Force deleting orphaned NSX Transport Node fails with error MP9538 Could not find HostSwitch.

Products

• VMware NSX
• VMware Cloud Foundation

Issue/Introduction

In a VMware Cloud Foundation (VCF) environment, after a cluster is successfully removed from SDDC Manager and vCenter Server, ESXi hosts may remain visible in the NSX Manager UI as orphaned Transport Nodes. These hosts appear under the Standalone tab at System > Fabric > Nodes > Host Transport Nodes with a status of Host Disconnected or Unknown.

When attempting to force-delete the stale Transport Node via the NSX UI or REST API, the deletion progress loops continuously, reaching 8% ("Disabling host status in controller" stage) before resetting to 0%.

The /var/log/proton/nsxapi.log on the NSX Manager reports the following error:

ERROR L2HostConfigTaskExecutor2 ProjectVpcVCSyncerUtils 6129 SWITCHING [nsx@6876 comp="nsx-manager" errorCode="MP9538" level="ERROR" subcomp="manager"] Could not find the HostSwitch [....] of type VDS
ERROR L2HostConfigTaskExecutor2 TransportNodeAsyncRequestHandler 6129 FABRIC [nsx@6876 comp="nsx-manager" errorCode="MP9538" level="ERROR" subcomp="manager"] TransportNodeConfigTask on host failed. Abort all unprocessed stages.
com.vmware.nsx.management.switching.common.exceptions.SwitchingException: null

Environment

VMware NSX
VMware Cloud Foundation

Cause

When force-deleting a decommissioned Transport Node whose associated VDS has already been removed from vCenter, a transactional lookup of the now-deleted VDS throws an unhandled exception, which aborts the deletion task with error MP9538.

Resolution

This issue is resolved in VMware NSX 9.1 and later, where the VDS lookup failure no longer throws an unhandled exception that aborts the deletion task.

Workaround: For manual clean-up of stale Host Transport Nodes, follow the API cleanup steps documented in Script for removing large number of stale transport nodes from NSX.

Prevention: To prevent orphaned Transport Nodes from occurring during cluster removal, ensure the proper decommissioning workflow is followed prior to deleting the cluster in SDDC Manager. Refer to the "Prevention Best Practices" section in Stale Transport Nodes in NSX After SDDC Manager Cluster Deletion - VMware Cloud Foundation.