Hyper-V VM blue screen

[RahamimL]

Hi everyone,

I have a terminal server VM which is part of a 2019 RDS farm.

The server is 40 giga memory with 16 cores and 320 GB disk (100 of them are free to make sure that a dump file can be created just in case.) I also have another 100 GB disk to dump the .DMP file.

I also configured the NUMA of the VM:

• max processors per V-NUMA node - 16
• Max memory per V-Numa - 20GB
• Max V-NUMA nodes per socket - 4

Up until the past 2 days we had minimal crashes. The server does have more than 150 profiles opened, and we do use OneDrive on the server installed to all users. We do have SAP and SPSS installed on the TS along with Office 365 shared license.

The reason I mention all these things is to maybe help point me in the right direction.

Now for the problem - Every time the server crashes I have a Hyper-V Worker error 18950 with the reason for the blue screen - "has encountered a fatal error. The guest operating system reported that it failed with the following error codes: ErrorCode0: 0x3B, ErrorCode1: 0xC0000005, ErrorCode2: 0xFFFFF8012FA65284, ErrorCode3: 0xFFFFD908E25FDD80, ErrorCode4: 0x0. If the problem persists, contact Product Support for the guest operating system".

When I do catch the blue screen, it reports something about cldflt.sys. Looks like a service issue which I can't figure out.

The dump file is 4 GB and when I try to debug with WINDBG it says the file is corrupt.

If someone has an idea other than rebuild the server. I would be more than happy to hear.

I can upload the dump file if someone thinks they can do something with it.

Thanks in advance. Rahamim.

[#2]

Hello,
What do you mean by "150 profiles opened" ? 150 simultaneously connected users ?

[#3]

Sorry, I meant there are more than 150 profiles in the server. Most of the time we have 30-40 users connected simultaneously. I do a lot of maintenance on the servers (Deleting old profiles, defragging the disks, using OD cleanup policies and storage sense.)

[#4]

cldflt.sys could point towards a OneDrive-related issue.
First, make sure all your hypervisors, guest VM and guest VM tools are fully updated with Microsoft updates. Check possible updates for hypervisor drivers too.
Do you have an issue on others servers, or just one ?

[#5]

@Alban1998 Thanks for the reply, I was afraid you'd say that.

This issue could happen on the second server on our farm, but the issue is random.

If that is the case and since we keep our environment fairly updated, we might have a problem where someone did an action with OneDrive that might cause a BSOD... 

[#6]

Just found this on the dump:
SYMBOL_NAME: cldflt!CldiStreamCancelSynchronousRequest+20

MODULE_NAME: cldflt

IMAGE_NAME: cldflt.sys

STACK_COMMAND: .cxr 0xffff8502c5e8dd80 ; kb

BUCKET_ID_FUNC_OFFSET: 20

FAILURE_BUCKET_ID: AV_cldflt!CldiStreamCancelSynchronousRequest

OS_VERSION: 10.0.17763.1