NVDS to VDS fails via CLI with: vds-migrate [cluster name] Failed to get Tns from Urt.

Products

• VMware NSX

Issue/Introduction

• Status of migration via API shows the following inconsistencies. 

POST https://<manager-ip>/api/v1/nvds-urt/topology/UUID?action=apply

"precheck.id": "[UUID]" ,
"precheckstatus" : "APPLY_TOPOLOGY_FAILED",
"precheckissue": [
"component": "Uplink profile",
"objid": "[objid]",
"warning": "Found inconsistent profiles: [[UUID], UUID].
Uplink profile mismatch. TransportVlan mismatch.
"recommendation": "Keep profiles consistent among transport nodes in the same NVDS",
"_protection": "NOT_PROTECTED"
｝
"component" : "NIOC profile",
"objid": "[objid]",
"warning": "Found inconsistent profiles: [[UUID], [UUID],
[UUID], [UUID], [UUID], [UUID]]. NIOC profile mismatch. ",
"recommendation": "Keep profiles consistent among transport nodes in the same NVDS",
"_protection":

NOTE: Other inconsistencies may show depending on the configuration within each specific environment

• A single NVDS is used between clusters. 
• Each cluster has their own Transport Node Profile
• Configuration within each cluster is consistent

Environment

VMware NSX

Resolution

Workaround:

This workaround involves running the pre-check against individual clusters rather than all clusters.

1. Get the Cluster ID 

GET [manager-ip]/api/v1/fabric/compute-collections 

2. Get the pre-check ID using the cluster ID obtained from Step 1

POST [manager-ip]/v1/nvds-urt/precheck-by-cluster/[Cluster-UUID:domain-cID]

3. Get the status using the pre-check ID obtained in Step 2

GET [manager-ip]/api/v1/nvds-urt/status-summary-by-cluster/[pre-check UUID]

4. Get the cluster topology using pre-check ID

GET [Manager-ip]/api/v1/nvds-urt/topology-by-cluster/[pre-check UUID]

5. Take response of Step 4 and put it in the body of Step 5

POST [Manager-ip]/api/v1/nvds-urt/topology-by-cluster/[pre-check UUID]?action=apply

6. Get the status, if host is ready put into maintenance mode

GET [Manager-ip]/api/v1/nvds-urt/status-summary-by-cluster/[pre-check UUID]

7. Once host is in maintenance mode trigger migration (Note: If you see Discover node error for this API call run step 5 again followed by step 7)

POST[Manager-ip]/api/v1/transport-nodes/<host_ID_from_above_call>?action=migrate_to_vds

8. Run call to confirm that host is migrated. 

GET[Manager-ip]/api/v1/nvds-urt/status-summary-by-cluster/[pre-check UUID]