"precheck_status": "FAILED" when doing NSX-T NVDS to VDS migration

Products

• VMware NSX

Issue/Introduction

• You are running NSX-T version less than 3.2.2.
• You are migrating the host switch from NVDS (NSX VDS) to CVDS (vCenter VDS).
• The host that fails the precheck is a standalone host and is not part of a vCenter cluster.
• Migration of switch fails with precheck status failed.
• In the NSX manager log /var/log/proton/nsxapi.log you will see the the following error:

        2022-09-01T15:59:48.261Z INFO http-nio-127.0.0.1-7440-exec-48 NvdsUpgradeReadinessCheckServiceImpl 10919 FABRIC [nsx@6876 comp="nsx-manager" level="INFO" reqId="1####dc9-b266-47c3-9##9-3#######44" subcomp="manager" username="admin"] NVDS-UPGRADE              triggerNvdsUpgradePrecheck precheckStatus FAILED
    ...
    2022-09-01T15:59:48.261Z INFO NvdsUpgradeTaskExecutor1 NvdsUpgradeReadinessCheckServiceImpl 18480 FABRIC [nsx@6876 comp="nsx-manager" level="INFO" subcomp="manager"] NVDS-UPGRADE exception in validateNvdsUpgradeConfig: java.lang.NullPointerException

• As admin on the NSX-T manager, when you run the following pre check, it fails:

        NSX-mgr01> vds-migrate precheck

    GetVDSUpgradeSummary: {"precheck_id": "a#####5-d##9-4##a-b########a", "precheck_status": "FAILED"}

Environment

VMware NSX-T Data Center 3.x

Cause

This issue occurs if the host is not part of any vCenter cluster and is a standalone host. Standalone hosts have a null payload in the internal DB and NSX-T incorrectly tries to process this entry, which results in the NPE (Null Pointer Exception).

Resolution

This issue is resolved in VMware NSX 3.2.3, available at Broadcom downloads, hereby NSX-T will correctly handle the null entry for the host. The host still needs to be part of a cluster in vCenter to be migrated.
If you are having difficulty finding and downloading software, please review the Download Broadcom products and software KB.

If in case of further concerns, please feel free to raise a Broadcom Support Case Creating and managing Broadcom support cases.

Workaround:
Ensure all transport nodes are under a vCenter cluster prior migration of NVDS to VDS.