steve1977

Thanks @david279. What video of @gridrunner are you referring to to handle virt-manager? Is this still really needed or is there an easy work-around within the image or with working with the xml files? Also, do you mind to elaborate a bit how you did the efti thing to migrate from the iso-disk to the local harddisk. I followed the steps in youtube, but it didn't work. What you describe above reads different and probably the way to go for me as well. Thanks again!

thanks. perfect. it also worked with 4200 for me and not with the later versions. did you follow some step to step guide to migrate from vnc to gpu. i don’t know virt manager and have never used a ubuntu vm. sounds a lot of work to get the nvidia drivers to work. can you share step by step the EFI things you went through. it reads like you faced the same issue and also couldn’t just follow the youtube guide? Sent from my iPhone using Tapatalk

quick follow up. what are kvm net drivers? all works well when gpu is set to vnc. Sent from my iPhone using Tapatalk

Great to hear @david279. What version of clover did you use and what guide did you follow to passthrough your Nvidia card? Did you succeed to "migrate" Clover to your local harddisk following the instructions in the youtube video from this thread?

This reads quite disturbing and may impact me (i7 7800X). I do have some issues these days, but I never thought they may be related to the CPU. At the same time I would anticipate that there are many Unraid users running Skylake/Kabylake CPUs. Is this issue for a real and a big deal? This thread is quite short and nobody for sure confirms a major issue. The link to the debian thread is a bit cryptic to me. Can someone advice whether my CPU is impacted and can be fixed via update or otherwise?

Thanks. If doing a straight swap, can I do it in the same HD slot? I shutdown Unraid and then take out "old" disk and add a new one to replace in the same slot. Then power up Unraid and start the array. And then the magic begins? I just realized that I will need to do 3-4 swaps: 1) Parity: 6TB to 10TB 2) Data 1: 6TB to 10TB 3) Data 2: 4TB to 6TB ("old" parity") 4) Data 3: 4TB to 6TB ("old data 1") I could save one swap going from 4TB to 10TB, but this would then require a lot of copy activity between the disks. So, 4 swaps may be needed? I think I'll take my chance when upgrading the parity disk without preclear. Worst thing that can happen is that building the array fails, bu it shouldn't bring data loss (unless one of the data disks fail at same time). For the upgrade of data 1, I will probably need to do the preclear not to risk losing data. When re-using my "old" parity disk to upgrade my data 2 disk, do I need to do anything differently or does this work same way as a new empty disk?

Thanks for your message. This is clear. If I understand you correctly, I can now make an attempt to pass through audio from my GTX1050, but you are suggesting that it would be even better to pass through all of the GTX1050 (both video and audio)? Is this understanding correct and is this possible in the meantime with High Sierra? I remember you said in one of your videos that the drivers may not be available yet? If available by now, do you have a video with the how-to as well? I remember there is plenty of complexity in passing through a primary GPU to a VM though? Also, don't want to be a pain, but I am still struggling to migrate Clover to my local harddisk. I followed all the steps of the videos (thanks again). What I don't get to work is what is being described from 16:01 in the youtube video onwards. The labels in disk utility don't match with what you show in the video. Also, only one of the two EFI partitions includes clover folder/files. Any thoughts what I am doing wrong?

+1 Sent from my iPhone using Tapatalk

Related question - any views on running two VM images on the same cache drive? Better one on cache and one on unassigned? Any material performance difference?

Got it, thanks.

Thanks for your message. This is clear and helpful. I don't have free internal slot. I may be able to do the pre-clear over USB if strongly advised. Or just skip it and go straight to parity creation?

Thanks, this is helpful. One related question about the cache pool: Assuming I reformat both disks as brfs, what are the use cases for using two cache disks? I don't think I can mirror them as they are different sizes. I don't really want to "pool" them as I don't want files across the two and keep the disks independent (allowing me to remove one of them at any time). Does my restrictions leave me with any use case for 2 cache disks?

No worries, thanks, much appreciated your help and the great video guides on youtube. Have been an active user. I have two audio cards. One is-on-board and one is part of my GPU. I have been trying to pass through the on-board one. See below IOMMU groups: IOMMU group 14 [8086:a2d2] 00:1f.0 ISA bridge: Intel Corporation Device a2d2 [8086:a2a1] 00:1f.2 Memory controller: Intel Corporation 200 Series PCH PMC [8086:a2f0] 00:1f.3 Audio device: Intel Corporation 200 Series PCH HD Audio [8086:a2a3] 00:1f.4 SMBus: Intel Corporation 200 Series PCH SMBus Controller IOMMU group 36 [10de:1c81] 65:00.0 VGA compatible controller: NVIDIA Corporation GP107 [GeForce GTX 1050] (rev a1) [10de:0fb9] 65:00.1 Audio device: NVIDIA Corporation GP107GL High Definition Audio Controller (rev a1) Any thoughts on my second question? Any more info required from me?

Thanks. My "old" brtfs disk was part of a cache pool. Assume that's the reason why I now cannot mount it. How an I format it? Turn it into a NTFS or XFS disk and just run it as mounted disk outside array possible?

I am planning to upgrade my parity disk from 6TB to 10TB. Also, I'd like to replace one 4TB disk within the array with a 10TB disk. Would love to get some guidance how to do this most effectively. Assume I cannot change two disks at once? So, I will need to change the parity disk first? Should be easy. Just take old parity out when array is offline and add the new disk? It will then rebuild parity? Same for old data disk in array (4TB). I just take it out and put a new 10TB disk in? It will then rebuild the 4TB on the 10TB disk and I have 6TB free?

I have a disk that is formatted as btrfs. I used to use it as a cache drive, but may want to re-use it for something else. Two questions: 1) I tried to mount it as a regular disk outside the array. Somehow, this does not work. Assume this may be because I cannot mount btrfs disks? If so, what to do? Can I format it? Or other work-around? 2) My new cache disk is formatted as XFS. Can I have my "old" cache disk (formatted as btrfs) as second cache disk? I don't want to merge it with the first, but can I run both separately as two disks?

Hope to get this sorted out. Let me provide you some more information. Context where the HW : only one GPU (GTX 1050) used in primary PCI slot, GPU used by Unraid and not assigned to VM Below how "lspci -v" gives me related to the GPU. You will notice that the kernel driver is not in use (this was different when I did this first and unbinded it). https://pastebin.com/8XFap1JA Followed the comments to bind the card again. See error message below: root@Tower:~# cd /sys/bus/pci/devices/0000:65:00.0/ root@Tower:/sys/bus/pci/devices/0000:65:00.0# echo 1 > rom root@Tower:/sys/bus/pci/devices/0000:65:00.0# echo 0 > rom root@Tower:/sys/bus/pci/devices/0000:65:00.0# echo "0000:65:00.0" > /sys/bus/pci/drivers/vfio-pci/bind -bash: echo: write error: No such device And some more info from tools/system devices in case this helps trouble-shooting: IOMMU group 36 [10de:1c81] 65:00.0 VGA compatible controller: NVIDIA Corporation GP107 [GeForce GTX 1050] (rev a1) [10de:0fb9] 65:00.1 Audio device: NVIDIA Corporation GP107GL High Definition Audio Controller (rev a1) How can I "bind" the GPU again? What happened when I "successfully" unbinded my card?

Any thoughts on above? My GPU (primary slot GTX 1050) is no longer binded and I don't know how to bind it again. I had unbinded to dump the bios, but then failed to bind it again. Any thoughts how to do so? Thanks in advance!

I followed the instructions. Hope I didn't brick anything. I have an GTX1050, which is in my primary PCI port. I dumped the bios using commandline. I didn't move the card to a secondary slot, which I hope was ok? Everything actually worked and I succeeded to dump the bios. The only thing that didn't work is to bind the card again. I get an error message that this card doesn't exist. I initiall unbinded it. Everything seems to be still working though, but I am worried that I bricked something by not binding the card again?

What are the KVM net drivers and how to install? The issue is replicable, which is the weirdest part. I can set a static IP when using VNC (no GPU). Once I switch to GPU, it automatically change to DHCP.

Thanks! I wish this was sorted. The two may be related though? I am now working with a DHCP IP instead of a static one. Not ideal, but not the end of the world. It is a symptom though that something is wrong and the Kodi issue is also a network issue, so probably related. Root cause must be the GPU as all is working when I run it via VNC (disabling GPU). XML: https://pastebin.com/4VMGL7zG IOMMU: Enabled (in info). Anything else needed? Devices: What devices you referring to?

I am facing issues with Kodi since passing through GPU to VM. Kodi stopped working properly since installing my new GPU (GFX 1050 pass-through in VM). I am running Kodi 17.5 on a Win10x64 VM.When scanning the library, Kodi and the full computer basically becomes unusable. I can no longer move the cursor within Kodi, but I still hear the keyboard clicks. Only way is to force-stop Kodi. Error log from Kodi below. Help appreciated!https://pastebin.com/6dVCKs1T I doubt that this is a Kodi issue as it was working well and still works well once I disable the GPU and emulate via VNC. Any thoughts appreciated!

And yet again me with some more progress. I realized that it is not feasible to set up VNC as primary and Nvidia as secondary. So, I dropped this idea. I am now back to running the KVM with primary Nvidia. I even succeeded to get RDC to work. The issue was that I had set a static IP and somehow I cannot set a static IP once I have my GPU passed through. I tried everything including the DHCP IP as static IP, but nothing is working. Could it be that static IP assignment for some reason does not work when passing through a single GPU?

I am making progress. I added VNC as first and Nvidia as second GPU. Did the manual changes in the XML to add the 3rd party bios. RDC is working now, which is great. However, I can no longer output from the nvidia to an external monitor. Also, when accessing device manager (through RDC), the nvidia card shows with error 43 (Windows has stopped this device because it has reported problems). Appreciate your help!

I just gave it a shot using a different dumped bios. It appears to do the trick. Good and bad news! The good news is that the VM is working when having a screen attached to the only GPU. The weird thing is that I cannot access the VM via RDC. Any idea?