Windows VMs are crashing into BSOD 0x50 when running on a specific ESXi host

Products

• VMware vSphere ESXi

Issue/Introduction

• Virtual machines with any Windows version are crashing with a Bluescreen code 0x50 when running on a specific ESXi host
• The same VMs do not experience similar issues when running on other hosts
• Microsoft as the operating system vendor was involved and pointed out that the crashes were caused by a bit flip

Environment

VMware vSphere ESXi 6.x/7.x/8.x

Cause

A bit flip is an unexpected state change of a single bit in memory. This is not a VMware-specific problem, but indicates a hardware-problem, usually for the physical memory of the host or for one of its CPU(s)

Resolution

This is not a VMware issue. 
Please reach out to the hardware vendor to thoroughly test the server hardware and replace the defective part.

Additional Information

BSOD code 0x50 means PAGE_FAULT_IN_NONPAGED_AREA. https://learn.microsoft.com/en-us/windows-hardware/drivers/debugger/bug-check-0x50--page-fault-in-nonpaged-area describes potential causes for such issues:

• Installation of a faulty system service
• Faulty driver code
• Antivirus software using a faulty filter driver
• A corrupted NTFS volume

However, each of these can be ruled out by moving the virtual machine to a different ESXi host. If the issue does not persist after the move, any potential issues within the guest OS can be ruled out, and if shared storage is used in this scenario, NTFS corruption can be ruled out as well, as the virtual machine will still access the same virtual disk on the same datastore when running on its new host.