What it does: Converge takes an existing standalone vSphere environment and turns the whole vCenter, with all of its clusters, into a VCF 9.1 management domain in place.
Depends on: An eligible vSphere 8.0 Update 3 environment, a deployed VCF Installer appliance, and free capacity for the new NSX and management appliances. NSX, SDDC Manager, VCF Operations and Fleet Manager stay mandatory in any VCF 9.1 deployment, so converge deploys whichever of them you do not already run.
- Converge reuses an existing vCenter and its ESX hosts. Your running virtual machines stay in place, and downtime is minimal.
- Source floor for VCF 9.1 converge is vSphere 8.0 Update 3a through 8.0 Update 3i. Version 8.0 Update 3j is import only, not converge.
- An existing NSX Manager attached to the vCenter blocks converge. NSX is deployed fresh during the conversion, as one node or as a three node cluster.
- Minimum host count for a management domain conversion is four, with vSAN or supported external storage on the cluster.
- Each NSX Manager uses the Medium form factor, six vCPU and twenty four GB of memory. SDDC Manager and VCF Operations are sized by the installer.
- Key steps: check eligibility, remediate versions, prepare records, run the VCF Installer convert workflow, deploy NSX and the management components, then validate.
You will convert a standalone vSphere environment into a VCF 9.1 management domain without rebuilding it. You start with a vCenter that already runs your workloads, and you finish with that same vCenter managed by SDDC Manager and VCF Operations, with NSX deployed and the fleet under lifecycle control. Converge suits you when no VCF fleet exists yet, or when you want an additional management domain built from hardware you already own. If a fleet already exists and you only want to add the vCenter as a workload domain, use import instead.
Before you touch the installer, get the source environment to a supported state and stage every network record. Read the VCF 9.1 network, DNS and NTP setup guide for the record set this process expects, and confirm the appliance from the VCF Installer appliance deployment guide is already running on a reachable management network.
Prerequisites
Confirm each item below on the source environment before you start. A missed prerequisite is the most common cause of a stalled conversion.
| Item | Requirement | Notes |
|---|---|---|
| vCenter and ESX version | vSphere 8.0 Update 3a to 8.0 Update 3i | Version 8.0 Update 3j is import only |
| Host count | Four or more in the cluster | Minimum for a management domain conversion |
| Storage | vSAN or supported external storage | vSAN gives the full stack |
| NSX | None attached to the vCenter | Fresh NSX is deployed during converge |
| Enhanced Linked Mode | Deactivated | Not supported in VCF 9.1 |
| Cluster lifecycle | vLCM image | Baselines are not supported |
| VCF Installer | Deployed and reachable | Drives the whole workflow |
Step 1, confirm the source environment is eligible
Match your current versions against the VCF 9.1 converge floor first, because the installer refuses an out of band source.
- Log in to the source vSphere Client as an administrator.
- Open Menu, then Administration, then Deployment, and note the vCenter build number.
- Confirm vCenter and ESX sit between vSphere 8.0 Update 3a and 8.0 Update 3i.
- Confirm no cluster runs vSphere 8.0 Update 3j, which converge does not accept.
- Confirm the target cluster has at least four hosts with vSAN or supported external storage.
- Record every cluster name, since converge brings all clusters into one management domain.
Step 2, remediate versions and cluster images
Fix anything outside the supported band, and move each cluster off baselines, since VCF 9.1 manages hosts with images only.
- In the vSphere Client, open each cluster, then Updates, then Image.
- Click Convert to an Image if the cluster still uses baselines.
- Select one existing host as the image source and click Save.
- Patch vCenter and ESX to a supported 8.0 Update 3 level if any host is below 8.0 Update 3a.
- Re run the cluster compliance check and confirm every host reports compliant.
Step 3, deactivate Enhanced Linked Mode
Enhanced Linked Mode is not supported in VCF 9.1, so break it before you convert.
- Confirm whether the vCenter joins Enhanced Linked Mode under Administration, then System Configuration.
- Follow the documented deactivation procedure to isolate the vCenter without downtime.
- Wait for the replication partners to drop, then confirm the vCenter is standalone.
Step 4, prepare network, DNS and NTP records
Stage forward and reverse records for every new appliance, because the installer prechecks both directions.
- Reserve management IP addresses for the NSX Manager or managers, SDDC Manager and VCF Operations.
- Create a forward A record for each new appliance FQDN.
- Create a matching reverse PTR record for each IP address.
- Point every appliance at the same NTP source used by the existing vCenter.
- Verify forward and reverse lookups resolve from the VCF Installer appliance.
Step 5, reach the VCF Installer appliance
You drive converge from the installer, so confirm it is running and reachable on the management network.
- Open a browser to the VCF Installer UI at its FQDN.
- Log in with the installer administrator account.
- Confirm the appliance resolves the source vCenter FQDN.
- Confirm outbound access to the software depot for any binaries the conversion pulls, as covered in the VCF Management Services and license server guide.
Step 6, start the convert workflow
Pick the path that reuses an existing vCenter rather than a greenfield management domain build.
- On the VCF Installer home page, click Deploy.
- Choose the option to convert or reuse an existing vCenter Server, not a new deployment.
- Accept the end user license agreement and click Next.
- On the General information page, enter a name for the management domain.
- Select whether this converge starts a new fleet or joins an existing fleet, then click Next.
Step 7, enter existing vCenter and host details
Give the installer access to the source vCenter so it can inventory clusters and hosts.
- On the vCenter Server page, enter the source vCenter FQDN.
- Enter the single sign on administrator account and password.
- Confirm SSH is enabled on the vCenter, which the converge workflow requires.
- Click Connect and wait for the installer to list the discovered clusters.
- Confirm all clusters and hosts appear, then click Next.
Step 8, configure the new NSX deployment
Converge cannot reuse an attached NSX, so define a fresh NSX deployment here.
- On the NSX page, choose a deployment model of one node for a lab or three nodes for production.
- For a three node cluster, enter the FQDN and IP for each NSX Manager and a cluster virtual IP.
- Select the Medium form factor for a management domain, which is six vCPU and twenty four GB of memory per manager.
- Enter the NSX admin and audit passwords.
- Enter the overlay and uplink details for the transport nodes, then click Next.
Step 9, configure VCF Operations and SDDC Manager
Supply details for the mandatory management components that converge deploys alongside NSX.
- On the VCF Operations page, enter the FQDN and IP for the VCF Operations node.
- Set the VCF Operations administrator password.
- Enter the SDDC Manager FQDN, IP and password.
- Accept the installer sized appliance form factors unless you have a documented reason to change them.
- Review the license or subscription assignment for the new fleet, then click Next.
Step 10, run validation and start the conversion
Let the installer precheck the plan, fix any flagged item, then commit.
- Click Validate to run prechecks against DNS, NTP, versions, storage and IP availability.
- Read each warning and error, then correct the source or the records where flagged.
- Click Retry until validation reports success across every check.
- On the Review page, confirm every value, then click Deploy to start the conversion.
- Keep the browser open and watch the task list until every subtask reports success.
NSX Manager sizing for converge
| Deployment model | Node count | Form factor | Use |
|---|---|---|---|
| Single node | 1 | Medium, six vCPU and 24 GB | Lab and proof of concept |
| High availability | 3 plus a cluster VIP | Medium, six vCPU and 24 GB each | Production management domain |
Convert wizard fields and example values
| Field | Example value | Wizard page |
|---|---|---|
| Management domain name | mgmt_domain_01 | General information |
| Source vCenter FQDN | vcenter.corp.example.com | vCenter Server |
| Single sign on admin | administrator@vsphere.local | vCenter Server |
| NSX deployment model | Three nodes | NSX |
| NSX cluster VIP | nsxvip.corp.example.com | NSX |
| VCF Operations FQDN | ops.corp.example.com | VCF Operations |
| SDDC Manager FQDN | sddc.corp.example.com | VCF Operations |
Verify the conversion
Confirm the converge finished cleanly before you hand the environment to operations.
- In the VCF Installer, confirm all conversion subtasks show a green success state.
- Log in to SDDC Manager and confirm the new management domain and its clusters appear.
- Open VCF Operations and confirm the vCenter, NSX and every host show in inventory.
- In NSX Manager, confirm the manager cluster status reads Stable.
- In VCF Operations, confirm the fleet shows the licensing subscription and core allocation.
Common errors and fixes
Precheck fails on vSphere version. Converge rejects a source outside vSphere 8.0 Update 3a to 8.0 Update 3i. Patch vCenter and ESX into that band, and avoid 8.0 Update 3j, which is import only. Re run the validation after patching.
Converge refuses because an NSX Manager is attached. An existing NSX cannot be converged. Detach or remove the existing NSX from the vCenter, then let the installer deploy a fresh NSX during the conversion.
Validation flags Enhanced Linked Mode. VCF 9.1 does not support Enhanced Linked Mode. Deactivate it on the vCenter using the documented no downtime procedure, then re run validation.
Reverse lookup missing for a new appliance. A precheck fails when a PTR record is absent for an NSX, SDDC Manager or VCF Operations IP address. Add the missing PTR record, confirm it resolves, then retry.
Common questions
Downtime for running virtual machines during a converge
Running virtual machines stay online during converge. Conversion mainly deploys and configures missing components. Any host or component you upgrade to meet the prerequisites needs its own maintenance window, so use maintenance mode and vMotion to move workloads while you remediate.
Whether an existing NSX deployment can be reused
No. Converge deploys a new NSX instance and does not adopt an attached one. Plan to remove the existing NSX first, and size the fresh deployment as a single node for a lab or a three node cluster for production.
Storage required on the management cluster
vSAN gives the full SDDC stack and is the common choice for a management domain. VCF 9.1 also supports external principal storage such as Fibre Channel and NFS, so a cluster on supported external storage can also converge.
Difference between converge and import
Converge builds a management domain from a vCenter when no fleet exists yet, or adds another management domain. Import adds an existing vCenter as a workload domain into a fleet that already runs. Import has a lower version floor and does not deploy a new management stack.
What happens to Enhanced Linked Mode
Enhanced Linked Mode must be deactivated before converge. VCF Operations takes over the cross vCenter visibility that Enhanced Linked Mode used to provide, so you do not lose a single pane view after conversion.
References
- Broadcom TechDocs, Converging Existing Virtual Infrastructure to a VCF or vSphere Foundation Platform
- VCF Blog, How to Converge a VMware vSphere Environment to VMware Cloud Foundation
- VCF Blog, Converging vSphere to VCF, Top 10 Questions Answered
- William Lam, VCF 9.1 vSphere Back in Time Support for Upgrade, Converge and Import
← Previous: How to Configure a Stretched Cluster in VCF 9.1
Complete Guide: VCF 9.1 Deployment
Next: How to Import a Brownfield vSphere, vSAN and NSX Environment →

