[Bug]: New-HolodeckInstance fails with -DeveloperMode when Cluster has limited view or resource pool

[certanet]

Holorouter Version

9.1

VCF Version

VCF 9.1

Target Type

vCenter Cluster

Describe the bug

Similar to #140, but when deploying New-HolodeckInstance with all appropriate env vars set, including $env:cluster_name = "MY-CLUSTER" the deployment fails with error:

PreChecks[1101]: [ERROR] Failed to verify target host hardware details.
PreChecks[1101]: [ERROR] Cannot process argument transformation on parameter 'Cluster'. Cannot convert value to type System.String.
PreChecks[1101]: [ERROR] Failed to complete Pre-Checks.

This likely happens when a Cluster has a user with limited permissions on it or a Resource Pool. As the table output with Cluster and ResourcePool objects displays multiple entries for the Cluster with the given name.

Per #140 it looks like a Resource Pool can now be selected, but is this exposed via an env var for -DeveloperMode and documented?

Reproduction steps

1. Create a vSphere Cluster in vCenter and a user with limited permissions on it
2. Set appropriate env vars for your environment per docs
3. Attempt a New-HolodeckInstance deployment with -DeveloperMode

Expected behavior

The deployment works using the first named instance of the Cluster found, or fails as expected and an env var is exposed for Resource Pool or Cluster MOID or similar?

Relevant logs

Additional context

No response

[dhruv-tyagi-broadcom]

Yes, you can specify a resource pool in Developer mode using the following environment variable:

$env:resource_pool_name = "<resource_pool_name>" (optional, if you want to deploy to a specific resource pool in vCenter)

Docs have been updated to reflect the same.

For the issue you mentioned, we will need more information around:

1. the account permissions specified at New-HoloDeckConfig command.
2. Also if you can run the New-HoloDeckInstance command without developer mode and see what gets listed.
3. How many clusters and/resource pools exist in the target env?
4. Share error logs

[certanet]

Thanks for updating the docs. I've created a Resource Pool and used the env var which has worked around it.

Answers:

1. A full priv account e.g. SSO admin
   2/3. When a single Cluster exists and no Resource Pool or limited privileged user the Cluster is listed once. With a user created with limited permissions e.g. can't add hosts to cluster, the Cluster is listed twice in the output. When a Resource Pool is added, the Cluster is listed 3 times and I can then use the resource_pool_name env var to pick the unique ID for it.