toastman

Try i440fx/Seabios with the emulated CPU <cpu> <topology sockets='1' cores='2' threads='1'/> </cpu>

I had this exact problem, Its caused by something wrong with how the CPU is handled. If you change the CPU to an emulated QEMU CPU, or emulate a Skylake processor your problems should be solved. Check that thread. About halfway down on the first page

Can you go into your BIOS boot options and manually select your boot device

Interesting. So it was more likely a sound problem than a video problem. Maybe there is something in how the USB sound interacts with the USB card... Odd. But at least you got it working. I would think that, all things being equal, passing through the dongle, versus passing through a whole card should have zero difference in sound quality itself - only in operation. You've discovered there is a difference - it doesn't work one way.

Not sure yet. Still some steps to do, but at least you know its not your system, unRAID, your VM, etc. And the fact that you say Ubuntu works properly says its less likely your GPU - but it still may be... I'd suggest creating a new VM. Same details as before, but passthrough your GPU and a keyboard/mouse. Nothing else, and no VNC. Presuming you have a monitor attached it should come up right away. If it does, install Windows - again, as bone-stock as possible. Don't install Nvidia drivers - just let Windows do its thing. One it's done, use Edge to watch a Youtube video. See if anything changes.

Any chance its disabled in BIOS?

Try going right back to basics: Create a new Windows VM. OVMF, Q35-4.2. Give it 4 cores and maybe 8GB of RAM. Makes sure its on your cache. And nothing else - just as bone-stock as you can get it. Install windows and use VNC or RDP to access. Play a video - it'll run like garbage (probably) but you just want to see if it runs for longer than 10 minutes.

As of 6.8.2 this is still an issue, although I can confirm using the emulated CPU still works

I love how easy it is to spin up a VM and add dockers I would love to see a little more granular control of VM settings and not not have to directly edit xml for most VMs I create

I love how easy it is to spin up a VM and add dockers I would love to see a little more granular control of VM settings and not not have to directly edit xml for most VMs I create

No, the Device Manager within the Windows VM itself. On first install I have about 6, but I can delete 4 of them without losing anything - the other two have my mouse/keyboard etc., attached. Delete those and setting HPET to yes drops my idle usage down from ~15-20% to 1-2%, and completely solves my audio sync problems

I had a similar issue caused by latency. What fixed it for me was to remove all unused USB root hubs in the VM device manager, then in the XML set HPET to present=yes

Yay

One of the reasons for high CPU usage is that Windows seems to throw a lot of interrupts. I've had mixed success with enabling HPET - sometimes it works, and sometimes it doesn't. What I have found that works all the time is disabling the unused USB Root Hubs. CPU usage will drop to nearly nothing at idle when I do that.

The problem is that I/O wait is CPU usage. While the CPU is waiting on I/O it can't do anything else. So when its showing 100%, even though its not doing anything productive, you can't do anything else with it.

An E3-1240 v2 has the gaming processing power of an i7-2600k. That's baremetal. You're giving it 75% of baremetal cores and thats before any VM-related overhead. You've got a great GPU that should easily handle fortnite, but you're seriously CPU constrained. I think its entirely possible you've reached the limits of your CPU. Try opening up Task Manager in your VM, select the "Performance" tab, and leave it open while you run Fortnite. If your CPU is pegged but your GPU is barely moving you've found your problem.

Hello, Something I've noticed since upgrading to 6.7 (on 6.7.2 now) is that when I open up the "main' page of the webui it takes a solid 2-3 seconds before the cache section populates with my cache drive. Everything else is there right away just like it always has - Dockers, VMs, arrays. If I navigate to a different tab then come back, the delay is still there. All my plugins are up to date, and everything else works just fine. Anyone else noticed this?

I went through the device manager and disabled all of the unused USB root hubs based on an obscure post I found elsewhere. CPU usage dropped to normal idle levels. Odd

This is the issue where the task manager within the VM shows nearly zero percent, as it should at idle, while the VM Dashboard shows ~25% usage. I suppose I should have mentioned that in my original post.

So I've been creeping everything I can find about KVM/QEMU/etc to try and figure this out. I have a consistent ~25% idle usage right now on the unRaid dashboard on a single core (of 6 passed through) and between 5-10% on the other 5. According to various posts around here, reddit, proxmox, etc., its caused by a windows interrupt issue. Tried everything I can find and spun up a bunch of VMs to get it diagnosed. Finally was successful with spinning up a VM with Hyper-V enhancements OFF, HPET='yes', Q35, OVMF, updated to 1809. Idle usage sitting at 1-3% on a single core and the rest barely tickling. And it stayed like that for 8 days. Then it went back to ~25%. Nothing changed near as I can tell: it just idled up. So I fired up a new Windows 10 image to test using an Arch template. HPET='yes'. Updated to 1809. Idles right down at ~3% within a minute of boot. This does not affect any of my other VMs (Manjaro, Ubuntu, Windows7). Only one VM is on at any given time. Anyone have any idea what is going on? Specs: unRAID 6.7 Ryzen 2700. Cores 4-7 w/HT isolated for VMs 32GB RAM - 8GB for Win10 VM ASUS B450-F Gaming GTX 950 for passthrough VM gets 3 isolated cores with respective hyper-threads

I'll contribute as well. Asus B450-F Gaming w/ ryzen 2600. unRaid 6.7.0 stable Stared with version 2008 - everything working Upgraded to v. 2301 - 127 / D3 problem Downgraded to 2008 - everythig working again.

Asus B450-f Gaming Ryzen 2600 GT720 (for unRaid) GTX950 (for passthrough) In trying to diagnose a number of errors (namely a 127 error affect one my my PCIe slots), as well as get Primary-GPU passthrough working, I noticed there was a BIOS update for my motherboard that included the new microcode for the 3-series ryzens, etc. So I updated. After that, no VMs needing GPUs would boot (127 / D3 error). Not only did a restart of the server not clear the problem, it never actually restarted. Watching the shutdown process I noticed it went through everything appropriately but the shutdown signal never actually turned the server off; I had to do the push-and-hold of the power button to actually get it to power off. After a bunch of messing around that didn't fix the problem I went back to the previous BIOS (2008) and all the problems disappeared. Luckily Asus' ez-flash allowed 'updating' to an older version. I suppose there's something between the BIOS and unRaid that isn't playing nice thats causing these problems. I don't know enough about much of it to offer any other conclusions than that.

I have a problem similar to this and it happens anytime I try to passthrough a PCIe device that's on my 3rd PCIe x16 slot. GPU, NIC, USB card - doesn't matter, anything in that slot. If I move that card to a different slot and pass it through, problems disappear. It seems that slot is controlled by the chipset rather than directly by the CPU. Specs: ASUS B450-F Ryzen 2600

If you haven't stubbed it out, unRaid detects it as a network device and keeps it for itself. I imagine if you went to Settings->Network Settings you will see it listed as an available network device (whether you can use it or not). You'll need to add (without quotes) the following into your syslinux config "pci-stub.ids=168c:002e" then restart the server If you don't understand what the above means, I suggest you do some watching of Spaceinvaderone on Youtube.

Yeah, USB hand-off and the like. I did try that