Stokkes

Ha! That is exactly what I'm doing now - streaming to my Apple TV 4K using Apollo/moonlight. It runs well, but even with the 5070 TI on this cpu, I kind of need to enable FrameGen which seems a bit ridiculous. I am not hyper sensitive to input latency. Enabling framegen I can keep the latency pretty stable at 8-16ms and I don't really notice perceptually if there is latency on the moonlight/apple tv front (it is wired however). My gut tells me I should probably tough it out until Nova lake comes at the end of the year which looks to be quite beefy and I've heard the core count for higher CPUs should be pretty crazy. I almost have to brute force it if I want to stick to VM gaming.

Ya makes sense, I have been flip flopping on sticking with an all in one or splitting and going with a 9800x3d.. This is what I'm telling myself too - the i9-10900 is absolutely fine for unraid. Thanks!

Wow thanks for the response. That's awesome - any issues you've noticed (I hear things about IOMMU, passthrough, etc)? I see in your OP that you have VMs, etc.. Do you do any gaming? If you run Plex, did you install another GPU for transcoding? I'm running an i9-10900 (non-K) and while it's been humming along since 2020, the gaming performance leaves a lot to be desired with my new 5070 TI. I know RAM pricing is nuts right now, but I am considering pulling the trigger because when I game, my GPU sits at 60-70% @ 1440p with max settings in many games, so it's being severely CPU bottlenecked.

Hey, sorry to resurrect an old thread, but did you end up ever upgrading to the 9950x3d? I am curious to hear from people who have this. It appears the sweet spot for high core count/threads and price vs. threadripper. Considering this or possibly waiting for Nova lake (panther lake has been getting pretty good reviews so hoping they can continue that with Nova lake). Thx

Howdy I'm seeing NVRM Out of Memory lines in my dmesg using a 5070ti + latest open source driver: [Fri Jan 23 21:21:59 2026] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc$ @ mem_desc.c:1359 [Fri Jan 23 21:22:00 2026] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1359 [Fri Jan 23 21:22:02 2026] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1359 The GPU is currently passed through to a Windows 11 VM, but it is not bound to VFIO (as I sometimes shut the VM down and use the GPU for transcoding using unmanic. Seems odd to me Unraid would complain about this with the GPU being passed through. If this is expected, that's fine, or if I need to attach a diag i can. The fix common problems plugin complains almost daily that I have oom errors, but the only oom in the entire log is this. Cheers,

@EDACerton I've been thinking a lot about this over the last month or two, the simplicity of having auto-unlock but with some modest security (understanding that if someone wants to really get in, they will) - doesn't mean I can't make it extremely difficult. The timing of this plugin was kind of odd, but good as provides additional perspective (and these conversations are golden). I may build something myself (either for Unraid or maybe bring my unraid server back under proxmox and use proxmox to do the decrypting) and thinking I'll adopt 3-4 different dimensions to derive the key.

This is a great plugin! Definitely something that should eventually be integrated into Unraid I think, great job. Agree with Niklas, would be good to have a configuration parameter to give it an IP to get the Mac address and use that. I've been doing something similar where I use a small device in my home that is hidden and go grab the key via SSH, and then delete the key once the array is started.

Hi there. Running 7.1.4 and have 5 "Connect" API keys and and 2 "CliInternal" API keys. I'm trying to better understand each of their use and if I can delete non-used ones. Any suggestion here? What happens if I go and delete all of them, will it break the API/Unraid? Will they just get recreated automatically by the system? I usually keep my API keys pretty tidy on any platform so this is bugging me a bit :) Thanks!

Well thanks to both of you for confirming this is clearly some weird bug that the Unraid team hasn't gotten around to fixing yet. Bugs the heck out of me, but at least I'm not alone

Hey there, This is more curiosity than anything (may be a bug, who knows). I have two servers, both have Unraid Connect 2025.03.18.2027 (I have even attempted to uninstall the plugin and reinstall / reconnect on multiple occasions). For some reason,, 1 of my servers isn't showing what I suspect is the version number (it's just blank), and the other server is showing an old 2024.05.15.1138 version number. Is this a known issue? Maybe just my OCD, but trying to figure this out and haven't been able to. See the attached screenshot. Thanks,

Alright what an ordeal (esp as I don't have a monitor, but managed to use a USB HDMI Capture card with my iPad and Orion app). Putting solution here in case someone comes across this. Issue is related to how Supermicro boards / IPMI works with an external GPU. Even though there are settings set, if you have any monitor (or in my case a dummy HDMI adapter) plugged in, it will switch to use that, which effectively means IPMI becomes useless. I had put the dummy adapter for my RTX 4060 in Windows to allow custom resolutions, I also wanted the ability to use VNC within Unraid in case the VM. Anyway, with the Dummy plugged in and booting, it would switch to the RTX as the default GPU and kill the IPMI. Solution: stop using the Dummy adapter and install the custom virtualdisplay driver: https://github.com/VirtualDisplay/Virtual-Display-Driver Now without the dummy plug, IPMI works and the Gaming VM works with the built-in VNC for accessing the VM. Still annoying it appears by the manual you can get Supermicro boards to not detect PCIE graphics adapters on boot, but on my X12SCA-F board this doesn't appear to do anything as with the dummy plug it still switches to this display for the BIOS, etc even with that option to skip enabled. Bottom line: if you have a supermicro board with IPMI and use a Gaming VM with a passthrough GPU, you basically can't have anything plugged in if you're using this headless. I am using this VM for Sunshine/Moonlight streaming to my Apple TVs, iPhone, etc. but because I have SUnshine switch to a custom resolution, enable HDR, etc depending on the device connected, I needed the dummy plug.. What a mess, but figured it out if anyone ever runs into this.

Just adding that I removed the nvidia driver just in case, still no joy. (safe or non safe).

No go unfortunately. Blacklisted the i915 as per the instructions and booted into Safe Mode, still black screen Attached a diag from safe mode. stewie-diagnostics-20241211-1050.zip

Hey thanks for the reply @JorgeB, appreciate it. yes the server does have IPMI, in fact that's how I was using it (sorry for leaving that detail out). Once I select the option from the Unraid boot (blue) menu, it loads bzroot and then goes blank on the IPMI. In the BIOS I set the iGPU as primary (instead of Auto) and nothing is plugged into either GPU. I don't have a monitor handy, would have to go borrow one (it's a headless server and my other computer is a Macbook Pro). Attaching diagnostic file. stewie-diagnostics-20241211-0833.zip

Hello, A bit at my wits end on this one. I have a Supermicro X12 series board with an Intel i9-10900 and onboard GPU. I also have an nVidia RTX 4060 that I pass through to a Gaming VM. My issue is that when the Server boots, I see the blue screen to choose which Unraid to boot (regular, GUI, etc.) and once it loads bzroot the screen goes black/blank. The server boots normally and I can access the Web GUI no problems. I'm actually unsure when this problem started until recently I had to troubleshoot locally and had no local console. I've tried a bunch of things (short of removing the nvidia card), even setting the Primary graphics adapter in the BIOS to the IGFX, ensuring the i915 driver wasn't blacklisted in /boot/config/modprobe.d, etc. Anyone have any tips?