ESXi host experiences a PSOD during a V2T migration

Products

• VMware NSX

Issue/Introduction

Symptoms:

• Migration Coordinator is used to migrate from NSX Data Center for vSphere to NSX-T Data Center
• During the ESXi host migration process, an ESXi host prepared for NSX-T may crash with a PSOD
• A backtrace similar to this example may be observed however it is possible the backtrace for this issue may vary

2021-11-13T03:39:21.994Z cpu69:6941753)@BlueScreen: #PF Exception 14 in world 6941753:vmnic6-pollW IP 0x420005233dfb addr 0x45ec1b0fa8c0
PTEs:0x4840088023;0x0;
2021-11-13T03:39:22.007Z cpu69:6941753)Code start: 0x420005000000 VMK uptime: 11:12:26:07.810
2021-11-13T03:39:22.029Z cpu69:6941753)0x45390d71b460:[0x420005233dfb]PktPartialCopyBuildPkt@vmkernel#nover+0x103 stack: 0x43470000006e
2021-11-13T03:39:22.053Z cpu69:6941753)0x45390d71b500:[0x42000523421b]Pkt_PartialCopyWithHeadroom@vmkernel#nover+0x30 stack: 0x45390d71b650
2021-11-13T03:39:22.071Z cpu69:6941753)0x45390d71b530:[0x420005273556]vmk_PktPartialCopy@vmkernel#nover+0xb stack: 0x0
2021-11-13T03:39:22.097Z cpu69:6941753)0x45390d71b540:[0x4200071134c4][email protected]#1.1.7.0.18844960+0xaa1 stack: 0x0
2021-11-13T03:39:22.123Z cpu69:6941753)0x45390d71b660:[0x420007117344][email protected]#1.1.7.0.18844960+0x59 stack: 0xffffffffffff
2021-11-13T03:39:22.148Z cpu69:6941753)0x45390d71b6b0:[0x42000716a32e][email protected]#1.1.7.0.18844960+0x43f stack: 0xa000193
2021-11-13T03:39:22.171Z cpu69:6941753)0x45390d71b770:[0x420007160bba][email protected]#1.1.7.0.18844960+0x743 stack: 0x1c48
2021-11-13T03:39:22.192Z cpu69:6941753)0x45390d71bdd0:[0x4200051fb137]IOChain_Resume@vmkernel#nover+0x270 stack: 0x430514e1b1f0
2021-11-13T03:39:22.211Z cpu69:6941753)0x45390d71be70:[0x42000523e0f2]Port_InputResume@vmkernel#nover+0xbf stack: 0x3fff
2021-11-13T03:39:22.230Z cpu69:6941753)0x45390d71bec0:[0x4200051c949b]Net_AcceptRxList@vmkernel#nover+0x264 stack: 0x202
2021-11-13T03:39:22.251Z cpu69:6941753)0x45390d71bf70:[0x420005275176]NetPollWorldCallback@vmkernel#nover+0xa7 stack: 0x1900000033
2021-11-13T03:39:22.270Z cpu69:6941753)0x45390d71bfe0:[0x420005369481]CpuSched_StartWorld@vmkernel#nover+0x82 stack: 0x0
2021-11-13T03:39:22.289Z cpu69:6941753)0x45390d71c000:[0x4200050be69f]Debug_IsInitialized@vmkernel#nover+0xc stack: 0x0
2021-11-13T03:39:22.303Z cpu69:6941753)base fs=0x0 gs=0x420051400000 Kgs=0x0

Environment

VMware NSX-T Data Center

Cause

This issue may occur during a V2T migration when an ESXi host prepared for NSX-T receives a VXLAN packet with the MTEP replication flag set.
This can result in an unhandled exception resulting in a PSOD of the ESXi host.

Resolution

This is a known issue impacting NSX-T Data Center.

Workaround:
If this issue occurs, reboot the host and continue with the migration.
The issue can only occur when some hosts continue to run NSX Data Center for vSphere.
Once the migration of all hosts to NSX-T Data Center completes, the PSOD cannot occur.