Reconnecting a failed host Returns the error "Credentials for host need to be specified"

Products

• VMware NSX

Issue/Introduction

• Reconnecting a failed host to NSX gives the error "Credentials for host need to be specified. (Error code: 26208)".
• Relevant nsxapi.log from NSX manager(/var/log/proton/nsxapi.log):

  • 2025-06-30T03:53:42.642Z  INFO http-nio-<Local Host IP>-7440-exec-256 UI_LOG 87643 - [nsx@6876 comp="nsx-manager" level="INFO" reqId="<Request ID>" subcomp="manager" username="admin"] {"user":"admin","message":"Api Errors->","messageData":{"headers":{"normalizedNames":{},"lazyUpdate":null},"status":400,"statusText":"Bad Request","url":"https://<NSX manager IP>/policy/api/v1/infra/sites/default/enforcement-points/default/host-transport-nodes/<host TN>host-15992","ok":false,"name":"HttpErrorResponse","message":"Http failure response for https://<NSX manager IP>/policy/api/v1/infra/sites/default/enforcement-points/default/host-transport-nodes/<host TN>host-15992: 400 Bad Request","error":{"httpStatus":"BAD_REQUEST","error_code":26208,"module_name":"service-installation-fabric","error_message":"Error: Credentials for host need to be specified. (Error code: 26208)"

• This Behavior also seen on NSX environment where the ESXi transport node was rebuilt and not removed from NSX before rebuilt.

Environment

NSX 4.x

Cause

This issue is caused due to stale entry of the Node when ESXi TN being rebuilt with same DNS and IP.

Resolution

The resolution depends on environment that is being managed.

Without vLCM controlling the hosts image:

1. Remove the default TransportNodeProfile from the cluster in question.
2. Force delete NSX from the orphaned host(s)
3. Re-apply the default TransportNodeProfile to the cluster. 
4. If the host is not currently in the cluster add it back.
5. It will then run through NSX config, and complete setting up the host.

If vLCM controls the hosts image, the TransportNodeProfile cannot be removed. The instructions below will work around this restraint.

(See "For a Single Host" in Troubleshooting NSX Uninstall Process for screenshots of this process.)

1. Using vCenter, place the host into Maintenance Mode.
2. Move the host to the Datacenter level of the vCenter Inventory.
3. Ensure NSX is properly removed from the host and NSX. (No VIBs installed on the host and NSX GUI reports the host is "Unprepared.")
4. Return the host to the cluster.
5. NSX will reinstall and prepare the host per the cluster's attached TransportNodeProfile.