Academy/Holodeck Lab Setup & Operations/Holodeck 9.0.2 Prerequisites, HoloRouter Deployment and Binary Staging
This lab targets VCF 9.0.2

Holodeck 9.0.2 Prerequisites, HoloRouter Deployment and Binary Staging

VCF 9.0.2Beginneradminarchitectvcdx⏱ 90 min

This foundational lab applies to VCF 9.0.x nested deployments with Holodeck 9.0.2 on a single ESX host or a vSphere cluster. Source-checked 1 Oct 2026 against the Holodeck 9.0.2 documentation at vmware.github.io/Holodeck/9.0.2/ (index: Holodeck Environment Overview, Holorouter Overview, Concepts, Pre-requisites, Deployment, Post Deployment, Day 2 Operations, How To; cmd_reference; FAQ) and, where VCF behaviour is described, the VCF 9.0.1 documentation. In Holodeck 9.x the toolkit is a PowerShell module built into the HoloRouter OVA (Photon OS): the operator SSHes to the HoloRouter, runs pwsh, and works under /holodeck-runtime. There is no Windows HoloConsole, no C:\Holodeck path, no -ConfigFile parameter and no separate Prepare phase. Rewritten for v8.2.2 after the 360-degree audit.

Objectives

  • Validate the target host against the Holodeck 9.0.2 prerequisites: vSphere 8.0 U3 or 9.0, NTP running, SSD or NVMe storage, MTU 9000, and a dedicated trunk port group with the documented security settings
  • Size the host for the chosen deployment from the documented tables (VCF 9.0 management-only or full stack with VCF Automation, vSAN OSA or ESA; VVF 9.0; VCF 5.2)
  • Deploy the HoloRouter OVA from the Broadcom Support Portal download, with SSH enabled, and verify the built-in Holodeck PowerShell module from pwsh
  • Stage the two binaries a VCF 9.0.2 deployment needs (ESX ISO and VCF Installer OVA) under /holodeck-runtime/bin/9.0.2.0/ and choose the depot method (online with a Broadcom build token, or offline depot)
  • Create a Holodeck configuration with New-HoloDeckConfig, understand the 1:1 config-to-deployment rule and the centralized config.json, and snapshot the HoloRouter as the pre-deployment baseline

Prerequisites

One physical ESX host (vSphere 8.0 U3 or 9.0) or a vSphere cluster with capacity for the chosen Holodeck deployment, SSD or NVMe backing storage, an external management network with one free IP for the HoloRouter, and a Broadcom Support Portal account entitled to VCF 9.0 downloads. A jump host or console with a browser and an SSH client on the external network.

Required skills:

  • ESX Host Client or vSphere Client administration (port groups, OVA deployment, snapshots)
  • Basic PowerShell and SSH
  • VCF 9.0 component names (VCF Installer, SDDC Manager, vCenter, NSX, VCF Operations)
📸 Starting State: S0-9x — Fresh Host (VCF 9.x)

Lab Environment

Physical ESX host (or cluster) with vSwitch0: an External port group for the HoloRouter management interface, and a Trunk port group (VLAN 4095, promiscuous mode, MAC address changes and forged transmits set to Accept, MTU 9000 on the switch) for Site A and Site B. The HoloRouter VM (Photon OS, 4 vCPU, 8 GB RAM from Holodeck 9.0.1, 75 GB) carries the Holodeck PowerShell module, the runtime under /holodeck-runtime, and the infrastructure services (DNS, DHCP, NTP, L3 routing, firewall, BGP, proxy, job scheduling, webtop).

Mermaid source
graph LR
  JH[Jump host / console] -->|SSH root, webtop :30000| HR[HoloRouter VM: External PG + Trunk PG]
  HR -->|/holodeck-runtime/bin/9.0.2.0: ESX ISO + VCF Installer OVA| BIN[(binaries)]
  HR -->|New-HoloDeckConfig -TargetHost| ESX[Physical ESX host or vCenter cluster]
  ESX --- TPG[Trunk PG VLAN 4095, security Accept x3, MTU 9000]

IP Addressing

NetworkPurposeVLAN
10.1.0.0/20Default 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)
10.2.0.0/20Default 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 needsExternal port group

Credentials

SystemUsernamePassword
HoloRouter (SSH and pwsh)rootSet 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
Broadcom Support Portalyour entitled accountNeeded to download the HoloRouter OVA, the ESX ISO and the VCF Installer OVA, and for the online-depot build token

Tasks

Task 1 Validate the target host and prepare the trunk port group

Manageability

Holodeck's pre-checks fail fast on four things: a missing trunk port group, NTP not running on the target host, missing binaries, and an unreachable target. Getting the host right first saves hours.

Step 1

Record the target host version and confirm it is one the documentation lists as tested and supported: vSphere 8.0 U3 or vSphere 9.0, either a stand-alone ESX host not managed by vCenter or a vSphere cluster managed by vCenter. Confirm the backing storage is SSD or NVMe.

Host version and storage type recorded; both within the documented Target Host Configuration.
Step 2

Confirm NTP is enabled and running on the target host (ESX Host Client > Manage > System > Time & Date, or esxcli system ntp get and esxcli system ntp start). If the target is a vCenter cluster, every host in the cluster must have NTP running and configured.

NTP service running with a reachable time source on every target host.
The pre-check error '[ERROR] NTP service is not running on target host' stops the deployment until NTP is started.
Step 3

Set the MTU of the vSwitch (vSS or vDS, or the NSX switch) that will carry the Holodeck trunk to 9000.

vSwitch MTU 9000.
Step 4

Create a dedicated trunk port group for the HoloRouter: VLAN ID 4095, on the vSwitch with the uplink; security policy Promiscuous Mode = Accept, MAC Address Changes = Accept, Forged Transmits = Accept. On a vDS, do the same on a distributed port group. If you use an NSX overlay segment instead, set its type to Overlay, allow VLANs 0-4094 (or at least 0, 10-25 for Site A and 40-58 for Site B), and attach IP Discovery, MAC Discovery and Segment Security profiles as the documentation shows.

Trunk port group present (for example Trunk-PG, VLAN 4095) with the three security settings on Accept.
The documented pre-check error for a missing trunk is '[ERROR] No trunk port groups found on the connected server'. With the default sites you can use the same trunk port group for Site A and Site B: they sit on discrete VLANs and subnets.
Step 5

If the target is a vCenter cluster, allow VLANs 0, 10-25 and 40-58 (or VLAN 0 plus your custom range) on the physical switch ports so nested hosts on different cluster hosts can talk.

Physical switch trunking covers the Holodeck VLANs (cluster targets only).
Step 6

Confirm an External (management) port group exists for the HoloRouter's management interface with one free IP (DHCP or static) and that NTP, DNS and, if used, internet access are reachable from it.

External port group identified; one management IP available.
Step 7

Check that memory tiering is not enabled on the host. The FAQ states memory tiering is not recommended for nested environments like Holodeck: deployments may complete but nested hosts can crash, deployment times grow and performance suffers.

Memory tiering off (or the risk recorded if you must keep it).
Step 8

Record the host's free CPU, memory and datastore capacity for Task 2.

Free cores, GB RAM and TB disk written down.

Validation Gate

Check: Host version, NTP, MTU 9000, trunk port group with the three Accept settings, external port group and capacity figures recorded.

Expected: The target passes the documented Configuration requirements and Target Host Configuration before the HoloRouter is deployed.

Common Errors

Pre-check: [ERROR] No trunk port groups found on the connected server
Cause: No port group with VLAN 4095 (or multiple VLANs) on the target.
Fix: Create the trunk port group (VLAN 4095, Promiscuous/MAC changes/Forged transmits Accept) on vSwitch0 or the vDS and rerun.
Pre-check: [ERROR] NTP service is not running on target host
Cause: NTP disabled on the ESX host.
Fix: Enable and start NTP (UI or esxcli system ntp set -e true; esxcli system ntp start) and rerun.
Nested hosts lose connectivity to each other on a cluster target
Cause: Physical switch ports do not trunk the Holodeck VLANs.
Fix: Allow VLANs 0, 10-25 and 40-58 (or the custom range) on the uplinks.
Slow or unstable nested ESX hosts
Cause: Memory tiering enabled, or HDD-backed storage.
Fix: Disable memory tiering; use SSD or NVMe backing storage as the documentation requires.

Task 2 Size the host from the documented requirement tables

Performance

The requirement tables are the design input for everything that follows; a lab that does not fit the table fails late and expensively.

Step 1

Copy the Physical Host requirements tables for your choice: VCF 9.0 with VCF Automation and vSAN ESA: 32 CPU, 325 GB RAM, 1.1 TB disk single site (64 / 768 GB / 2.5 TB dual site); VCF 9.0 with VCF Automation and vSAN OSA: 24 CPU, 325 GB, 1.1 TB (48 / 768 GB / 2.5 TB); VVF 9.0: 12 CPU, 256 GB, 1 TB (24 / 512 GB / 2 TB); VCF 5.2: 16 CPU, 384 GB, 2 TB (32 / 1 TB / 4 TB). Add the HoloRouter itself: 4 CPU, 8 GB RAM (9.0.1 and later), 75 GB on SSD/NVMe.

The applicable row and the HoloRouter footprint recorded.
Step 2

Compare with the free capacity from Task 1 step 8 and decide the deployment shape for holodeck-02: -ManagementOnly (4 nested hosts) is the smallest VCF 9.0 shape; a full stack adds 3 workload hosts; -DeployVcfAutomation and NSX Edge clusters add more.

Deployment shape chosen with headroom documented.
Step 3

Note the documented timings for VCF 9.0.0.0 as planning figures: -ManagementOnly about 4-5 hours; with -NsxEdgeClusterMgmtDomain 5-6 hours; full stack with VCF Automation and Supervisor 12+ hours. These depend on the physical environment and network.

Maintenance window planned from the documented timings.
Step 4

Decide the vSAN mode. The FAQ warns that nested vSAN ESA on non-vSAN physical storage shows unexpected datastore consumption and poor efficiency; the documentation recommends running nested ESA on physical infrastructure that also runs vSAN. On a single host with local SSDs, OSA (the default) is the safer choice.

vSANMode chosen (OSA default unless the physical layer is vSAN).
Step 5

Record the licensing model: Holodeck 9.0 deploys VCF 9.0.x in License Later mode, fully functional for 90 days (60 days for VCF 5.2.x); afterwards redeploy or add licences. Plan the lab calendar around that window.

90-day window noted with the planned redeploy date.
Step 6

Download from the Downloads page and the Broadcom Support Portal: the HoloRouter OVA (Holodeck 9.0.2), the ESX ISO and the VCF Installer OVA for VCF 9.0.2.0 (the documentation recommends VCF Installer 9.0.2.0 for 9.0.0.0, 9.0.1.0 and 9.0.2.0; Installer 9.0.0.0 is no longer supported). Record file names and sizes.

Three files downloaded (or queued for download inside the webtop later) with names and sizes recorded.
The GitHub releases page carries no binaries. The OVA and VCF binaries come from the Broadcom Support Portal and require VCF 9.0 entitlement.

Validation Gate

Check: Requirement row, deployment shape, timings, vSAN mode, licence window and the three downloads recorded.

Expected: A sized, scheduled plan for the nested deployment.

Common Errors

Deployment runs out of memory during bring-up
Cause: Host sized below the documented table for the chosen options.
Fix: Use -ManagementOnly, drop -DeployVcfAutomation or the edge cluster, or add RAM; the tables are minimums for the stated options.
vSAN ESA datastore appears fully consumed in the nested lab
Cause: Nested ESA on non-vSAN physical storage (FAQ).
Fix: Use OSA in single-host labs, or run nested ESA only on a vSAN-backed physical layer.
Download of the VCF Installer OVA is refused
Cause: No VCF 9.0 entitlement on the Broadcom account.
Fix: Sign in with an entitled account; contact Broadcom support if access problems persist (FAQ).

Task 3 Deploy the HoloRouter OVA and verify the Holodeck module

Manageability

The HoloRouter is the toolkit. Everything you will run in later labs is a cmdlet of the Holodeck PowerShell module shipped inside this appliance, driven from pwsh after SSH.

Step 1

Stand-alone host: in the ESX Host Client, right-click Virtual Machines > Create/Register VM > Deploy a virtual machine from OVF or OVA file, name the VM (for example holorouter) and select the HoloRouter OVA. vCenter target: right-click the cluster > Deploy OVF Template and select the OVA.

OVA deployment wizard open with the HoloRouter OVA selected.
Step 2

Select the storage and the networks: the External port group for Management, and the Trunk port group for Site A and for Site B (the same trunk port group may serve both with the default sites). Accept the EULA.

Management on the External port group; Site A and Site B on the trunk port group.
Step 3

Networking properties: leave the fields blank for DHCP (fill in DNS if your DHCP does not provide it), or set a static Management IP, CIDR, Gateway and DNS with access to the rest of your network. This IP is how you reach the HoloRouter and every nested component. Set the HoloRouter password.

Management addressing decided and entered; password set.
Step 4

Select the SSH checkbox and, if you want the light desktop, the Webtop checkbox. The 9.0.2 documentation notes the webtop on port 30000 has no authentication: do not expose it externally, or leave it unselected.

SSH enabled; webtop enabled only if the network is private.
Webtop in 9.0.2 is unauthenticated. Treat the HoloRouter external IP as lab-internal.
Step 5

Click Finish and power on the HoloRouter. From the jump host run ssh root@<holorouter-ip> and log in with the OVA password.

SSH prompt on the HoloRouter (Photon OS).
Step 6

Start PowerShell with pwsh and verify the module: Get-Command -Module Holodeck lists the cmdlets (New-HoloDeckConfig, Import-HoloDeckConfig, Get-HoloDeckConfig, New-HoloDeckInstance, Update-HoloDeckInstance, Get-HoloDeckInstance, New-HoloDeckNetworkConfig, Set-HoloRouter, Reset-HoloRouter, Get-HoloDeckSubnet, Get-HoloDeckAppNetwork, Get-HoloDeckBGPConfig, Get-HoloDeckDNSConfig, Reset-HoloDeckState and others). Run Get-Help New-HoloDeckInstance to see the parameter sets. List the runtime: ls /holodeck-runtime (templates, config, bin, logs, specs).

Holodeck cmdlets listed; /holodeck-runtime present with templates/config.json.
The module lives on the HoloRouter and nowhere else: there is nothing to install on a workstation, no git clone, and no C:\Holodeck path.

Validation Gate

Check: HoloRouter deployed on the right port groups with SSH; pwsh opens; Get-Command -Module Holodeck lists the cmdlets; /holodeck-runtime exists.

Expected: The toolkit is reachable and verified on the HoloRouter.

Common Errors

Cannot SSH to the HoloRouter
Cause: SSH not selected at deployment, wrong management IP, or DHCP gave a different address.
Fix: Check the VM console for the address; redeploy with the SSH checkbox selected if needed.
Nested components are unreachable later although the HoloRouter is
Cause: Site A/B interfaces were attached to a non-trunk port group.
Fix: Edit the VM's network adapters: Site A and Site B must be on the trunk (VLAN 4095) port group.
Webtop reachable from outside the lab
Cause: Unauthenticated webtop exposed on a routed network.
Fix: Restrict access at the firewall or redeploy without the webtop option.

Task 4 Stage the VCF 9.0.2 binaries and choose the depot method

Manageability

A VCF 9.0.x nested deployment needs exactly two files on the HoloRouter: the ESX ISO and the VCF Installer OVA. The remaining VCF components are downloaded by the VCF Installer from the online depot (Broadcom build token) or from an offline depot you prepare.

Step 1

Create or confirm the version folder on the HoloRouter: ls /holodeck-runtime/bin/9.0.2.0/ (folders exist for 9.0.2.0, 9.0.1.0, 9.0.0.0, 5.2.2, 5.2.1 and 5.2).

/holodeck-runtime/bin/9.0.2.0/ present.
Step 2

Copy the two binaries from your jump host: scp /<local-path>/<ESX ISO File Name> root@<Holorouter-IP>:/holodeck-runtime/bin/9.0.2.0/ and scp /<local-path>/<VCF Installer OVA File Name> root@<Holorouter-IP>:/holodeck-runtime/bin/9.0.2.0/. Alternatively, with the webtop enabled, download them from the Broadcom Support Portal inside the webtop browser straight into that folder.

scp /<local-path>/<ESX ISO File Name> root@<Holorouter-IP>:/holodeck-runtime/bin/9.0.2.0/
scp /<local-path>/<VCF Installer OVA File Name> root@<Holorouter-IP>:/holodeck-runtime/bin/9.0.2.0/
ls -lh /holodeck-runtime/bin/9.0.2.0/ shows the ESX ISO and the VCF Installer OVA.
The pre-check '<Component> binary was not found. Download the binary from the Broadcom support portal and place it in /holodeck-runtime/bin/<version> folder' fires when either file is missing from the exact version folder.
Step 3

Decide the depot: Online (default -DepotType) needs internet access from the nested VCF Installer (directly or through the HoloRouter proxy) and a Broadcom support site build token, which the deployment prompts for after the pre-checks. Offline needs the Offline Depot Appliance deployed and populated beforehand (Offline Depot page); the deployment then prompts for the depot IP, port, protocol and credentials. The offline depot applies to VCF 9.x only; VCF 5.2.x uses Cloud Builder and needs no depot.

Depot method chosen; token available or offline depot ready.
Step 4

If your lab uses a proxy for outbound access, note the proxy protocol, IP, port and credentials: the interactive deployment (or the -DeveloperMode environment variables enable_proxy, proxy_protocol, proxy_ip, proxy_port, enable_proxy_auth, proxy_username, proxy_password) will ask for them.

Proxy details recorded or 'no proxy' confirmed.
Step 5

For an offline depot, confirm with the documentation note that the depot manifest is the latest from the Broadcom Support Portal if the depot already held earlier VCF 9 binaries.

Depot manifest current (offline only).
Step 6

Write down the two file names exactly as staged; you will see them referenced by the pre-checks in holodeck-02.

File names recorded.

Validation Gate

Check: Both binaries in /holodeck-runtime/bin/9.0.2.0/; depot method and proxy details decided.

Expected: The deployment pre-checks for binaries will pass and the depot prompts can be answered.

Common Errors

Pre-check reports a binary not found although the file was copied
Cause: File placed in the wrong version folder (for example 9.0.0.0 instead of 9.0.2.0), or an interrupted scp.
Fix: Place both files in /holodeck-runtime/bin/9.0.2.0/ and compare sizes with the downloads.
VCF Installer cannot download bundles
Cause: No internet path from the nested network, missing proxy settings, or an invalid build token.
Fix: Supply the proxy details when prompted or in -DeveloperMode variables; regenerate the token; or switch to the offline depot.
Offline depot deployment fails on bundle validation
Cause: Depot populated with an old manifest.
Fix: Use the latest manifest from the Broadcom Support Portal before deploying (documentation note).
Cloud Builder OVA staged for a 9.0.x deployment
Cause: Confusing the 5.2.x toolchain with 9.x.
Fix: VCF 9.0.x uses the VCF Installer OVA; Cloud Builder is for 5.2.x.

Task 5 Create the Holodeck configuration and take the baseline snapshot

Manageability

Holodeck 9.x is built around a centralized config file created and loaded by cmdlets; you never edit it by hand, and one config maps to exactly one deployment. The HoloRouter snapshot gives you a clean baseline to return to in holodeck-07.

Step 1

On the HoloRouter in pwsh, create the configuration: New-HoloDeckConfig -Description 'lab01-9.0.2' -TargetHost <target vCenter or ESX IP/FQDN> -Username <username> -Password <password>. The cmdlet copies /holodeck-runtime/templates/config.json to /holodeck-runtime/config/<config-ID>.json with a unique config ID and loads it into $config.

New-HoloDeckConfig -Description 'lab01-9.0.2' -TargetHost <target-host> -Username <username> -Password <password>
A new config ID is returned and $config holds the loaded configuration.
Step 2

Inspect the loaded configuration: run $config and read the config ID and description. Then list all configurations with Get-HoloDeckConfig, and retrieve yours with Get-HoloDeckConfig -configID <id>.

Your config ID and description are listed; the file exists under /holodeck-runtime/config/.
Step 3

Practise re-loading: exit pwsh, start it again, and run Import-HoloDeckConfig -ConfigID <id>. $config is per PowerShell session; a new session always needs the import.

$config is populated again after the import.
Step 4

Write down the two rules from the documentation: (a) each configuration maps to a single deployment; create a new config for every new deployment and never reuse one, or state conflicts follow; (b) customers are not expected to interact with the config file directly or edit it; the templates/config.json is a template only.

Both rules in your notebook.
Step 5

Review the pre-check list the deployment will run (Built-In PreChecks): binaries present in the right location, target host reachable, trunk port group present, NTP running. Tick each against Tasks 1 and 4.

All four pre-check inputs satisfied.
Step 6

Power off nothing yet; from the outer layer (ESX Host Client or vSphere Client) take a snapshot of the HoloRouter VM named holodeck-01-baseline. This is the clean, pre-deployment state.

Snapshot holodeck-01-baseline exists on the HoloRouter VM.
Step 7

Record the lab's identity data for holodeck-02: config ID, target host, trunk port group name, datastore name, depot method, planned InstanceID and vSANMode.

Deployment card completed.

Validation Gate

Check: Config created and re-imported; the two config rules recorded; pre-check inputs satisfied; HoloRouter snapshot taken; deployment card written.

Expected: Ready to run New-HoloDeckInstance in holodeck-02.

Common Errors

$config is empty in a new session
Cause: $config is session-scoped.
Fix: Run Import-HoloDeckConfig -ConfigID <id> after every new pwsh session.
A second deployment behaves unpredictably
Cause: The same config was reused for two deployments.
Fix: Create a new config with New-HoloDeckConfig for each deployment (1:1 mapping).
Config edited by hand no longer loads
Cause: Manual edits to /holodeck-runtime/config/<id>.json.
Fix: Create a fresh config with the cmdlets; do not edit config files directly.

Final Validation

Target host validated and sized from the documented tables, trunk port group with the documented security settings, HoloRouter deployed with SSH and the Holodeck module verified in pwsh, the ESX ISO and VCF Installer OVA staged under /holodeck-runtime/bin/9.0.2.0/, depot method chosen, configuration created with New-HoloDeckConfig, and a baseline snapshot taken.

✓ Host prerequisites → vSphere 8.0 U3 or 9.0; NTP running; SSD/NVMe; MTU 9000; trunk PG VLAN 4095 with Promiscuous/MAC changes/Forged transmits Accept

✓ Sizing → Chosen deployment row recorded with HoloRouter 4 vCPU / 8 GB / 75 GB added; 90-day licence window noted

✓ HoloRouter → Deployed on External + Trunk port groups; SSH works; Get-Command -Module Holodeck lists the cmdlets

✓ Binaries → ESX ISO and VCF Installer OVA in /holodeck-runtime/bin/9.0.2.0/

✓ Depot → Online token or offline depot ready; proxy details recorded

✓ Configuration → New-HoloDeckConfig created; Get/Import-HoloDeckConfig practised; 1:1 rule recorded

✓ Baseline → Snapshot holodeck-01-baseline on the HoloRouter VM

✓ Deployment card → Config ID, target, trunk PG, datastore, depot, InstanceID, vSANMode written

✓ Webtop exposure → Webtop disabled or confined to a private network

Cleanup / Restore

Snapshot: holodeck-01-baseline

• Keep the HoloRouter powered on with the baseline snapshot in place

• Keep the staged binaries; they are reused by every deployment of that version

• Store the Broadcom build token and HoloRouter password in your password manager

Design Reflection (VCDX)

A panel does not test Holodeck itself, but it tests whether you understand what a nested lab can and cannot prove. Be ready to explain the HoloRouter's role (one appliance providing DNS, DHCP, NTP, routing, firewall, BGP and proxy so the lab needs one external IP and no physical VLANs), why isolation at the vSS/vDS trunk level with VLAN tagging works, what the licence window implies for a study plan, and which physical-world constraints (vSAN ESA on non-vSAN storage, memory tiering, 10 Gbps NIC requirements) the nested lab masks. Use the lab to practise VCF 9.0 operations, not to size production.

Requirements

  • Target host on vSphere 8.0 U3 or 9.0 with NTP, SSD/NVMe and MTU 9000
  • Dedicated trunk port group (VLAN 4095) with Promiscuous Mode, MAC Address Changes and Forged Transmits set to Accept
  • HoloRouter deployed with SSH; one external management IP
  • ESX ISO and VCF Installer OVA staged in /holodeck-runtime/bin/9.0.2.0/
  • A fresh New-HoloDeckConfig per deployment

Constraints

  • Resource tables: VCF 9.0 management-only or full stack with Automation needs 24-32 CPU, 325 GB RAM, 1.1 TB (single site)
  • HoloRouter 4 vCPU, 8 GB RAM, 75 GB
  • Licence Later mode: 90 days for VCF 9.0.x, 60 days for 5.2.x
  • Webtop is unauthenticated in 9.0.2
  • Config files are created and loaded by cmdlets only; never edited by hand
  • Offline depot applies to VCF 9.x only

Assumptions

  • Broadcom Support Portal account with VCF 9.0 entitlement
  • Jump host with SSH and a browser on the external network
  • Physical switch trunking in place for cluster targets
  • Internet access or an offline depot for VCF bundle downloads
  • Physical storage is SSD/NVMe; OSA chosen unless the physical layer is vSAN
  • Memory tiering disabled
  • A single Holodeck environment per external IP
  • Holodeck is used for testing and training only (not a VMware supported product; similar to a Fling)

Risks

  • Undersized host — Impact: bring-up fails late — Mitigation: size from the tables with headroom; start with -ManagementOnly
  • Trunk security settings missing — Impact: nested hosts cannot communicate — Mitigation: verify the three Accept settings before deploying
  • Webtop exposed — Impact: unauthenticated access to the lab — Mitigation: private network or webtop disabled
  • Wrong version folder for binaries — Impact: pre-check failure — Mitigation: ls the exact folder before New-HoloDeckInstance
  • Config reuse — Impact: state conflicts — Mitigation: one config per deployment
  • Licence window overrun — Impact: lab stops working — Mitigation: calendar the 90-day redeploy

Self-Assessment Discussion Prompts

  1. Why does Holodeck need only one external IP, and what does the HoloRouter do that a physical VCF design needs the data centre to provide?
  2. Why must the trunk port group accept promiscuous mode, MAC changes and forged transmits, and what does that tell you about nested networking?
  3. Which of the documented host requirements would you push back on in a production design, and why?
  4. What does License Later mode allow you to study, and what can it not show you about a licensed estate?
  5. Why is the configuration file created by a cmdlet and never edited, and how would you defend a similar 'generated, never hand-edited' rule in a production design?
  6. Where does nested vSAN ESA mislead you about ESA in production?

Extensions

Deploy the Offline Depot Appliance

Follow the Offline Depot page to deploy and populate the Offline Depot OVA with the VCF 9.0.2 bundles, then record the depot IP, port, protocol and credentials for an offline deployment in holodeck-02.

harder

vCenter cluster as the target

Repeat Task 1 against a vSphere cluster target: create the distributed trunk port group, allow the Holodeck VLANs on the physical uplinks, and confirm NTP on every host.

same

NSX overlay segment as the trunk

Instead of a vSS/vDS port group, prepare an NSX overlay segment with the documented IP Discovery, MAC Discovery and Segment Security profiles and VLANs 0-4094 allowed.

harder

Developer-mode variables

Prepare the -DeveloperMode environment variables (brcm_build_token, enable_proxy, datastore_name, trunk_port_group_name, and cluster_name/dc_name for a vCenter target) so holodeck-02 can run without interactive prompts.

same

⚠ Known Pitfalls

Looking for the toolkit on GitHub releases or a Windows console COMMON
Problem: Older Holodeck generations used a Windows HoloConsole and local scripts; 9.x ships the module inside the HoloRouter OVA.
Resolution: Download the HoloRouter OVA from the Broadcom Support Portal; SSH to it and run pwsh.
Staging five binaries and a manifest COMMON
Problem: VCF 9.0.x needs only the ESX ISO and the VCF Installer OVA on the HoloRouter.
Resolution: Place the two files in /holodeck-runtime/bin/9.0.2.0/; the installer fetches the rest from the online or offline depot.
Unauthenticated webtop on a shared network COMMON
Problem: Port 30000 has no authentication in 9.0.2.
Resolution: Keep the HoloRouter on a private network or deploy without webtop.
Nested ESA on local disks COMMON
Problem: Datastore shows as fully consumed with poor performance.
Resolution: Use OSA in single-host labs (FAQ).
Memory tiering to make the host fit COMMON
Problem: Deployment completes but nested hosts crash or crawl.
Resolution: Disable memory tiering; size with physical RAM (FAQ).
Editing templates/config.json COMMON
Problem: Hand edits break later cmdlets.
Resolution: Use New-HoloDeckConfig / Import-HoloDeckConfig only.
Reusing one config for a second deployment COMMON
Problem: State conflicts and unpredictable behaviour.
Resolution: One config per deployment.
Using VCF Installer 9.0.0.0 COMMON
Problem: No longer supported by Holodeck for nested deployments.
Resolution: Use VCF Installer 9.0.2.0 for 9.0.0.0 through 9.0.2.0.

References

Was this page useful?
Type to search. ↑ ↓ to move, Enter to open, Esc to close.