TType85

I am having an issue provisioning a SSL cert on my server. When I hit the button to provision I get this in the console of the web browser. When I set it to Auto for SSL/TLS it did generate a self signed cert. I am on 6.9.1; diagnostics are attached. tower-diagnostics-20210406-2022.zip

I think you got it there. Mine was 2.11 now 3.0. That was the only change I could find.

It was similar to jerkey_san showing 2/2/16/16 way instead of the correct 8/4/8/16 way. It does seem to have smoothed out the VM a bit. World of warcraft now feels more like it does bare metal. Just as a warning though, I did have to re-activate my windows 10 licenses.

Wow runs a little bit smoother, not much frame-rate improvement but it *feels* a bit smoother.

I have been having issues getting the cache to show up right using this tweak. I ended up creating a new VM and the cache showed up correctly; so I created a new VM XML for my existing windows install and now it shows correctly. I will have to test games when I get home.

Sonarr + NzbGet dockers are what I use. It has been working great for a long time.

Mine feels great. I had some weirdness running WoW and youtube in chrome at the same time but now that seems ok. Will have to play with it more though.

Yeah, really odd. The Wifes VM usually starts first. I would assume the numatune setting would push it over to node 1, when mine starts it would be in node 0. Will have to wait till tomorrow to play with it more. Numastat -c below, that numa_miss....

Unraid sees all 32GB. Currently at 62% used. 64GB upgrade is on the list but will have to wait till after the holidays.

32GB total, 8GB per vm, 2 VM's. I do have a bunch of dockers too.

Wifes VM is set to strict, nodeset 1, mine was strict nodeset 0 but I rebooted the server and my vm wouldn't start because it ran out of memory. I had to take out the numatune to get my vm up. Edit, both are set to 8GB of ram

Currently there are 4 dimms installed. What command shows the per-node usage?

Oddness, on my wife's VM memory bandwidth is half and latency is 50% higher. All settings the same except she is on the other side of the chip. CPUz scores are the same (better than they were before, in the 410 range instead of the 360 range)

I set mine to Channel this weekend and noticed an increase in performance in WoW. I just added the numatune to my configs pointing to the correct nodeset (wifes VM is on node 1, mine is on node 0). I'll have to wait till I get home to check gaming performance. since I am RDP'd in right now. If I can get it to read the cache right maybe it will get even better.

I just tried the settings and I still get the same as the Virtual Old. My latency for Memory and L1 are close to yours but the write/copy are about half. This VM only has 8 vCpu (4c/8t). The L2 and L3 don't even run, I am on the trial version of aida64 so maybe that is why they don't show?

Sounds like your hardware is not stable. Try installing windows directly, no unraid, and runs some stress tests. I have had zero issues running unraid VM's on multiple hardware platforms (Intel and AMD).

I had an issue with having 2 of the same USB card being passed through. If you can, try removing one of the cards and see if the other passes through.

are the cards set as virtio or did you specify e1000? I know for OSX you need to change the type for it to work. <interface type='bridge'> <mac address='xx:xx:xx:xx:xx:xx'/> <source bridge='br0'/> <model type='e1000-82545em'/> <address type='pci' domain='0x0000' bus='0x09' slot='0x01' function='0x0'/> </interface>

this is how one of my cards (the primary card) is passed through to a VM with the bios. Once it starts the VM, unraid will not be able to take it back until reboot. <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x49' slot='0x00' function='0x0'/> </source> <alias name='hostdev0'/> <rom file='/mnt/user/isos/BIOS/GP106-1060.rom'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </hostdev>

Is this the only GPU in the system? Are you passing the Bios through if it is?

Are you assigning to all four of your cores? Are you leaving core 0/4 for Unraid? Try it with core 0/4 not used for the VM.

you need to go in to your OVMF bios (hit esc on the bios screen) and set the resolution to 1920x1080 check out SpaceinvaderOnes video:

Are you up to date on the windows 10 updates? The GPU in task manager was added in the last big update I think.

I don't think RDP uses the graphics card in windows to display, it uses it's own driver. If you use something like teamviewer it will use the GPU but there still is latency.

SeaBIOS vs OVMF: I think SeaBIOS is used for older OS that do not support UEFI and does not play as well with video card passthrough. I440fx vs Q35: i440fx is a older, more supported hardware profile for windows, Q35 is a newer one that works best in Linux (and osx). Your code 43 issue is because Nvidia does not want you to pass your video card to a VM. I think KVM can hidethe HyperV flag so that the video card does not see it and it works. FWIW: I have 2 Windows 10 VM's running as i440fx/OVMF with Nvidia GTX 1060 & 1070 cards passed through along with a usb card and nvme ssd with no issue. I also have a OSX VM that is Q35/OVMF with the GTX 1070 passed to it along with the USB card and it works well also.