Can't install some apps or unzip big files on Windows 11 VM

[brnogu]

Hello everyone! I would like to ask your help for a very strange behavior that I'm seeing.

I have several Windows and Linux machines running on Proxmox. Every machine is working great, with the exception of one.
Already tried to delete it and recreate a lot of times but the issue persists. Also tried to unlink the GPU, but the same thing happens.

If I have a big .zip file, when I try to extract it, I get an error informing that the file is corrupted.
When I install some games, it fails during installation, and when I check the logs it says something like: HashMismatchError: Expected: 12345 <-> Actual: 54321 (Just an example).

At first I thought that it would be the physical disk, so I have created a new virtual disk on another physical disk, install the game on that location (Or download and zip the file to that location) and still the same issue.

Do you have any idea of what might be wrong here?

Proxmox version: 7.4-3

VM configuration:

agent: 1
balloon: 0
bios: ovmf
boot: order=virtio0;ide0;net0
cores: 5
cpu: host
efidisk0: Samsung_2TB_SSD:vm-104-disk-1,efitype=4m,pre-enrolled-keys=1,size=4M
ide0: local:iso/virtio-win-0.1.229.iso,media=cdrom,size=522284K
machine: pc-q35-7.2
memory: 12288
meta: creation-qemu=7.2.0,ctime=1679202233
name: GamingVM
net0: virtio=2E:3E:A1:BA:72:79,bridge=vmbr0
numa: 1
onboot: 1
ostype: win11
scsi0: Crucial_500GB_SSD:vm-104-disk-0,backup=0,cache=writeback,size=250G,ssd=1
scsihw: virtio-scsi-pci
smbios1: uuid=9417cc0b-284c-457f-8c66-c0379f047e83
sockets: 1
tpmstate0: Samsung_2TB_SSD:vm-104-disk-0,size=4M,version=v2.0
virtio0: Samsung_2TB_SSD:vm-104-disk-2,aio=native,backup=0,cache=directsync,size=768G
vmgenid: f0963086-b9a8-4cb6-a3cf-76aa7c106479
hostpci0: 0000:01:00,pcie=1,x-vga=1

 

[Dunuin]

I would boot memtest86+ on your PVE server and see if you got a bad RAM module. The last three times checksums of files didn`t matched it was a failing RAM module silently corrupting all data for weeks.

 

[brnogu]

I would boot memtest86+ on your PVE server and see if you got a bad RAM module. The last three times checksums of files didn`t matched it was a failing RAM module silently corrupting all data for weeks.

That's a good idea and it actually make sense. Do you know if memtest86+ can point to the specific module? Or I need to take them out and run one by one?

I will try to do this in a few hours

 

[Dunuin]

It will show you the address in memory that failed but not the module. I just removed/swapped DIMMs until I found the bad one.

And in case it is actually a failing RAM. Hash all your data and compare it to some weeks old backups. ZFS scrubs will tell you that eveything is fine even if it is not.

 

[brnogu]

Hey! Just run the RAM scan. It output errors right after the test start. With a few remove and add I was able to identify the problematic memory.

I tested it alone and:


I'm buying another one tomorrow to replace it. The other 2 that I had also installed pass without any issues
I have also tried to unzip and install the games and now everything goes as it should!!!

Thank you very much for the help!

 

[Dunuin]

Glad it helped. But really try to check your data. Here it corrupted hundreds of GBs of data over weeks, when I actually recognized it because of increasing crc/checksum errors or bluescreens. Look for photos where the bottom half of the picture is missing, videos that stop working in the middle and so on.