ESXi hosts might fail with purple diagnostic screen due to rare memory fragmentation issue

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms -

• Under rare conditions of high host memory utilization, severe memory fragmentation may cause delays in network device driver memory allocation requests, which can result in a heartbeat miss and trigger a PSOD.
• The PSOD backtrace typically includes entries similar to - 
  
  yyyy-mm-ddThh:mm:ss.Z cpu2:267947)Panic: 547: Panic from another CPU (cpu 2, world 267947): ip=0x420013cf7434 randomOff=0x13c00000:
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)Saved backtrace from: pcpu 2 Heartbeat NMI
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951b990:[0x420013d42528]FastSlabAllocPages@vmkernel#nover+0xf9 stack: 0x420013d424f0
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951bc20:[0x420013d429a8]FastSlabCreateNewBuffer@vmkernel#nover+0xe9 stack: 0x45388951bce0
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951bc80:[0x420013d43600]FastSlabCreateObj@vmkernel#nover+0x4cd stack: 0x100000000
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951bd30:[0x420013d43e2a]FastSlabReplenishCPU@vmkernel#nover+0x22f stack: 0x430074602ae8
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951bd80:[0x420013d4510f]FastSlabAllocSlow@vmkernel#nover+0x84 stack: 0x9
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951bda0:[0x420013dded5e]Pkt_SlabAllocPkt@vmkernel#nover+0x133 stack: 0x1f
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951bde0:[0x420013dde281]Pkt_AllocHandleWithSize@vmkernel#nover+0xee stack: 0x4520c01a8850
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951be00:[0x420013dde570]Pkt_AllocWithFlags@vmkernel#nover+0xd stack: 0x45388951bf20
yyyy-mm-ddThh:mm:ss.Z cpu0:400001)0x45388951be20:[0x420013e82f71]vmk_PktAllocForDMAEngine@vmkernel#nover+0x46 stack: 0x0
  
  

Environment

• VMware vSphere ESXi 7.0.0

Resolution

• This issue has been fixed in VMware ESXi 7.0 Update 3f | ESXi 7.0.3 P05 (Build 20036589).