When deploying a new cluster or adding hosts via SDDC Manager, the workflow fails during the "Prepare Transport Node Collection" task.
search cancel

When deploying a new cluster or adding hosts via SDDC Manager, the workflow fails during the "Prepare Transport Node Collection" task.

book

Article ID: 448989

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

  • When adding a new cluster or adding hosts via SDDC Manager, the workflow fails during the "Prepare Transport Node Collection" stage.
  • SDDC Manager Error: Unable to create transport node collection with profile [UUID] on compute collection [UUID].
  • NSX Manager Error: Validation Errors 9560: Transport Zone [UUID] cannot be relocated from HostSwitch [ID] to HostSwitch [ID] in a single transaction. Delete HostSwitch from current Transport Zone to add later to another Transport Zone.

Environment

VMware NSX

Cause

  • This issue might occur when stale Transport Node Profile (TNP) entries or duplicate profiles exist in the NSX database, often left over from a previous failed deployment or stretch cluster workflow.

Resolution

  • To resolve this issue, perform a clean removal of the failed configuration before rerunning the SDDC Manager workflow.
    1. Place the affected hosts in Maintenance Mode and move them out of the cluster to the Datacenter level in vCenter.
    2. In the NSX Manager UI, navigate to System > Fabric > Nodes > Managed Nodes and select the affected hosts under the Other Nodes tab.
    3. Select Delete NSX and check the Force Delete option to remove stale configurations.
    4. Navigate to System > Fabric > Profiles > Transport Node Profiles and delete any duplicate or stale profiles associated with the cluster.
    5. Wipe the hosts and delete the temporary cluster object from vCenter to ensure no stale metadata remains.
    6. Rerun the Cluster Deployment workflow from SDDC Manager.

Additional Information

  • Triggering Uninstallation from the vSphere Web Client Technical Documentation.
  • Similar issue with the same error described in KB 319037, but the error appears when trying to apply a Transport Node Profile to an existing cluster.