ocean_breeze

It shouldn't be bound there at all. But you're right, it was. I've removed that and nvidia-smi as well as the plugin page are both working now. No clue how it got bound. Thank you so much!

Okay, so to update you on this: Booting into GUI mode produced no video output from the GPU, however I was able to monitor progress via iLO. After loading SMB, it get stuck on a flashing cursor. The webpage does still work via a browser. I logged in, checked the plugin, and saw that it was displaying a GPU ID and that nvidia-smi was working still. However, seconds following this, the web UI became completely non-functional and refused to load pages. I tried to perform a graceful reboot via iLO but it errored out (attached a screenshot of it). It hung here for several minutes before I had to forcibly reboot the system. Booting back into non-GUI mode works fine, but the plugin ceases to work. It did cause a parity check to be started, but no big deal. Two things to note: This is a fresh unRAID install (less than a week old) and switching to legacy 470.XX drivers resulted in the same issue. Super weird, right?

So I enabled legacy mode and PCI Express 64-bit BAR support was already enabled. If anyone else reads this and wants to know where to find it, you have to hit CTRL+A while in the RSBU to open a secret "Service Options" menu. Still had the same problem, unfortunately. I know this server/GPU configuration works though as I've previously used it on a baremetal Ubuntu Server install on the same machine.

Having an issue where the plugin (and console) reports "Failed to initialize NVML: Unknown Error" after the plugin is initialized. If I run "nvidia-smi" prior to opening the plugin page, I do get an output. After opening the plugin page, I only receive that error. I've attached an image of the console/plugin page as well as my diagnostics file. This is a ProLiant ML150 Gen9 with a GTX 1080 inside of it. Thanks! fileserver-diagnostics-20230423-1350.zip

Found the issue! It was caused by permanent ARP entry on my Opnsense router's ARP table that had an outdated MAC and seemed to be readded by the old DHCP server I used to use on there. Not a problem with unRAID at all!

Hey, this is my first post on the forums. I've dug around everywhere and not been able to find anyone else with my same issue. My unRAID machine has inexplicably stopped communicating with my gateway. It can ping everything else on my network, but not my gateway. This is the only machine on my network with this issue. I've currently only got one onboard NIC plugged in. I've checked firewall settings and not found any rules or routes that prohibit traffic. Weirdly, the VMs using the br0 interface have no problem getting out to the gateway. It may be important to note that this machine hosts an OpnSense VM (that has its own two NICs separate from the onboard NIC). I have attached my diagnostics file along with screenshots of my eth0 configuration and my routing. Am I missing something here? Is there any more information that I can/should provide? Thanks! fileserver-diagnostics-20230421-1924.zip