Deploy VCF 9.0.x Management Domain Using Holodeck 9.0.2
Objectives
- Choose the New-HoloDeckInstance parameters for a management-only VCF 9.0.2 deployment (Version, InstanceID, -ManagementOnly, vSANMode, CIDR, VLANRangeStart, DNSDomain, DepotType, optional NsxEdgeClusterMgmtDomain, DeployVcfAutomation, DeploySupervisorMgmtDomain) and understand the parameter sets
- Run the deployment from pwsh on the HoloRouter, answer the interactive pre-check prompts (datastore, trunk port group, build token or offline depot) and read the pre-check output
- Follow the deployment phases (networking setup on the HoloRouter, nested ESX build, VCF Installer deployment, bundle download, management domain bring-up) and use idempotency to resume after a failure
- Validate the management domain through the documented endpoints: VCF Installer, SDDC Manager, management vCenter, VCF Operations and management NSX
- Record the instance with Get-HoloDeckInstance, power it off and on with Stop-/Start-HoloDeckInstance, and take the post-deployment snapshot set
Prerequisites
holodeck-01 complete: HoloRouter deployed with SSH, binaries staged in /holodeck-runtime/bin/9.0.2.0/, a configuration created with New-HoloDeckConfig, depot method decided, baseline snapshot taken.
Prior labs: holodeck-01
Required skills:
- Holodeck configuration handling (New/Get/Import-HoloDeckConfig)
- VCF 9.0 management domain components and bring-up concepts
- vSphere Client and NSX Manager navigation
Lab Environment
HoloRouter on the External and Trunk port groups. New-HoloDeckInstance -ManagementOnly builds 4 nested ESX hosts (esx-01a..esx-04a) as vSAN ready nodes on the Site A VLANs of 10.1.0.0/20, deploys the VCF Installer (vcfinstaller-a), and the installer brings up the management domain: SDDC Manager (sddcmanager-a), management vCenter (vc-mgmt-a), management NSX (nsx-mgmt-a), VCF Operations (ops-a), optionally VCF Automation (auto-a), an NSX Edge cluster and a Supervisor.
Mermaid source
graph TB HR[HoloRouter: DNS/DHCP/NTP/BGP/proxy, Site A VLANs 0,10-25 on 10.1.0.0/20] --> E1[esx-01a] & E2[esx-02a] & E3[esx-03a] & E4[esx-04a] HR --> VI[VCF Installer vcfinstaller-a] VI -->|bring-up| SM[SDDC Manager sddcmanager-a] VI --> VC[vCenter vc-mgmt-a] VI --> NSX[NSX nsx-mgmt-a] VI --> OPS[VCF Operations ops-a] VI -.optional.-> AUTO[VCF Automation auto-a]
IP Addressing
| Network | Purpose | VLAN |
|---|---|---|
10.1.0.0/20 | Default Holodeck Site A master CIDR, subdivided by New-HoloDeckNetworkConfig into the Site A subnets (management, vMotion, vSAN, host overlay, edge uplinks, and others) | VLANs 0 and 10-25 (Site A default) |
10.2.0.0/20 | Default Holodeck Site B master CIDR (dual-site only) | VLANs 40-58 (Site B default) |
<external management network> | HoloRouter external management IP (DHCP or static, set at OVA deployment); the only external IP a Holodeck environment needs | External port group |
Credentials
| System | Username | Password |
|---|---|---|
| HoloRouter (SSH and pwsh) | root | Set in the OVA Additional settings at deployment |
| HoloRouter webtop | (no authentication in 9.0.2) | http://<holorouter-ip>:30000; the docs warn not to expose it externally |
| VCF Installer | admin@local | Holodeck documented default (Post Deployment table) |
| VCF Operations | admin | Holodeck documented default (Post Deployment table) |
| Management vCenter | administrator@vsphere.local | Holodeck documented default (Post Deployment table) |
| SDDC Manager | administrator@vsphere.local | Holodeck documented default (Post Deployment table) |
| Management NSX | admin | Holodeck documented default (Post Deployment table) |
| Nested ESX hosts | root | Holodeck documented default (Post Deployment table) |
Tasks
Task 1 Choose the deployment parameters
ManageabilityNew-HoloDeckInstance has four parameter sets (VVF, management-only, full stack, and the dual-site sequence). Reading the syntax carefully is the difference between a 4-5 hour deployment and a 12+ hour one, or one that does not fit the host.
SSH to the HoloRouter, run pwsh and load your config: Import-HoloDeckConfig -ConfigID <id>. Confirm with $config.
Read the parameter sets: Get-Help New-HoloDeckInstance. Mandatory for management-only: -Version, -InstanceID, -ManagementOnly. Optional: -CIDR (/20, default 10.1.0.0/20 for Site A), -vSANMode (ESA or OSA, default OSA), -NsxEdgeClusterMgmtDomain, -DeployVcfAutomation (9.x only), -DeploySupervisorMgmtDomain (9.x only), -LogLevel, -ProvisionOnly, -VLANRangeStart (default 10 for Site A; at least 16 consecutive VLAN IDs), -DNSDomain (default vcf.lab), -Site (default a), -DepotType (Online or Offline; 9.x only), -DeveloperMode.
Decide the values: -Version 9.0.2.0; -InstanceID holo (a prefix applied to every nested VM name); -ManagementOnly; -vSANMode OSA (or ESA if the physical layer is vSAN); defaults for CIDR, VLANs and DNS domain unless your network conflicts with 10.1.0.0/20 or VLANs 10-25; -DepotType Online or Offline per holodeck-01. Leave -NsxEdgeClusterMgmtDomain, -DeployVcfAutomation and -DeploySupervisorMgmtDomain off for the first run unless the host meets the larger requirement rows.
If you use a custom CIDR or VLAN range, note the documentation rule: the CIDR must be a /20, and the same custom CIDR and VLAN start must be given to New-HoloDeckInstance (and, for dual site, to New-HoloDeckNetworkConfig first) or the defaults overwrite them.
Record the expected duration from the documented table: -ManagementOnly about 4-5 hours for 9.0.0.0 (longer with an edge cluster or Automation), and plan to leave the SSH session attached or run the deployment in a terminal multiplexer so the output stays visible.
Optional comparison for the design notebook: -ProvisionOnly deploys only the nested ESX hosts and the VCF Installer and writes JSON API specs to /holodeck-runtime/specs/ so you can walk through a greenfield VCF deployment manually; -VVF deploys vSphere Foundation instead of VCF. Note both for later labs.
Validation Gate
Check: Config imported; parameter sets understood; final command composed with justified values; duration planned.
Expected: A deployment command that fits the host and the lab goal.
Common Errors
Task 2 Run the deployment and pass the pre-checks
ManageabilityThe deployment is one cmdlet. Its first phase is the built-in pre-check, interactive for the datastore, the trunk port group and the depot credentials. Every documented pre-check error maps to a holodeck-01 step.
Run the composed command, for example New-HoloDeckInstance -Version 9.0.2.0 -InstanceID holo -ManagementOnly -vSANMode OSA -DepotType Online. Watch the pre-checks: binaries in /holodeck-runtime/bin/9.0.2.0/, target reachable, trunk port group present, NTP running.
New-HoloDeckInstance -Version 9.0.2.0 -InstanceID holo -ManagementOnly -vSANMode OSA -DepotType OnlineAt the interactive prompts select the datastore for the nested VMs and the trunk port group on which the nested ESX hosts and the VCF Installer are deployed.
Online depot: provide the Broadcom support site build token when prompted; answer the proxy questions if your network needs one. Offline depot: provide the offline depot IP, port, protocol and credentials instead.
Observe 'Networking setup completed on Holorouter': New-HoloDeckInstance configures the HoloRouter networking for the instance (Site A VLANs, subnets, DNS records, BGP) before any nested VM is created. In a second SSH session, in pwsh after importing the config, inspect it: Get-HoloDeckSubnet -Site a | ft -AutoSize and Get-HoloDeckAppNetwork -Site a.
Validation Gate
Check: Pre-checks passed, depot configured, HoloRouter networking set up, subnet and appliance network listings captured.
Expected: Deployment initiated with no red errors.
Common Errors
Task 3 Follow the deployment phases and use idempotency
RecoverabilityAfter the pre-checks the cmdlet builds the nested ESX hosts, deploys the VCF Installer, downloads the VCF bundles and drives the management domain bring-up through the installer. Holodeck stores the deployment state on the HoloRouter, so a failed run is resumed by re-running the same command.
Watch 'Nested ESX hosts getting built'. The FAQ explains the repeated message 'SSH service on <IP> is not responding, Trying after 10 seconds' with 'WARNING: Host key is not being verified since Force switch is used': the hosts are still booting; wait 10-15 minutes. If it persists beyond 15 minutes, confirm the ESX VMs are powered on and reachable and try ssh root@<nested-host-ip> yourself.
Watch 'VCF Installer being deployed' and 'VCF bundles being downloaded'. Open https://vcfinstaller-a.site-a.vcf.lab (admin@local) from the webtop or a console that uses the HoloRouter as DNS, and follow the bundle download and bring-up progress in the installer UI.
Watch 'Management Domain deployment initiated through VCF Installer' until 'Deployment completed successfully'. For a management-only run the documentation's indicative time is 4-5 hours.
If the run stops with a red error: fix the cause and re-run exactly the same New-HoloDeckInstance command. Holodeck's idempotency picks up where the code failed instead of restarting the whole deployment. Only if the state is corrupted should you clear it with Reset-HoloDeckState (cannot be undone) and start again. If validation or deployment fails inside the installer, check /var/log/vmware/vcf/domainmanager/domainmanager.log on the VCF Installer appliance (VCF 9.0.1 documentation, Log Files and Locations).
Validation Gate
Check: All phases observed; installer UI reached; deployment completed, or resumed correctly after a failure.
Expected: Management domain bring-up complete.
Common Errors
Task 4 Validate SDDC Manager, vCenter and VCF Operations
AvailabilityThe documentation's Post Deployment table is your acceptance checklist: every management component has a documented FQDN and account.
From the webtop (http://<holorouter-ip>:30000) or a console using the HoloRouter external IP as DNS server, open https://vc-mgmt-a.site-a.vcf.lab and log in as administrator@vsphere.local. Confirm the management cluster with four hosts esx-01a..esx-04a, vSAN healthy, and the management VMs (SDDC Manager, NSX Manager nodes, VCF Operations, fleet management appliance) running.
Open https://sddcmanager-a.site-a.vcf.lab and log in as administrator@vsphere.local. Note the deprecation notice for the SDDC Manager UI and check Inventory > Workload Domains shows the management domain as Active.
Open https://ops-a.site-a.vcf.lab (admin or admin@local) and verify Inventory lists the VCF Instance with the management domain, and Fleet Management > Lifecycle lists the fleet management components.
If you deployed with -DeployVcfAutomation, open https://auto-a.site-a.vcf.lab (organization system, user admin).
Open https://vcfinstaller-a.site-a.vcf.lab and confirm the deployment report shows the bring-up complete. In VCF 9.0 the installer remains as an appliance after bring-up.
From pwsh on the HoloRouter: Get-HoloDeckInstance -InstanceID holo. Record the deployed components, networking and deployment state it returns from output.json.
Validation Gate
Check: vCenter, SDDC Manager, VCF Operations (and Automation if deployed) and the installer validated; Get-HoloDeckInstance output recorded.
Expected: Management domain healthy and documented.
Common Errors
Task 5 Validate NSX Manager and transport nodes
AvailabilityNSX health is the usual first casualty of nested networking problems; confirm the fabric before you build anything on it.
Open https://nsx-mgmt-a.site-a.vcf.lab as admin. System > Appliances: confirm the NSX Manager cluster status is Stable and the nodes are available.
System > Fabric > Hosts: confirm the four management hosts are prepared as transport nodes with configuration state Success and node status Up.
If you deployed -NsxEdgeClusterMgmtDomain: System > Fabric > Nodes > Edge Transport Nodes shows the edge nodes Up with tunnels Up; Networking > Tier-0 Gateways shows BGP neighbours to the HoloRouter established. Compare with Get-HoloDeckBGPConfig -Site a on the HoloRouter (router AS, edge AS, BGP password).
Networking > Segments: list the segments created by bring-up and compare with Get-HoloDeckOverlaySubnet -Site a for the overlay subnets Holodeck reserved.
Record NSX Manager version and build (System > Appliances or Help > About) against the VCF 9.0.2 BOM in the VCF Operations inventory.
Validation Gate
Check: NSX Manager cluster Stable; host transport nodes Up; edge/BGP healthy if deployed; versions recorded.
Expected: NSX fabric healthy.
Common Errors
Task 6 Power operations and the post-deployment snapshot set
RecoverabilityStop-/Start-HoloDeckInstance are power operations only: they do not deploy or remove anything. Combined with outer-layer snapshots they give you a reusable, known-good management domain for every later lab.
In pwsh on the HoloRouter run Stop-HoloDeckInstance -InstanceID holo (confirm, or add -Force). Watch the nested VMs power off in the outer vSphere layer.
Stop-HoloDeckInstance -InstanceID holoFrom the outer layer, snapshot every nested VM of the instance (prefix holo-) and the HoloRouter with the name holodeck-02-complete.
Run Start-HoloDeckInstance -InstanceID holo and wait for the management components to return; re-check vCenter, SDDC Manager, VCF Operations and NSX.
Start-HoloDeckInstance -InstanceID holoRecord the time to a healthy state after Start-HoloDeckInstance as your lab's cold-start figure.
Write the lab's end-state card: InstanceID, config ID, FQDNs, versions from the BOM, snapshot name, cold-start time. Keep it for holodeck-03 onwards.
Validation Gate
Check: Stop and Start cycle completed; snapshot set taken; consoles verified after restart; card written.
Expected: A reusable, documented management domain.
Common Errors
Final Validation
A management-only VCF 9.0.2 instance deployed with New-HoloDeckInstance from the HoloRouter: pre-checks passed, phases followed, idempotent resume understood, SDDC Manager, vCenter, VCF Operations and NSX validated against the documented endpoints, instance recorded with Get-HoloDeckInstance, Stop/Start cycle done and a snapshot set taken.
✓ Deployment command → Management-only parameter set with justified values
✓ Pre-checks and depot → No red errors; datastore, trunk PG and token/offline depot accepted
✓ Phases → Networking setup, nested ESX build, installer, bundles, bring-up completed (or resumed)
✓ Validation → vc-mgmt-a, sddcmanager-a, ops-a, nsx-mgmt-a (and auto-a if deployed) healthy
✓ Record and snapshot → Get-HoloDeckInstance output, Stop/Start cycle, holodeck-02-complete snapshot set, cold-start time
Cleanup / Restore
Snapshot: holodeck-02-complete
• Leave the instance powered on for holodeck-03 or stop it with Stop-HoloDeckInstance to free resources
• Keep the end-state card with FQDNs and versions
• Do not delete the config; Remove-HoloDeckInstance is covered in holodeck-07
Design Reflection (VCDX)
The bring-up you just watched is the VCF 9.0 deployment model: the VCF Installer deploys SDDC Manager, vCenter, NSX and VCF Operations into the management domain from a depot, with the fleet management appliance alongside SDDC Manager. Be ready to explain what the installer needs (DNS, NTP, a depot, four vSAN ready nodes), why management-only is the minimum shape, how idempotent automation changes your operational risk model, and what a nested lab cannot show you about physical bring-up (10 Gbps NICs, host certificates, HCL compliance).
Requirements
- Four nested vSAN ready nodes for the management domain
- VCF Installer 9.0.2.0 and the matching ESX ISO staged
- Online token or offline depot reachable from the installer
- DNS and NTP provided by the HoloRouter
Constraints
- -ManagementOnly about 4-5 hours (9.0.0.0 figure)
- One config per deployment
- Custom CIDR must be /20 and repeated on every cmdlet
- Licence Later: 90 days
Assumptions
- holodeck-01 completed
- Host sized for the chosen options
- Console can use the HoloRouter as DNS or the webtop
- No conflicting 10.1.0.0/20 or VLANs 10-25 on the external network
Risks
- Transient download failure — Impact: stalled run — Mitigation: re-run the same command (idempotent)
- Corrupted state — Impact: wrong resume — Mitigation: Reset-HoloDeckState and restart
- Nested networking misconfiguration — Impact: NSX transport nodes fail — Mitigation: trunk settings and MTU verified before deploying
- Resource exhaustion during bring-up — Impact: appliances crash — Mitigation: sized from the tables; start management-only
Self-Assessment Discussion Prompts
- What does the VCF Installer do that Cloud Builder did in 5.2, and what stays with SDDC Manager after bring-up?
- Why is idempotent deployment automation a recoverability feature, and how would you argue for it in a production automation design?
- Which pre-check failures map to design decisions (time, DNS, networking) rather than lab mistakes?
- How does the depot choice (online with proxy vs offline) map to air-gapped production designs?
- What did Get-HoloDeckInstance tell you that the VCF consoles did not?
- How would you defend 'power operations are separate from deployment operations' as an operational control?
Extensions
Deploy VCF 5.2.x with the same toolkit
Create a new config, stage the ESX ISO and Cloud Builder OVA in /holodeck-runtime/bin/5.2.2/, and run New-HoloDeckInstance -Version 5.2.2 -InstanceID holo52 -ManagementOnly. Compare the Cloud Builder bring-up with the VCF Installer bring-up (no depot, 60-day licence window, vSAN OSA only).
sameManagement domain with an NSX Edge cluster and Supervisor
Re-deploy with -NsxEdgeClusterMgmtDomain -DeploySupervisorMgmtDomain (host permitting) and validate the Tier-0 BGP peering with Get-HoloDeckBGPConfig and the Supervisor in the vSphere Client.
harderProvision-only greenfield
Run New-HoloDeckInstance -ProvisionOnly and complete the VCF deployment yourself through the VCF Installer UI using the JSON API specs in /holodeck-runtime/specs/.
harderDeveloper-mode unattended run
Set the documented $env variables and run the same deployment without interactive prompts.
same⚠ Known Pitfalls
References
- Holodeck 9.0.2 documentation (index): Environment Overview, Holorouter Overview, Concepts, Pre-requisites, Deployment, During Deployment, Post Deployment, Day 2 Operations, How ToTier 1 — Official
Primary source for every HoloRouter and toolkit step in this lab. Fetched 1 Oct 2026. - Holodeck 9.0.2 command referenceTier 1 — Official
New-HoloDeckInstance, Update-HoloDeckInstance, Get-HoloDeckInstance, New/Get/Import-HoloDeckConfig, Reset-HoloDeckState, New-HoloDeckNetworkConfig, Set-/Reset-HoloRouter, Get-HoloDeckSubnet/AppNetwork/BGPConfig/DNSConfig/OverlaySubnet/AppIpPools, Get-HolodeckServiceIPPools. - Holodeck 9.0.2 FAQ and troubleshootingTier 1 — Official
Pre-check errors (binary not found, no trunk port group, NTP not running), ESX build SSH wait, memory tiering, nested vSAN ESA. - Holodeck 9.0.2 downloads and release notes (known issues)Tier 1 — Official
HoloRouter OVA and Offline Depot OVA; release notes at /release_notes/ carry the Known Issues per release. - Holodeck GitHub repository: issues and discussionsTier 2 — Supporting source
Where to report bugs; community troubleshooting. The releases page carries no binaries (OVA downloads are on the Broadcom Support Portal). - VMware Cloud Foundation 9.0 documentation: VCF Installer log files (/var/log/vmware/vcf/domainmanager/domainmanager.log) and bring-upTier 1 — Official
VCF 9.0.1 PDF p.1298 Log Files and Locations. Task 3.