vMotion Fails to Recently Upgraded Hosts During NSX Upgrade

Products

• VMware NSX

Issue/Introduction

During NSX upgrade, vMotion operations fail when attempting to migrate virtual machines to ESXi hosts immediately after their upgrade completes. The issue manifests when DRS attempts automatic migrations to recently upgraded hosts, with both automated and manual vMotion attempts failing to these hosts.

The upgraded hosts appear online in vCenter but reject incoming vMotion operations. The issue persists for a period after each host completes its upgrade process before vMotion functionality resumes normally.

Environment

• VMware NSX
• VMware vSphere ESXi
• VMware vCenter Server

Cause

Recently upgraded ESXi hosts require additional time to fully initialize all services after the upgrade process completes. While the hosts appear online and show as upgraded in vCenter, the vMotion service and related components are not fully operational immediately following the upgrade. The initialization process takes longer than the typically expected 15 minutes and can require up to a couple of hours to complete.

Resolution

Implement the following workaround during NSX upgrade:

1. Disable DRS on the cluster before beginning host upgrades
2. Proceed with host upgrades according to the NSX upgrade workflow
3. After each host completes its upgrade:
   • Mark the host as excluded from vMotion operations
   • Continue upgrading remaining hosts
   • Allow additional time for host initialization (can take up to a couple of hours)
4. Manually control VM migrations during the upgrade window:
   • Direct migrations only between fully operational hosts
   • Avoid migrations to recently upgraded hosts
5. After allowing sufficient initialization time for each host:
   • Test vMotion to the host with a non-critical VM
   • If successful, resume normal operations for that host
6. Once all hosts are upgraded and operational:
   • Re-enable DRS on the cluster
   • Resume normal automated migration operations

Note: While hosts typically complete initialization within 15 minutes, this issue may require up to a couple of hours. Allow sufficient time before contacting support for troubleshooting.

If the error persists after following these steps and allowing adequate initialization time, contact Broadcom Support for further assistance.