this post was submitted on 30 Nov 2023
1 points (100.0% liked)

Homelab

371 readers
9 users here now

Rules

founded 1 year ago
MODERATORS
 

Hey everyone! Hoping someone came across this before. Googling didn't help much. I might not be thinking correctly here.....

So I run EAXI 7 hosts in my home lab, and have been running into new VM builds corrupting. Windows 10 and windows server 2022 constantly has errors after first installation. It's always the same issues, BSOD or failed updates. The usual attempts at repairing the windows image fails. I'm completely out of ideas. I have another 13 windows 2022 servers running happily without issues. They update completely fine!

I thought maybe the ISOs were corrupted, so i downloaded new ones of each, no dice.

Anyone running into the same issues? This is happening on all of my 2 hosts running the same ESXI version. Photo of my lab for fun

top 10 comments
sorted by: hot top controversial new old
[–] Financial_Guess6458@alien.top 2 points 1 year ago (1 children)

Which version of 7 are you running? I'm aware of issues with the vmtools and 2022. I'd look at the VMs themselves - if you're using paravirtual driver that could be of issue - and it's not easy to fix. Update to the latest 7 vsphere, then your esxi hosts, then update the tools on the affected hosts in safe mode.

Another thing to check - if you can boot them into safe mode, its a virtual layer driver causing the hub bub.

This wouldn't be a hardware level thing cause you'd get exceptions or callbacks from vmware before the OS gets a chance to load and BSOD. Hope this helps.

Btw: where did you get your rack? I'd love that setup!

[–] thatfrostyguy@alien.top 1 points 1 year ago

I'll check those out! Thx!

Here's a link to the rack I use SysRack

[–] AdderallBuyersClub2@alien.top 2 points 1 year ago

Check iDRAC or iLO on your servers. They are able to pick up memory and cpu errors on the VMs.

You wouldnt happen to have energy saving features enabled would you?

[–] NorthernDen@alien.top 1 points 1 year ago (1 children)

did you try the install with nothing but bare minimum? So no nics, only one hdd presented to the guest (just for the c drive), usb 2 only.

Nothing different about the iso, so nothing slipstreamed in?

If you clone one of the working machines, then do a sys prep on it (so treats it like a new install) does this work?

I'm going to be offline for a few days, so may not respond right away.

[–] thatfrostyguy@alien.top 1 points 1 year ago

I'll clone and run a sysprep on it, see if that makes any difference. I got the ISOs directly from microsoft

[–] TheButtholeSurferz@alien.top 1 points 1 year ago (1 children)

Where did you source the ISO's from.

I realize you've tried and used multiple ones. I'm just curious, because I've ran both 7 and 8 clusters and I have never seen this issue except on poorly sourced or broken images.

[–] thatfrostyguy@alien.top 1 points 1 year ago

Sourced directly from Microsoft

Granted they are ms evaluations

[–] Capt_Offended@alien.top 1 points 1 year ago

Any storage connection issues? I’m running iSCSI on a synology without any issues.

[–] davedunaway@alien.top 1 points 1 year ago (1 children)

Do you have EVC mode on for your cluster? Your hosts look like they could be pretty different. Is it possible your VM boots up or migrates to the other host with a much different CPU?

[–] thatfrostyguy@alien.top 1 points 1 year ago

No ability for it to migrate. No vsphere. Just two independent hosts. Although I can manually migrate the VMs by unregistering and reregistering on the other host.