual Data Centers

Each VergeOS tenant becomes its own VergeOS Manager and engine ID in Veeam Backup & Replication, so multi-tenant backup follows the Virtual Data Center boundary the platform already enforces.

VergeOS multi-tenant backup follows a boundary the platform already enforces, not one a backup administrator builds on top of it. Most Veeam coverage of the VergeOS integration treats it as a hypervisor story, and multi-tenant behavior gets one line in a feature list near the bottom of the page. The architecture behind that line carries the real argument. Each VergeOS tenant is added to Veeam Backup & Replication as its own VergeOS Manager, with its own engine ID. Jobs, retention, and reporting stay separated per tenant, and that separation follows the Virtual Data Center, the isolation unit VergeOS already enforces for compute, storage, and networking.

Key Takeaways
  • VergeOS multi-tenant backup follows the Virtual Data Center boundary the platform enforces for compute, storage, and networking. Veeam inherits that boundary rather than reconstructing it.
  • Enterprises carry the same tenancy problem service providers do. Production and development, a subsidiary under its own compliance rules, an acquired company, and a regulated department all need the same separation.
  • VergeOS multi-tenant backup means Veeam adds each tenant as its own VergeOS Manager with its own engine ID, so jobs, retention, and reporting stay separated without a backup administrator maintaining the boundary by hand.

The Enterprise Case for Virtual Data Centers

Most readers categorize a Virtual Data Center as a cloud provider feature, and that categorization costs VergeOS an audience that skips past anything with “multi-tenant” in the title. An enterprise runs the same problem under different names. Production and development need real separation, not a naming convention two engineers agreed to follow. A subsidiary carries its own compliance obligations. An acquired company arrives with an infrastructure estate that has to stay intact through an integration that runs two years. Departments charge back their own infrastructure costs, and regulated workloads live under audit scope that cannot spread into the rest of the estate.

VDC 01 Production VMVMVMVM Isolated tenant VDC 02 Subsidiary VMVMVMVM Isolated tenant VDC 03 Dev / Test VMVMVMVM Isolated tenant Shared Physical Infrastructure — VergeOS

VergeOS multi-tenant backup, visualized: three isolated tenants, each running its own VMs, on one shared platform.

Every one of those is a tenancy problem, and enterprises solve it today by buying separate clusters. Separate hardware is the enterprise version of a tenant boundary. It sits idle between peaks, and moving a workload from one cluster to another means a migration project rather than a configuration change. The VDC delivers the same isolation without the second cluster, and the Veeam integration lets the backup design inherit that isolation instead of fighting for it.

The Virtual Data Center as the Isolation Unit

A VDC is a full encapsulation of compute, storage, and networking, the same way a virtual machine encapsulates its own resources, scaled up to data center size. VergeOS describes the result as complete isolation at the infrastructure level, and draws a direct contrast with a VLAN, which segments only the network and operates at the data link layer. A VLAN separates traffic. A VDC separates the infrastructure the traffic runs on.

That distinction carries the whole piece. On most platforms, multi-tenant backup is a discipline the backup administrator imposes on infrastructure that has no idea tenancy exists. Folder hierarchies, job naming conventions, tag policies, and scoped credentials all approximate a boundary the hypervisor never had. The approximation holds until someone edits a job, renames a folder, or reassigns a credential. On VergeOS the boundary already exists at the infrastructure layer before a single backup job gets created, so the backup topology inherits it instead of reconstructing it.

How Veeam Inherits VergeOS Multi-Tenant Backup

VDC 01 Tenant VDC 02 Tenant VDC 03 Tenant VergeOS Manager Engine ID 01 VergeOS Manager Engine ID 02 VergeOS Manager Engine ID 03 Veeam Backup & Replication Job 01 Job 02 Job 03

Each isolated VDC becomes its own VergeOS Manager and Veeam job — no shared path between tenants.

Veeam Backup & Replication 13.1 connects directly to that structure. A VergeOS provider system, or an individual tenant inside it, gets added to Veeam under Virtualization Platforms as its own VergeOS Manager, and each one carries its own engine ID. Jobs point at a specific VergeOS Manager, so a job built against one tenant has no path to the VMs living inside another. Retention policies, schedules, and reporting all follow the same separation, tenant by tenant, without a shared job that a backup administrator has to scope by hand.

One deployment step earns a place in this piece rather than a footnote. The provider’s External network has to extend into each tenant Veeam will protect, so Veeam Workers can reach the VMs they back up. It is a real step in a real deployment plan, and a piece that skips it loses credibility with the reader who tries the integration next.

State the division of labor plainly, since getting it backwards inverts the whole argument. VergeOS enforces the tenant boundary, and Veeam inherits it. Veeam reads a structure that already exists and reports against it, tenant by tenant, a smaller job than enforcing isolation and a more reliable one.

 Most Backup PlatformsVergeOS with Veeam
What separates tenantsFolder structure, job naming, tag policy, scoped credentialsThe Virtual Data Center, enforced at the infrastructure layer
Who enforces the boundaryThe backup administrator, job by jobThe platform, before a backup job exists
What happens when a job is misconfiguredThe boundary can be crossedThe job has no path across it
Where the boundary livesInside the backup productInside the infrastructure the backup product reads

What the Design Buys Operationally

VergeOS multi-tenant backup pays off at restore time. A tenant-scoped restore returns exactly what that tenant owns, and nothing from a neighboring tenant surfaces in the recovery, since the VergeOS Manager backing the job never had a path to that data in the first place. An enterprise running production and development as separate VDCs gets the same guarantee a service provider gets for two customers, without asking a backup administrator to police the line between them every time a job changes.

Tenant isolation answers one restore question. Whether the restored VM opens the way the application left it is a separate one, and it turns on application-consistent backup, not on which VDC the job ran against. VergeOS earns that guarantee through the same Veeam Deployment Toolkit mechanism vSphere and Hyper-V already use, and the tenant boundary has no bearing on it either way.

That is the part worth carrying past this article. A tenant boundary built from folders and naming conventions holds until somebody edits a job. A tenant boundary built into the infrastructure has nothing to edit.

Key Terms
Virtual Data Center (VDC)
A VergeOS construct that encapsulates a dedicated allocation of compute, storage, and networking, delivering isolation at the infrastructure level rather than through network segmentation alone.
VergeOS Manager
The object added under Virtualization Platforms in the Veeam console, representing either a VergeOS provider system or an individual tenant to Veeam.
Engine ID
The identifier tied to each VergeOS Manager in Veeam, unique per tenant, that keeps jobs, retention, and reporting from crossing between tenants.
Veeam Universal Hypervisor API
The Veeam interface through which VergeOS is integrated and shipped as a fully supported hypervisor in Veeam Backup & Replication 13.1.

See the Tenant Model in a Live Console

Live Webinar · September 3, 2:00 PM ET
Veeam’s Role in a VMware Exit
Veeam’s Rick Vanover joins VergeIO’s Dave Vincent and George Crump. Dave adds VergeOS to a Veeam console and walks the tenant model directly.
Register

Read the full capability set, including the multi-tenant deployment step, in the datasheet: Leave VMware. Keep Veeam.

Frequently Asked Questions
Is Virtual Data Center tenancy only useful for service providers?
Enterprises need VergeOS multi-tenant backup as much as service providers do. A service provider separating customers and an enterprise separating production from development, a subsidiary, or a regulated department all need the same boundary. Enterprises solve it today with separate clusters, and the VDC delivers the same isolation on shared hardware.
How does Veeam separate tenant backup jobs on VergeOS?
Each tenant is added to Veeam as its own VergeOS Manager with its own engine ID. Jobs, retention, and reporting all follow that separation, so a job built against one tenant has no path to another tenant’s VMs.
What has to be configured before Veeam can protect a tenant?
The provider’s External network has to extend into each tenant Veeam will protect, so Veeam Workers can reach the VMs inside it. Plan for that step alongside adding the VergeOS Manager.
Does Veeam enforce the tenant boundary, or does VergeOS?
VergeOS enforces it, at the infrastructure layer, before a backup job exists. Veeam reads that structure and reports against it, tenant by tenant, through the VergeOS Manager and engine ID model.

Further Reading

Beyond the VMware Exit Consolidate Every Platform onto VergeOS with Veeam

Most VMware alternatives only promise a home for vSphere workloads. With official VergeOS support in Veeam Backup & Replication 13.1, one restore job moves workloads from vSphere, Hyper-V, Nutanix AHV, AWS, Azure, GCP, and physical servers onto VergeOS. Consolidate onto VergeOS from every platform you run, and retire the rest.
Read More

The Maintenance Line You Can’t Escape

Scott Bartgis stopped paying VMware maintenance and froze Saratoga Casino Holdings at version 7. The license type was never the problem. The renewal curve was. Here is what a VMware perpetual license buys you, what it costs to hold one unsupported in a regulated environment, and the terms that got him back under support.
Read More

Storage Tiering Without the Capacity Tax

Storage tiering was never an array feature. It was a placement decision that lived where the intelligence sat. Lab measurements show a VergeOS tier change acknowledged in 1.3 seconds with the VM still running, and per-node licensing that decides who captures the data reduction: the customer or the vendor's meter.
Read More