VMware NSX V2T edge-migration fails with error message "Failed to allocate vtep label for TransportNode"
search cancel

VMware NSX V2T edge-migration fails with error message "Failed to allocate vtep label for TransportNode"

book

Article ID: 377417

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

  • A v2T Migration is in progress.
  • A number of ESXi hosts have been migrated to NSX-T.
  • Following a failure the edge-migration was reverted by running the command:
    "python3 main.py -s prepare-infra revert"
  • The v2T edge-migration retry fails with error message "Failed to allocate vtep label for TransportNode. Please check logs for error received from IDAS" (ID Allocation Service).
  • The below logs are encountered on the NSX Manager node where the migration coordinator is running in /var/log/migration-coordinator/v2t/cm.log

    CM.engine.xlate_engine INFO   --------------- Starting stage runtime runtime_stage: prepare-infra type: None id: None
    CM.plugins.switching_base_plugin INFO  Reserving VTEP labels info for host host-######, ########-####-####-####-############...
    CM.clients.base_client DEBUG  API tracker: REQUEST method=POST, url=http://localhost:7440/nsxapi/api/v1/transport-nodes/vtep-labels, non-session-headers=None, params=None, data {"node_id": "########-####-####-####-############", "vteps": [{"vtep_name": "vmk#", "host_switch_name": "#####"}, {"vtep_name": "vmk#", "host_switch_name": "#####"}]}
    CM.clients.base_client ERROR  Failed to POST http://localhost:7440/nsxapi/api/v1/transport-nodes/vtep-labels with status: 400 and reason: {   "httpStatus" : "BAD_REQUEST",   "error_code" : 8205,   "module_name" : "NsxSwitching service",   "error_message" : "Failed to allocate vtep label for TransportNode. Please check logs for error received from IDAS." }

Environment

VMware NSX 4.x

VMware NSX-T Data Center 3.x

Cause

The VMware NSX v2T migrator does not support reverting "prepare-infra" stage after some hosts were migrated, if a user does so by mistake, the migrator can delete the VTEP labels allocated for the migrated ESX hosts. As a result, NSX-T IDAS releases the VTEP labels but the NSX-T database still has the labels in the Transport Node state objects. When IDAS requests to allocate a new label, it fails to allocate it because it is in a Transport Node state object.

Resolution

This is a condition that may occur in a VMware NSX environment.

    
If you believe you have encountered this issue and are unable to upgrade, please open a support request with Broadcom Support and refer to this KB article.