We are running W2012 R2 guests on ESXi 5.0 U2 hosts. Guests fail and reboot often with Windows event 1001 Bugcheck stop 0x00000109. We created a CPUID mask on these
guests as per VMware KB 2060019. That seems to prevent the failures as long as the guest stays on the same host. If we vmotion or there is a HA or DRS execution the guests end up
running on different hosts. The bugcheck then starts happening again. All hosts are the same hardware and BIOS settings are consistent. We don't understand why the CPUID mask
made via vcenter to edit the guest settings no longer works. If we migrate the guests back to where they were - then all works OK again. Anyone experience this condition ?
Based on the KB 2060019 this issue is resolved on vSphere ESXi 5.0 Update 3, this way, there is any chance to upgrade this host instead of use the workaround provided on KB ?