gltovar

Actually one thing problem I am having, I cannot passthrough my unraid gpu to a VM. In the past all I needed to do was specify a (hex edited ala SpaceInvaderOne) GPU bios dump and it would work. Now I am getting that classic "first pinned cpu maxes out at 100%" freeze and have to force stop the VM. I did try some rudimentary trouble shooting, but I'll check the prior RC threads to see if it has been covered before. I am attaching a diag zip. rc7_cannot_passthough_unraid_gpu.zip

Works for me upgrading from 6.6.7 What I am hoping to test out with the RC is stability for 5 (or worst case scenario 4) gpus (gtx1080s) running on one system with windows 10 VMs. I was doing some stress testing as I get the VMs really working, I start to see video artifacts (short horizontal white lines randomly on the windows desktop) and then one of the game clients starts to have serious issues rendering the 3d environment. I have done a decent chunk of trouble shooting thinking it was one or two GPUS but turns out they seem fine. Some more google searches turn up that people setting the VM to use Q35 resolves issues. So with this update i'll try not to modify anything, then switch things to Q35 and I'll report my findings to the thread linked below, in an attempt to also respond back to jonp's request for feedback: "Feedback is critical here. We are seriously considering removing the option to select Q35 for Windows-based VMs from VM manager. If you can provide a use-case for why we shouldn't, please do!!"

on 6.6.7 and I get this when trying to update to this version here is my output: plugin: installing: /var/tmp/unRAIDServer.plg plugin: downloading: https://s3.amazonaws.com/dnld.lime-technology.com/next/unRAIDServer-6.7.0-rc6-x86_64.zip ... done plugin: downloading: https://s3.amazonaws.com/dnld.lime-technology.com/next/unRAIDServer-6.7.0-rc6-x86_64.md5 ... done wrong md5 plugin: run failed: /bin/bash retval: 1 I've attached my diagonostics. Probably something dumb I am forgetting tower-diagnostics-20190404-0041.zip

I am having the same issue. I am running a machine with multiple GPUS. 4 out out the 5 are working fine. I can likely reinstall this one from scratch and get it working. Before I do that I'll point out something I noticed about the ones that work, vs the one that does not work: The ones that work have "alias names" for the gpus example: <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x6e' slot='0x00' function='0x0'/> </source> <alias name='hostdev0'/> <!--- THIS ---> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </hostdev> <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x6e' slot='0x00' function='0x1'/> </source> <alias name='hostdev1'/> <!--- THIS ---> <address type='pci' domain='0x0000' bus='0x00' slot='0x08' function='0x0'/> </hostdev> And the one that doesn't work, does not have the alias tags, and if I manually ad them, they do not persist after updating the xml. Hopefully this offers a clue to fixing this problem.