meep

The easiest way is to install the VFIO-Config plugin from Community Applications, mentioned by @T0a, above This provides a visual listing of devices, allows you click checkboxes to add them an generates the .cfg file for you.

I think I've been fortunate with AMD and I don't see the reset bug. Likewise, I think you'll find people will have smooth sailing with nVidia. My trouble for me was the -43 error. It seemed that no matter what I tried, I could not get nVidia drivers to install and in my most recent bout of attempts ended up installing Win10 about 10 times, trying different combos of machine type, bios etc. (nVidia drivers do not like to install if there's any hint that the OS is a VM) I finally found the magic formula, but it was hard fought. YMMV, as they say. Batteries not included. Investments may fall as well as rise.

You are welcome. Good luck!

Hi You are running CA Backup/Restore at 18:00. This is stopping your doclkters to allow the appdata backup take place. It tells you this in the logs; May 11 18:00:02 Tower CA Backup/Restore: ####################################### May 11 18:00:02 Tower CA Backup/Restore: Community Applications appData Backup May 11 18:00:02 Tower CA Backup/Restore: Applications will be unavailable during May 11 18:00:02 Tower CA Backup/Restore: this process. They will automatically May 11 18:00:02 Tower CA Backup/Restore: be restarted upon completion. May 11 18:00:02 Tower CA Backup/Restore: ####################################### It seems to be scheduled for this time every day. Are the dockers coming back up for you when this is completed? I'd suggest scheduling CA Backup for a more convenient time - 3.00AM?

So, the basic principle is that if one device in an IOMMU group is stubbed, that all devices in that group will be impacted. Therefore, if the device you want to passthrough is in a group with other devices, all must be stubbed. It sounds like you've got an IOMMU group from which you've stubbed a device that also includes your system network card. Therefore, unRAID OS has no access to this card and therefore no network. (look at the last couple of lines on your boot screen to see the IP addresses unRAID is assigned. It's likely you'll see a self-assigned address there rather than one recognisable from your LAN address range. Take great care with this - its very easy to stub something thats in the same group as your USB controller, and then unRAID won't boot at all as it loses access to the USB thumb-drive! So your first step is to try to get your passthrough device in its own IOMMU group. There's no point trying to stub it until you get this far. You're looking for a neat arrangement something like this; You need to play with ACS settings, Bios settings etc. until you achieve this. Only then can you start stubbing. Of course, it's entirely possible that your MB does not support this and then you're stuck. If you do manage to get your device in its own group, please post a fresh listing of your IOMMU groupings and we can take it from there.

Have you stubbed the network card, either in your syslinux file, or using vfio-config plugin? (Second option is easiest) for a device to show up for passthrough at vm creation, you. Must tell unraid at boot to ignore it and make it available. That’s assuming you’ve got it to show up in its own iommu group, of course.

Can you try installing something like Splashtop or teamviewer in your vm. Then, when you boot the vm connected to the big tv with no picture, you can remote in and see what’s going on. You might need to change resolutions, refresh rates etc, and you can do it that way. this is assuming your vm is booting to desktop, of course. At the very least, You’ll need to run netplwiz on the windows vm to bo bypass the windows login and ensue that the hidden vm is getting to desktop where you can remote in.

You’d need to check the current state of thunderbolt support in unRAID. As of recently, it’s a feature request; I dont one use it so cannot speak authoratively. i have toyed with external enclosures via SAS / SFF-8088 and found it to destabilize my unRAID to the point of unusability, though I never pinned down the exact cause. YMMV.

Maybe not of direct help, but as an ex-Plex user, I moved to an Emby/Kodi set up years ago and find it much, much better. If you're into HiFi, you'll understand when I liken Plex to an integrated system (all in one source/pre./amp), while an Emby/Kodi system is more of a separates set up with much greater flexibility. I got really fed up with Plex costs as well as restrictive functionality. Though I really liked Plex unification and the 'start watching anywhere, finish somewhere else' approach. Here's some of the reasons I switched, and stayed there; Emby server is an excellent standalone media manager / organiser Using Emby server is free, no need for paid functionality The free fork 'Jellyfin' is available if required The Emby plugin for Kodi is excellent, exposing all your media in Kodi categories Kodi has tons of customisation options so it can look however you want Kodi has excellent playback configuration options (video sync, audio bit-streaming etc.) I can decouple my PVR from my media organiser so not stuck with the limited tuner devices supported by Plex/Emby Kodi runs on almost anything I can build out any other functionality I want in Kodi - it's a great piece of software with an undeserved bad rep. for dodgy streaming I'm not locked in to a proprietary system I'm sure, for some people, Plex is absolutely the right solution. I'm a BetaMax / Syquest EZ kind of guy - I always end up with the best if not most popular technology. IMO, Emby/Kodi is the superior solution.

Using hardware know working for someone else will be a big win for you. Check out the build threads and peoples sigs and reach out to people with similar configs to what you propose. Based on my experience, I'd stay well away from nVidia GPUs. I know people make them work, some without issue, but I've set up lots of VMs on AMD RX5n0 cards without issue, but getting anything from nVidia running has been a royal PITA. I run 2x daily driver workstations on my system. Albeit both MacOs - one for me, one for the family. We alos have Windows VMs we can switch into for gaming, mostly the games you mention, as it happens. As well as those 2x systems, my unRAID runs pretty much everything else in VMs or dockers; NAS (obviously!), whole house audio, whole house PVR, home security system, home automation and lots more. The only real issue I have is that it's really difficult to schedule any maintenance or upgrades as there's always someone wanted to use something the system is doing! Sounds like you have this sorted, but switching to unRAID as a daily driver, I would still strongly advise that you retain some kind of physical PC (laptop) for troubleshooting. Good luck!

Splashtop is free for up to 5 target systems on local network. It works well. Ah, sorry, just re-read OP and noticed that its not local LAN. Splashtop free wont work in that case.

Or you can add something like an Intel RES2SV240 SAS Expander. This connects to your HBA and provides lots more ports. Despite appearances, it does not need a PCIe slot. It can use a slot for power, but theres a molex connector as well, so it cam be mounted anywhere inside your case. You can pick these up used on eBay quite cheaply, even cheaper if you want to source from China.

Since nobody has yet outlined the core tenet around here, I’ll add for completeness; parity is not backup parity is excellent protection against drive failure, and (usually) makes data recovery or drive swaps a breeze. However, consider that in the case of a failed drive, your data is unprotected during a rebuild, just at the point where your system is under most stress with all drives in simultaneous use. What if another drive fails at that point? You will likely lose data on that drive. you should always have a separate backup of critical data stored on your array. Sure, you can re-rip your media, but what about the family photos? It happened me just a few few weeks ago. My parity failed, and just as parity rebuild was starting to a new drive, another disk went down. I was super lucky that, while the disk wouldn’t add to the array, I could mount it outside the array and recover all the data. I managed to copy it to other drives, then back to an array disk before starting the whole thing up again and allowing parity calculate. not a pleasant experience, but a testament to the flexibility of unRAID that I could get that data of an otherwise failed disk (it was an unfixable XFS input/output error on the second failed disk).

Your proposed scheme is fine, and using the 10TB as parity will allow you add drives in the future up to 10TB each.

Also, if you haven’t seen it already, check out The Black Bear build thread. @testdasi has documented his designaire build, and is super helpful on the forums, so I’m sure he can answer any specifics you have on that board.

I have the same processor. When I was building, I’d researched extensively and chosen the Designaire, though they are very close in features etc. however, in the end, I found a lightly used cpu/mb combo for sale locally that was just too much of a bargain to pass up., so I ended up with a taichi x399. im happy to answer any questions you might have about the board, I’ve gotten to know it very well! You might like to check out my blog (see sig.), a lot of the recent articles include build notes around this system. The taichi is brilliant at iommu groups, I have A LOT of expansion devices , and it breaks them out no bother without ACS override. One advantage that the designaire has is the ability to choose your boot gpu. The taichi insists on booting to the first gpu it finds, and you can’t change that. ive also have an issue recently with memory. I’ve had 64gb in 16gb modules since build and it’s been rock solid. I recently added 4x 8GB sticks of otherwise identical RAM and things went to pot with constant crashing. I haven’t been able to fully test if the issue Is bad ram, a bad slot or just general incompatibility but those new sticks are now out of the system and it’s stable again. apart from that though, I’m super happy with the taichi. If I was doing it again, and pricing was equal, I’d likely go for the gigabyte, just for the gpu selection capability. You can’t really go wrong with either though. Just spec as much (known compatible!) Ram as you can if you intend using VMs - and why wouldn’t you with all those cores! And get heat sinks for any nvme drives you might be using - those puppies run hot. heres a peek at my iommu grouping, I’d you’re interested. https://drive.google.com/file/d/1r2oTe5mm9T0Kc7HIDAwGGIpXHs1drHza/view?usp=sharing GOOD LUCK

Hi Programmers - First post, be gentle! I'm dabbling with a new plugin related to VMs, and learning as I go. Looking at the Dynamix-VM-Manager plugin, there's a bunch of functions buried in helper includes that are perfect for what I need to do (generating lists of GPU, USB and PCI devices). Just wondering if it's considered good practice or bad form to call such functions directly from within my own plugin. Otherwise, I'd need to spend a lot of time writing exactly the same thing myself. Advice appreciated.

This works for me as well (ASRock Taichi X399, multiple AMD X5n0 cards). In a previous configuration, I had up to 4 discrete GPUs in my system all passed through to VMs. As noted, passing through the unRAID boot GPU can be finicky, but is entirely possible with the right hardware combination.

Mine got blacklisted recently after a 6.8.2 -> 6.8.3 update i pulled it from my unraid server, popped it in a windows PC and allowed windows repair the errors it found. This fixed the blacklist issue sometimes this happens, and the windows repair process resolves it. Worth a shot if this has happened out of the blue.

Result! Well done on the perseverance.

Yes, it seems XP needs a floppy at that point, or a USB spoofed as a floppy. Did you try to et primary vDisk bus to 'IDE' in VM config?

Was the article above no help? set primary vDisk bus to IDE then, and see how that goes....

More good news. I found this for you; https://www.raymond.cc/blog/install-xp-setup-did-not-find-any-hard-disk-drives-installed-in-your-computer/ It looks lies WinXP doesn't support SATA for install, and VIRTIO is causing a problem for you. Also, it seems that the drivers need to be on floppy (!) or USB, so likely why pressing 'S' results in no dice. Try setting your primary vDisk bus to IDE specifically. Or see if you can follow the instructions oin the above article. Good luck!

Actually, here's the correct guide for drive install during the setup process; https://wiki.unraid.net/index.php/UnRAID_Manual_6#Loading_the_VirtIO_Drivers_During_Installation

When you have the opportunity to load drivers, take it! You need to follow the driver installation sequence described here; http://lime-technology.com/wiki/index.php/UnRAID_6/VM_Management#Step_5:_Install_the_VirtIO_drivers_from_inside_the_VM_.28Windows_Guests_Only.29 That's actually for post-OS-install updates, but you my be getting that BSD because Windows Setup doesn't have drivers for some of your VM devices. You need to load the relevant drivers from the ISO as part of the setup process. (you need to do this for Win10, why not XP?) That page specifically indicates that the viostor drivers should be installed last. You'll need to load Balloon, NetKVM and vioserial drivers first, then viostor. You do them one at a time, so its a bit repeditive.