ESXi 6.5 and 6.7 host fails with a PSOD PF Exception 14 in world xxxxx:Unmap Helper IP

Products

• VMware vSphere ESXi

Issue/Introduction

To avoid ESXi host crash.

Symptoms:

• ESXi 6.5 and 6.7  host fails with a PSOD backtrace PF Exception 14 in world xxxxx:Unmap Helper IP
• We see backtrace similar to below

2019-04-16T16:32:44.690Z cpu3:66390)World: 2780: TR 0x4020 GDT 0x4390f9821000 (0x402f) IDT 0x418024d3f000 (0xfff)
2019-04-16T16:32:44.690Z cpu3:66390)World: 2781: CR0 0x80010031 CR3 0x1484c8000 CR4 0x42768
ESC[7m2019-04-16T16:32:44.742Z cpu3:66390)WARNING: kbdmode_set:519: invalid keyboard mode 4: Not supportedESC[0m
2019-04-16T16:32:44.742Z cpu3:66390)Backtrace for current CPU #3, worldID=66390, fp=0x4305d3836af0
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1bd30:[0x418025b0b55e]Res6AffMgr_UpdateState@esx#nover+0x4e stack: 0x4305d3836de0, 0x4305d36b1320, 0x0, 0x4305d3836af0, 0x4305d3836de0
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1bd80:[0x418025a840c0]J3PerformPrePostCommitActionsInt@esx#nover+0x78 stack: 0xc1d00014, 0x4305d3836af0, 0x0, 0x418025a87e34, 0x0
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1bdb0:[0x418025a87e34]J3_CommitTransactionWithErr@esx#nover+0x8f0 stack: 0x4305d36ca8b0, 0x4305d36f10b0, 0x0, 0x4390d71a7100, 0x0
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1be60:[0x418025a88622]J3_CommitTransaction@esx#nover+0x6 stack: 0x0, 0x418000000000, 0x4305d385e6e8, 0x856, 0x7050ce0000011003
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1be70:[0x418025b148a6]UnmapAsyncClusterProcessing@esx#nover+0x37e stack: 0x4305d385e6e8, 0x856, 0x7050ce0000011003, 0x0, 0xc0000000
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1bf50:[0x418024cca4de]helpFunc@vmkernel#nover+0x4b6 stack: 0x430099ae2050, 0x0, 0x0, 0x0, 0x46
2019-04-16T16:32:44.742Z cpu3:66390)0x4390dab1bfe0:[0x418024ecb675]CpuSched_StartWorld@vmkernel#nover+0x99 stack: 0x0, 0x0, 0x0, 0x0, 0x0
2019-04-16T16:32:44.764Z cpu3:66390)ESC[45mESC[33;1mVMware ESXi 6.5.0 [Releasebuild-11925212 x86_64]ESC[0m
#PF Exception 14 in world 66390:Unmap Helper IP 0x418025b0b55e addr 0xb8
PTEs:0x186d812027;0x14a30b027;0x0;

Note: The preceding log excerpts are only examples. date, time and environmental variables may vary depending on your environment.

Environment

VMware ESXi 6.7.x
VMware ESXi 6.5.x

Cause

This issue doesn't happen every time there's unmap activity. This happens whenever there's background unmap going on and host is either rebooted or the volume is closed  or  unmounted.

Resolution

This issue is resolved in:
VMware ESXi 6.5 Patch 04. 
VMware ESXi 6.7 U3. 


Workaround:
Ensure there is no background unmap going on while the host is either rebooted or the volume is closed or unmounted.

For example : Do not try to delete a large  file on a datastore immediately before rebooting the host.