SmokeyColes

This emulates a nvme and it worked

<target dev='nvme0n1' bus='nvme' rotation_rate='1'/> delete: <address type='drive' controller='0' bus='0' target='0' unit='2'/>

Hello In a linux distribution a vdisk shows as /dev/disk/sda. How can I make it display as: /dev/nvme0n1 In the VM template I can see a drop down which has VirtIO / SCSI / SATA / IDE / USB Is it at all possible? Could I amend: <disk type='file' device='disk'> <driver name='qemu' type='raw' cache='writeback' discard='unmap'/> <source file='/mnt/virtualmachine_cache/domains2/SteamOs/vdisk1.img'/> <target dev='hdc' bus='sata' rotation_rate='1'/> <serial>vdisk1</serial> <boot order='1'/> <address type='drive' controller='0' bus='0' target='0' unit='2'/> </disk> Thanks Chris

Hello This maybe the first ever post I can help. (1) Download image (2) WinRAR extract img file (3) Use Rufus to flash USB - on the img is 5 partitions. (4) Plug into UNRAID (5) Mount - not my 5th partition would not mount (to check if this is a problem with the USB) Partitions 1 - esp vfat 2 - efi vfat 3 - rootfs - btfs 4 - var - ext4 5 - home - ext4 (unmountable?) (6) In terminal type: ls -lha /dev/disk/by-id/ (7) Find usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0 lrwxrwxrwx 1 root root 9 Dec 5 14:50 usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0 -> ../../sdx lrwxrwxrwx 1 root root 10 Dec 5 14:50 usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0-part1 -> ../../sdx1 lrwxrwxrwx 1 root root 10 Dec 5 14:50 usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0-part2 -> ../../sdx2 lrwxrwxrwx 1 root root 10 Dec 5 14:50 usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0-part3 -> ../../sdx3 lrwxrwxrwx 1 root root 10 Dec 5 14:50 usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0-part4 -> ../../sdx4 lrwxrwxrwx 1 root root 10 Dec 5 14:50 usb-Kingston_DataTraveler_3.0_60A44C426518BEA0DB429420-0:0-part5 -> ../../sdx5 (8) In SteamOS VM template; I used OVMF for bios (9) In SteamOS VM template; add a second vdisk, select manual and type: /dev/disk/by-id/***copy the device in step 7 and dont include these 6 stars*** (10) Ensure usb devices you have not selected the USB (11) Select virtual; note I found the RTX3060ti sluggish; VNC worked better for the install. (12) Run steamOS template (13) At shell type exit to go into boot manager (14) Select "UEFI Misc Device" (15) SteamOS installer boots (16) At GUI select Wipe Device & Install SteamOS icon At this point I had issues with "NVME" - if you have passed through a NVME you will unlikely have this issue, if you have a vdisk like me. (A) Force Stop VM (B) Into the template, select XML. (C) Under the vdisk (not your USB) - change <target dev='hdc' bus='sata' rotation_rate='1'/> to <target dev='nvme0n1' bus='nvme'/> (D) delete the <address type='pci'> under vdisk1. (E) Save and restart the SteamOS (17) Select "Wipe Device Device & Install SteamOS icon (18) With the format caution, click proceed. (19) Click proceed to reboot after "Reimaging Complete" (20) Enter username, password and steam guard. (21) Hit shutdown to shutdown VM, configure GPU in VM template. Remove rufus steam.img USB - I used a NVIDIA RTX3060ti (22) Restart VM (23) When booted open Konsole type "flatpak install flathub org.mozilla.firefox" (24) Once downloaded type "flatpak run org.mozilla.firefox" You have to set a admin password to unlock read-only, enter console and type: passwd if you have locked yourself out you will need to reboot, and do the next steps otherwise ignore. Goto "Terminal with repaire tools" and type: sudo ~/tools/repair_device.sh chroot rm -f /var/lib/overlays/etx/upper/passwd rm -f /var/liv/overlays/etc/upper/shadow then reboot removing USB. Goto Settings > User and set password (25) Type sudo steamos-readonly disable (26) Type sudo pacman-key --init (27) Type sudo pacman-key --populate archlinux (28) sudo pacman -S archlinux-keyring (29) For nvidia: sudo pacman -S nvidia nvidia-utils nvidia-settings (30) hit proceed with installation (31) sudo reboot After this point I just got a black screen, so went back to VNC. No matter what I did I could not progress. I gave up with the NVidia 3060ti, and used my Arc - it worked.

@JorgeB perfect That's what I will do.

@JorgeB - could you help explain; how can a single drive create a mirror. Can you create two ZFS pools independently; where one mirrors the other? Or is it a case of adding a drive in that pool with a setting for mirror?

Hi I have purchased a mechanical SATA HDD, in the UK we seem to be having a bit of a HDD shortage. So I will be get a second with a pre-order; though SCAN have told me Seagate have a backorder on the EXOS. I want to set up a RAID 1 style cache. However I am only going to have the one drive for now. I think with two drives - I'd like ZFS RAIDZ1. I intend to store vdisks on it and iso's - whilst keeping my system cache (NVME) for dockers and apps. So speed + 1disk failure redundancy is what I want is the goal. I realise this is non-conventional to know you want mirroring but have only one disk - but I want to start using it asap as a cache pool. Then take large vdisks off my array and into a cache pool. I know ButterFS is good for flexibility. My question is - (1) What initial file system should I use? (2) If I can use ZFS can i then install a second of the same size; and select RAIDZ1? Thanks Chris

In fact I don't want all my hard drives spinning; so I think the option here is simple! I need an individual drive cache pool with domains2 on. It just means buying a hdd, which I probably need to replace one other whilst at it. Correct me if I have misunderstood or you have other idea's. Otherwise it maybe that in the current state its doing exactly what its meant to do and I don't have a problem.

Sorry had to take a read (https://forums.unraid.net/topic/50397-turbo-write/) My Tunable (md_write_method): setting is "Auto"; should I select "reconstruct write"? After rechecking I noticed the tooltip says: auto selects read/modify/write. Maybe I need to use a separate single drive cache outside the array. I do like the idea of it being backed up.

Hi I am transferring about 600GB of 20GB zip files to a 4.5TB vdisk which I have running in a non-cache drive on a dual parity protected UNRAID array. The transfer of data is SMB via windows explorer using VIRTIO-Net ethernet. Transfer speed is about 270MB's; and then drops to 0bytes/s after roughly 10GB. It then sits at 0bytes/s for a minute or two, than kicks up again. This also happens when I unrar large files already in the vdisk - the progress bar just pauses; though the unzipping doesn't give me any transfer info - so I assume data is being written to vdisk temp folder (the vdisk stored in domains2 mechanical SATA drive in UNRAID) domains - holds the VM C drive - that's on a NVME cache domains2 - mechanical drive on array protected by parity for large vdisks such as 4.5TB Basically my question is that it feels very slow doing data movement. Is this actually a problem or is this OK given its an array mechanical drive vdisk2 is sat on. Windows does not lock up, I can still use the start menu, explorer is a bit of a problem (unresponsive on the affected drive) I've done some basic work - I can see posts about dirty_ratio and dirty_background_ratio. I don't really want to touch as even though I have read the guidance - I'm not entirely sure this will help. Equally the solution maybe something different. Does anyone know a little more, I'm hoping this will resonate with others and not confuse. VM info: RAM: 16GB min, 32GB max, 256GB total UNRAID 6 cores of the CPU (33% of cores) Purpose: activated win11 home running launchbox & emulators. Used ref: https://forums.unraid.net/topic/88484-transfers-stalling-two-most-common-remedies-not-working/

OK thank you for confirming.

Hi I am having an issue with my notifications menu. I think its only started after the upgrade to 7.2.2.- I'm probably about 75% sure it was OK before. That said I'm half expecting someone to correct me which is OK, I've included my theme setting just in case. I wonder if someone could confirm?

@ich777 apologies

Hi Is there anyway I can install an older driver 474.82 for my old GT730; rather than the latest? My use case : I don't really need the latest drivers for modern GPUs (I don't have the room or PCIe for them), and the GT730 fits into a PCI-E x4 slot. I've tried pretty much everything to get a VM to work with it so suspect its the new driver causing QEMU failing errors. Thanks Chris

Hello I have started using Disk14 rather heavily over the last few days - particularly one large vdisk as a second drive to a arcade VM. Could someone please tell me if I need to be concerned, if I have lost or corrupted data and if I need to replace the drive. The drive is old - I believe was manufactured in 20/07/2015. Also I would appreciate but only if willing, you could explain to me how you've come to your decision. Maybe it will help me in the future self-resolve. Thank you Chris ST6000VN0001-1SF17Z_Z4D2RR7S-20251128-1828.txt 187 Reported uncorrect 0x0032 094 094 000 Old age Always Never 6

Its possible this maybe too old a card and not supported by the NVIDIA driver. I may have to accept it a lost device. I am however going to try playing with the drivers and vbios dump - see if it makes any difference.

Hello I am receiving the error: I have tried the following (cluelessly self troubleshooting): Trying PCIe Override to Multifunction and now Downstream I have bound the graphics and sound. I'm not sure now how to further troubleshoot. When I run the linux build - I have issues with the graphics in VNC using QXL. I can only seem to get it working with Virtio(3d):Auto So I am confident the image is not the problem. It's probably a very simple fix, but I am at a loss and the log is somewhat cryptic to me.

Just wanted to feedback for any who may read this - the Arc A380 Sparkle Genie card arrived from Spain and was installed. The above instructions followed and I have had multiple transcodes occur from the two different servers (two docker containers) at the same time - and its worked perfectly. It detected it in PLEX with no issue, and was selected in the transcode settings (image below). Its great and easy to setup.

Its someone else's screenshot: And doing that for both should work from what I'm reading. From Beta 7 - Arc became supported. @Niklas; thank you will mark as solution

In answer to the first bit of the question. I have misunderstood. SR-IOV refers to i915 plugin. If it says GVT-g and not SR-IOV then its the other plugin. Seems simple when someone else pointed it out This does generate a new question; if A380 is part of the A-Series (formerly known an alchemist) - it states it is not supported by either plugin. So how have people got it working with UNRAID? 2nd part of the question, would still be appreciated - can it be shared with 2 docker containers and how?

Hello I have bought the A380 Intel GPU. Its due to arrive this Wednesday from Spain. Spec below: https://www.sparkle.com.tw/files/20240729114753160.pdf I want to assign it two plex docker containers. Could someone tell me the required plugin: (Why I am confused as reading the 915 beta description sends you to a site which confirms if GVT-g (the left plugin) works) I also have the ability to use an iGPU, so I have one plugin already installed. Setup Containers: Is it just a case of adding --device /dev/drv to the extra parameters of both containers? Then in both PLEX severs selecting the A380 under the hardware transcoding device dropdown (located in settings > transcoder)? Forgive my limit understanding, is it capable to run transcoding at the same time without I/O conflicts? Thank you Chris

It is also noted in spaceinvaders GlutunVPN setup (which I have had to go to as enabling tailscale stopped the VPN from working; I originally skipped this part of the video as at that time I found unclear what it was) but outbound firwall and a few other things need to be enabled.

Embarrassingly it was late last night, and a google search found: This has the solution. The host path 1 is "config" is a binhex plex; therefore the state directory is /config/.tailscale_state. Really nice step by step guide from @ich777

I think this may answer another question I have just posted at this exact moment, however its 1am here so I won't really take it in well. I am going to give this a good read though tomorrow 👍

Hello I have a issue, I'm undergoing a bit setup and keep hitting hurdles. I use star link as I don't get good broadband, though we are in the process of the village having fttp setup. Star link is a CGNAT which causes all sorts of problems with remote access. My container setup: GlutunVPN binhex-qbittorrent binhex-radarr binhex-sonarr firefox binhex-overseer binhex-prowlarr Plex (br0 + tailscale) I have tried setting a custom server URL in plex but of course sonarr and plex don't speak. Data travels from sonnarr to the VPN and to the web. So I realise I need to setup tailscale on Glutun. However I get this issue in the container log: I'm not sure if this is a docker, glutun or tailscale issue. I am also not sure what the persistent Docker directory is? I do see under tailscale advanced options: Am I doing something wrong? Can anyone help. I have had to amend port 3001 to 4001 as firefox uses 3001. Can any one help? Thank you Ref how I plan to make the two containers speak: https://forums.plex.tv/t/remote-access-using-tailscale-magicdns-and-https/799941