Windows worker nodes get bluescreen

    Since upgrade from OCP 4.11 to 4.12 and also 4.13 we are experiencing random bluescreen (BSOD) on windows worker nodes. There was a suspicion that BSOD happens after reboot, that's why we established hourly drains and reboots of windows worker nodes. During the week we have already received second BSOD from which we have memory crash dump. Bluescreen reports NDIS.SYS. The only known way to fix the node is to reinstall it, or restore from backup. BSOD can be reproduced on affected node by reboot. It's even possible to start affected worker node in non-network mode and invoke the bluescreen by starting network or kubernetes based services, e.g. kube-proxy, containerd, or network list service. Attaching crash dump and must-gather. Define the value or impact to you or the business causing long downtimes because node is not fixable by reboot .

Microsoft Support Case
2402200030004965

When does this behavior occur? Frequency? Repeatedly? At certain times? 

    8.1.1 WMCO - Windows Server 2022 Standard

    Repeatedly, randomly, after reboot of windows worker node.

    1. Reboot always produce a BSOD in customer environment.

    Having to reinstall the node.

    Not having to reinstall the node and not receiving BSOD.

    memory dump with analyse -v of windbg output
*****************************************
*                                                                             *
*                        Bugcheck Analysis                  *
*                                                                             *
*****************************************

SYSTEM_SERVICE_EXCEPTION (3b)
An exception happened while executing a system service routine.
Arguments:
Arg1: 00000000c0000005, Exception code that caused the BugCheck
Arg2: fffff80088462e79, Address of the instruction which caused the BugCheck
Arg3: ffffb88828952150, Address of the context record for the exception that caused the BugCheck
Arg4: 0000000000000000, zero.