ESXi host fails with PSOD #PF Exception 14 in world xxx:hostd-worker IP addr 0x28

Products

• VMware vSphere ESXi

Issue/Introduction

• The ESXi might fail with a purple diagnostic screen

  #PF Exception 14 in World xxxxxx:hostd-worker IP 0x42002c916def addr 0x28
  PTEs:0x0;
  
  Module(s) involved in panic: [deltadisk Version 1.1, Built on: xxxxx
  Code start: 0x42002bc80000 VMK uptime: 3:09:56:81.886
  Đx4539348 1b178: 18x12082c916def SESparseCommit@esx#nover + 0x0 stack: 8x12d8cd00000
  base fs=0x0 gs=0x128044000008 Kgs=0x0
  <timestamp> cpu4:2097524)Jumpstart plugin nmp-nimble_psp_directed-rules activation failed.

   

• We observed performance latency(performance has deteriorated) issues right before the crash of the ESXi host

  YYYY-MM-DDTXX:XX:XX.999Z Wa(180) vmkwarning: cpu13:2097756)WARNING: ScsiDeviceIO: 1772: Device naa.XXXXXXXXXXXX performance has deteriorated. I/O latency increased from average value of 187 microseconds to 10237 microseconds.
  YYYY-MM-DDTXX:XX:XX.154Z In(182) vmkernel: cpu8:2097757)ScsiDeviceIO: 1772: Device naa.XXXXXXXXXXXX performance has improved. I/O latency reduced from 30049 microseconds to 5920 microseconds.
  YYYY-MM-DDTXX:XX:XXZ In(182) vmkernel: cpu19:2097753)ScsiDeviceIO: 1772: Device naa.XXXXXXXXXXXX performance has improved. I/O latency reduced from 10237 microseconds to 3867 microseconds.
  YYYY-MM-DDTXX:XX:XXZ Wa(180) vmkwarning: cpu13:2097756)WARNING: ScsiDeviceIO: 1772: Device naa.XXXXXXXXXXXX performance has deteriorated. I/O latency increased from average value of 187 microseconds to 9273 microseconds.
  YYYY-MM-DDTXX:XX:XXZ In(182) vmkernel: cpu19:2097753)ScsiDeviceIO: 1772: Device naa.XXXXXXXXXXXX performance has improved. I/O latency reduced from 9273 microseconds to 3743 microseconds.
  YYYY-MM-DDTXX:XX:XXZ Wa(180) vmkwarning: cpu13:2097753)WARNING: ScsiDeviceIO: 1772: naa.XXXXXXXXXXXX performance has deteriorated. I/O latency increased from average value of 187 microseconds to 8399 microseconds.

Note: The preceding log excerpts are only examples. Environmental variables and memory addresses may vary depending on the environment

Environment

• VMware vSphere ESXi 7.x
• VMware vSphere ESXi 8.x

Cause

Slow storage is causing the consolidation to take lot of time and this resulted in sesparse clearing/freeing the consolidation task while the task is getting accessed from other threads, causing this PSOD.

 

Resolution

This issue is resolved in the following 8.0 u2b ESXi version

8.0u2b Release Notes