ESXi host crash due to frequent PSOD (purple screen of death) with different backtrace for each crash

Products

• VMware vSphere ESXi 8.0

Issue/Introduction

• ESXi crash frequently without any unique functions in the backtrace, each crash might show different backtrace.
• Following are some of the observed backtrace on the same ESXi host:
  
  PSOD - 1
  

  VMware ESXi 8.0.2 [Releasebuild-23825572 x86_64]
#PF Exception 14 in world <WorldID>:hostd-worker IP 0x42000741d1ff addr 0x420000000000
PTEs:0x20bcc2c023:0x20a6537023: 0x0;

  Module(s) involved in panic: [vmkernel Version Releasebuild-23825572]
*PCPU30:<WorldID>/hostd-worker
PCPU 0: UUUUUISSSUUISSIIIISSSUSUSUUUUUUSSSSS
[0x42000741d1ff]DeviceDupID@vmkernel#nover+0x53
[0x42000741e875]Device_GetDeviceIDOvmkernel#nover+0x4a
[0x4200077f4b9e]PCI_GetPCIDeviceFromVMKDevice@vmkernel#nover+0x2b
[0x4200074bde28]vmk_PCIGetDevice@vmkernel#nover+0x15
[0x4200079119ce]PsaStorVsiAdapterInfoGet@vmkernel#nover+0x40f
[0x420007401f49]VSI_GetInfo@vmkernel#nover+0x192
[0x4200079f0a0b]UW64VMKSysca1lUnpackVSI_Get@vmkernel#nover+0x168
[0x4200079841c8]User_UWVMK64SyscallHandler@vmkernel#nover+0x129
[0x420007564c54]Sysca11UWVMK64@vmkernel#nover+0x90

  PSOD - 2

VMware ESXi 8.0.2 [Releasebuild-23825572 x86_64]
#PF Exception 14 in world <WorldID>:HELPER_MISC_ IP 0x4200382516fa addr 0x3ea228
PTEs :0x20ab3a8023:0x14426f023:0x0;

  Module(s) involved in panic: [vmkernel Version Releasebuild-23825572]
*PCPU30:<WorldID>/HELPER_MISC_INFRA_QUEUE
PCPU 0: SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
[0x4200382516fa]dict_repeat@vmkernel#nover+0x42
[0x420038251d0f]lzma_main@vmkernel#nover+0x5d4
[0x420038252735]xz_dec_lzma2_run@vmkernel#nover+0x6be
[0x420038252e06]xz_dec_run@vmkernel#nover+0x28f
[0x420037c17af2]Decompress_XZSingle@vmkernel#nover+0x8b
[0x420037c1940b]BootModuleDecompressModule@vmkernel#nover+0x1b0
[0x420037cf303c]HelperQueueFunc@vmkernel#nover+0x19d
[0x42003802c149]CpuSched_StartWorld@vmkernel#nover+0xe2
[0x420037cdbe7f]Debug_IsInitialized@vmkernel#nover+0xc

  PSOD - 3

VMware ESXi 8.0.2 [Releasebuild-23825572 x86_64]
#PF Exception 14 in world <WorldID>:storageRM IP 0x42002c965fc6 addr 0xffffffffffffff95
PTEs :0x0;

  Module(s) involved in panic: [vmkernel Version Releasebuild-23825572]
*PCPU30: <WorldID>/storageRM
PCPU O: SSUSUUIISIIIISSSUSSSUUIIUUIIUIUSIUSI
[0x42002c965fc6]StorageLookupPathByName@vmkernel#nover+0x8a
[0x42002c8e51e4]StorageDevicePathsIteratorNext@vmkernel#nover+0x105
[0x42002c8aa2a5]PsaStorVsiPerDevicePathList@vmkernel#nover+0x86
[0x42002c40297b]VSI_GetListInfo@vmkernel#nover+0x1d4
[0x42002c9f030f]JUW64VMKSyscal1UnpackVSI_GetList@vmkernel#nover+0x214
[0x42002c9841c8]User_UWVMK64SyscallHandler@vmkernel#nover+0x129
[0x42002c564c54]Sysca11UWVMK64@vnkernel#nover+0x90

Resolution

This may be a hardware problem as the host is crashing with different backtrace for each PSOD; please contact your hardware vendor to diagnose the hardware.

Replacing the CPU has helped to resolve the issue in similar failures.