Leaderboard

Hi Guys. So I have found that recently i have had to use a vbios with both my GPUs in my server whether Primary or Secondary GPUs. So i thought i would try and write a script that would automatically dump the vbios from a GPU in the server even if its a Primary GPU without having to do any command line work or without having to download a vbios from techpowerup and then hex edit it. I have tested this in about a half a dozen servers and it has worked for all of them. Before running the script I recommend shuting down the server completely (not rebooting) the start it back up. Make sure no containers are running (disable docker service) and make sure no VMs are running. This is because if the GPU is Primary the script will need to temporarily put the server to sleep for a few seconds to allow the GPU to be in a state where it can have the vbios dumped from it. If alot of containers are running the server doesnt always goto sleep and the script fails. How the script works. It will list the GPUs in the server showing the ids. Then with your choosen GPU id the script will dump the vbios using these steps. It will make a temporary seabios vm with the card attached, then quickly start and stop the vm with GPU passed through. This will put the GPU in the correct state to dump the vBios. It will then delete the temporary vm as no longer needed. It will try to dump the vbios of the card then check the size of the vbios. If the vbios looks correct (which normally if gpu is NOT primary it will be) it will finish the process and put the vbios in the location specified in the script (default /mnt/user/isos/vbios) However if the vbios looks incorrect and the vbios is under 70kb then it was probably dumped from a primary GPU and will need an additional stage. This is because the vbios was shadowed during the boot process of the server and so the resulting vbios is a small file which is not correct. So the script will now disconnect the GPU then put the server to sleep. Next it will prompt you to press the power button to resume the server from its sleep state. Once server is woken the script will rescan the pci bus reconnecting the GPU. This now allows the primary gpu to be able to have the vbios dumped correctly. Script will then redump the vbios again putting the vbios in the loaction specified in the script (defualt /mnt/user/isos/vbios) Here is a video showing how to use the script Of course you will need squids excellent user scripts plugin to use this script. You wil find the dump script on my github here https://github.com/SpaceinvaderOne/Dump_GPU_vBIOS I was hoping that if this script works for most people then we could get a big repository of dumped vbioses (not hexedited) and I will put them in the github. I have made a start with a few of my GPUs. So if you want to contribute then please post a link too your vbios and i will add it. Please name the vbios like this. Tith type, vendor and model and memory size. For example RTX 2080ti - Palit GAMING PRO - 11G The vbios collection is here https://github.com/SpaceinvaderOne/Dump_GPU_vBIOS/tree/master/vBIOSes Well I hope the script works for you and you find it useful. Oh and as its the now 12 midnight here in the UK - its Christmas day. So merry christmas fellow Unraiders I hope you have a great day and a good new year

Same. I now have catalina (instead of Big Sur which is what I chose) running though and I see the apple update thing for Big Sur. Can I update from with Mac OS Catalna to Big Sur or is that gonna cause problems? **EDIT** I completely deleted then reinstalled the container with method 2 instead of just changing after the vm was installed and it worked.

So, I've created a couple of custom themes for the login page that I think people might enjoy. The themes and installation instructions can be found here: https://docs.theme-park.dev/themes/addons/unraid/login-page/retro-terminal/ The themes are all highly customizable, so If you got some cool ideas I'd love to see them! @jonp The wait is over A couple of screenshots

personal experience, there is nothing really good compared to rdp windows when using mac remote, but thats my personal experience ... vnc, speed well ... but scaling, audio, ... teamviewer, ok .... anydesk, ok .... splashtop, never tested, give it a try, something similar to rdp would be really nice

I use splashtop personal as Spaceinvader has recommended it. I'm able to connect from another VM, laptop or phone

Ich antworte mir mal selbst. Habe des Rätsels Lösung gefunden. Ich hatte das VPN innerhalb des Opera-Browsers genutzt. Zeitgleich kam ich nicht mehr auf die Lokale IP des Servers. VPN deaktiviert, und es ging wieder.

It really comes down to two things: Quality Do you want the best quality and efficiency for transcoding at a reasonable price? If so, you’ll want the GTX 1650 Super or above. It has the Turing encoder and decoder and it works very well. It does require a PCIE power cable though but it doesn’t use much power for transcoding. If you can handle a small step down in quality, a P400 or GTX 1050 or 1050 TI will work wonderfully. They are both 75W cards and the P400 only sips power. Number of Transcodes All Quadro cards under the P2000 and all GTX/RTX (except for the RTX 4000) cards are locked at 3 transcodes. The P2000 can handle about 20 1080p right out of the box. Going above the P2000 is really unnecessary for practically all servers unless you have a ridiculous number of people you share your server with.

I believe that disk include/exclude are only applied when writing new files to a User Share and are ignored when reading files which automatically includes all drives.

Hi, this is the expected behavior, sound through hdmi/dp is expected to not be controlled via software but via hardware, mechanical buttons on tvs or monitors or such. However you can use third party applications, for example sound control, to be able to adjust the volume via software. Merry christmass!

Have you clicked Refresh in tge Server Browser after you addet it? From the system usage perspective it looks like everything is running, I will take another look ASAP. Will take a look at it ASAP.

There are 2 ways you can approach this, don't know if either will satisfy. Set to cache-only if you want new writes to go to cache but nothing gets moved. No need to exclude any disk in this scenario since only cache gets written. Or set to cache-no if you want new writes to go to the array but nothing gets moved. Include/Exclude if you want.

As far as I know, Mover when set to Prefer has always moved from the array to the cache drive / pool. It doesn't think about included / excluded disks when moving that direction.

Merry Christmas and a healthy new year Ed. Cheers mate!!! And thanks for the excellent video as usual, couldn't do without you.

Thank you! I had to wait for the disks to preclear before starting but it worked great. New config allowed me to restart the server with disks 1-6 data in tact and usable. After parity-sync completed I tried to mount the disks in UD. Disk 7 mounted and I was able to recover all 1.28 TB of data off of it. Disk 8 wouldn't mount and Disk 9 is completely dead. I ended up giving up on the 169 GB of data from Disk 8 but it's no biggie. Nothing critical. Thank you again and Merry Christmas, Happy belated Hannukah, Happy Kwanza, Happy New Year, Happy almost end to 2020, etc 😃

Have you seen this?

I did some more work to get it to look like the CPU sidebar that comes with unraid

Another check you can make. Go to Shares - User Shares and click Compute All button at bottom, will take a while to produce results and you may need to refresh page. That will show you how much of each disk each User Share is using, so you can see if appdata, domains, system is all on cache. The triangles in your previous screenshot just means some is not protected but doesn't really indicate it is all on cache.

Sounds like you've got it figured out. You are correct on all

The stick is USB 3.1 but I assume backwards compatible. USB 3.0 blue port, the MB manual(2013) just says USB 3.0. Unraid labels it as "Flash SanDisk_3.2Gen1", the "2" is USB 3.0 port 2 of 2.

Here is a guide from Valve itself: https://developer.valvesoftware.com/wiki/Rust_Dedicated_Server Just append the command line parameters to the Game Parameters and you should be good to go. Keep in mind that this is not that different to a Linux Dedicated Server.

Developer / Maintainer Profiles CA supports developer profiles. This allows a user to look at and compare the developers who are creating / supporting the various containers / plugins. In order to take advantage of this, create a file within your template repositories called ca_profile.xml For most developers, the profile would be something akin to: <?xml version="1.0" encoding="utf-8"?> <Maintainer> <Icon>https://raw.githubusercontent.com/ich777/docker-templates/master/ich777/images/avatar.jpg</Icon> <Profile> Creating Docker containers with a little different approach than most developers, the container only has the dependencies installed and also the start script and downloads the application itself on the first start and updates the application on every start/restart of the container. </Profile> <DonateLink>https://www.paypal.me/chips777</DonateLink> <DonateText>If you like my work please consider Donating.</DonateText> </Maintainer> However there are also other entries which can be utilized if so desired: <Forum> - A link to a forum page here (announcement thread, general support thread etc) <WebPage> - A link to a webpage you operate <Facebook> - A link to a facebook group / page <Reddit> - A link to a reddit group <Discord> - A link to a discord site <Twitter> - A link to a twitter page Whenever a user happens to look at the repositories that are present in any given search result, the ones which have the profile filled out are displayed first in the list.

Assuming you are not getting into the mess of running a cluster just yet and the folder paths are stock, you can set mods up for auto update pretty easily... Install a copy of SteamCMD that the ARK server can control From tower/docker, Go to the containers console and issue these commands (minus the quotes of course) 'cd /serverdata/serverfiles' './SteamCMDInstall.sh' 'ln -s /serverdata/serverfiles/Engine/Binaries/ThirdParty/SteamCMD/Linux/steamapps /serverdata/Steam/steamapps' then you need to add the mods to the Game.ini ether by cating them in from the console or how ever you prefer to edit your config files. They do not need to be in the same order as they are in the GUS.ini (or on the command line depending on how you load them) but I tend to keep them that way for sanity. 'cd /serverdata/serverfiles/ShooterGame/Saved/Config/LinuxServer/' 'cat << EOF >> Game.ini [ModInstaller] ModIDS=2229978458 ModIDS=1083349027 ModIDS=1211297684 EOF' You should add -automanagedmods to the command line but I have found that the existence of [ModInstaller] in the Game.ini is enough to trigger the server to fire the subroutine that fetches the mods from the workshop, unpacks them, and writes the .mod file. be aware that if you add the -automanagedmods or [ModInstaller] and have not set up the SteamCMD correctly the server will fail to start with a segfault because it is calling an external application that isn't available.

Nope, i forked the repository and updated the version. (To use my version) edit your docker and change your repository to flubster/gsdock and it'll pull my updated version You'll need to add the mapping i stated above as i haven't created any UnRAID template, just pushed another docker image. YMMV. Be careful as I started from scratch (after pulling the initial image from the template) so haven't tried any upgrades etc Rubberbandboy

Getting docker and libvirt images onto cache and off array may help but should be done in any case. In fact, you don't even need libvirt since you aren't doing VMs. Go to Settings - VM Manager and disable VMs then delete the libvirt image from that same page. Go to Settings - Docker, disable dockers and delete docker image from that same page. Go to Main - Array Operation and Move Now. Wait for Mover to complete and post new diagnostics.

Changing the model to e1000 did the trick. Thank you for your assistance!

Thanks for the link, I came across that soon after your first post. For now I have a 1.5tb vdisk on a 2tb btrfs striped cache pool and so far it seems to be working great. I am transferring my steam library now so we will see what happens once it exceeds 1tb (Although I am certain it will be fine now that the data is striped). Once I have everything stable I will dismantle my old LXD server and use its SSD for the VM's OS, keep the cache pool for games, and use the rest of my disk for less sensitive data storage (mostly infrequently accessed Plex media and temporary storage). Not exactly what I intended but it accomplishes the same thing. BTW, I did notice that each share can independently enable/disable the cache, which is quite useful for me. Thanks for your help, I think I'll stick with unraid if everything looks good over the next few weeks. I am using the trial now but it looks like a license may be in my future (never thought I would pay for a linux distro, but lime-tech seems deserving enough) Just came across your post and thought of your problem. You don't want multi vdisks correct. I guess as you want one large drive in windows. Well as you are not using parity then the writes to your array will not be slowed by a parity write. So here is what you should do. Create a vdisk on each drive you have in your array so eg on disk 1,2 and 3. Attach all the vdisks to your windows VM. Go to disk management in windows and using those attached vdisks create a striped or spanned volume. This way to windows you will have one large disk but it is going across multiple vdisks. I just tried it on mine and seems to work ok edit .....Any reason you are not just using a mapped drive in windows to the array? You could just install the os on your ssd then map a network drive to your array for data storage. Shares on the array span across drives. A vdisk will not as it is just one file

i second the notion to have a stop mover functionality in the gui. a friend of mine has setup his first unRAID after seeing it at my place. now (cause of the wording "prefer") he changed a share and the mover started to copy all contents onto the cache drive – which is a SSD – and chokes becuase of too less space. now i tried alle commands which were recommended here in this thread – all DON'T work. they don't stop the mover! i would bet this situation will pop up more often and because of the misleading wording chosen for it's different kind of use cases. if nobody has a perfect one/two word naming, then (and why not), give the thing a small description in form of "Prefer (move/keep all share contents)" now, back to the question: how to stop a running mover script?