vMotion migration fails at 20% with Timeout 195887137 due to switch mismatch

Products

• VMware vSphere ESXi

Issue/Introduction

vMotion migrations between ESXi hosts fail consistently at approximately 20% progress.
The vSphere Client displays the error: Migration to host <####> failed with error Timeout (195887137).

Continuous vmkping sessions between the vMotion VMkernel interfaces show immediate ICMP packet drops the moment the vMotion task initiates.

Environment

VMware vSphere ESXi 7.x / 8.x
Source Host: VMkernel (vmk) interface configured on a Standard Switch (VSS) using standard active/active uplinks.
Destination Host: VMkernel (vmk) interface configured on a Distributed Virtual Switch (VDS) using LACP (Link Aggregation Control Protocol).

Cause

A virtual networking configuration mismatch exists between the transport hosts. The source host uses standard switch teaming (hashing based on originating virtual port ID), while the destination host uses a Distributed Switch with LACP (which relies on physical switch hashing). This discrepancy in how the physical network handles traffic hashes for the state transfer phase results in packet dropping and eventual timeout.

Resolution

To restore migration capabilities, align the virtual switch topology between the hosts:

• Identify the host utilizing the Distributed Switch with LACP.
• Temporarily migrate the vMotion VMkernel interface (vmk) from the Distributed Switch to a Standard Switch that matches the topology of the peer host
• Ensure the switch utilizes a single or standard active/active uplink (e.g., vmnic1).
• Retest the migration.
  If the migration succeeds, engage your physical networking team to investigate the physical switch port configurations to determine why packet drops occur when traversing the LACP configuration on the distributed portgroup.