raysv

Main unraid version or also the smaller one *.*.2? Useing a RTX5050, mainly used for encoding. But nvidia has some trouble with atleast current driver, so when useing tune uhq, it segfaults on some sources On version 7.2 :)

General question, how long does it usually take from when a opensource driver gets released to that its available in this plugin?

Hi, I just installed a RTX 5050 in the server, and installed the plugin. However the plugin when opening (have ofcourse rebooted x number of times), says it dosnt find a GPU. However, in system devices i find this: IOMMU group 12: [10de:2d83] 01:00.0 VGA compatible controller: NVIDIA Corporation Device 2d83 (rev a1) [10de:22ec] 01:00.1 Audio device: NVIDIA Corporation Device 22ec (rev a1) I can also choose that NVIDIA card and set it to be used on a VM, and it manages to output a picture. When the VM is running, im getting this inside the plugin: Installed GPU(s): Failed to initialize NVML: No supported GPUs were found Unable to determine the number of GPUs Any idea of what happening here or how to fix it?

Yeah, probably the plextractsync docker, so not related to this issue

Aha , don't know why i have containers restarting at the middle of the night for that sake either, but thats a problem for another day. Anyway, have run the diagnostic tool, and supplying the logs here. enterprise-diagnostics-20250202-1301.zip

Hi, After upgrading to Unraid 7 (did not notice this problem on the beta that i was running before upgrading), i have a problem with my network seem to reset every night at alittle bit after 04 each night. This results in VM's loosing connection to their shares, same with dockers and that that server becomes unreachable for a short while. I have tasks running on the vm's and dockers 24/7, and the result is that all of this tasks errors out. The only thing i find in the logs are: Feb 2 04:40:02 Enterprise kernel: docker0: port 8(vethc0899d8) entered blocking state Feb 2 04:40:02 Enterprise kernel: docker0: port 8(vethc0899d8) entered disabled state Feb 2 04:40:02 Enterprise kernel: vethc0899d8: entered allmulticast mode Feb 2 04:40:02 Enterprise kernel: vethc0899d8: entered promiscuous mode Feb 2 04:40:05 Enterprise kernel: eth0: renamed from veth5008833 Feb 2 04:40:05 Enterprise kernel: docker0: port 8(vethc0899d8) entered blocking state Feb 2 04:40:05 Enterprise kernel: docker0: port 8(vethc0899d8) entered forwarding state Feb 2 04:42:30 Enterprise kernel: docker0: port 8(vethc0899d8) entered disabled state Feb 2 04:42:30 Enterprise kernel: veth5008833: renamed from eth0 Feb 2 04:42:30 Enterprise kernel: docker0: port 8(vethc0899d8) entered disabled state Feb 2 04:42:30 Enterprise kernel: vethc0899d8 (unregistering): left allmulticast mode Feb 2 04:42:30 Enterprise kernel: vethc0899d8 (unregistering): left promiscuous mode Feb 2 04:42:30 Enterprise kernel: docker0: port 8(vethc0899d8) entered disabled state I guess these messages is related to what is happening. Any idea of what is going on here, and eventually how to fix it?

Yeah loooked at the video, adn used his script to clean up! But there wasnt realy anything to clean up or volumes that wasn't mapped correctly. And the values (virtual size) that the script mappes up is much smaller than what unraid is reporting the docker img size to be. Still it keeps growing with 5g during a 24h and gone from 12gb to 33 gb in a couple of days! Weird thing, can't remember doing any changes to get this to start either.

Hi, I have some issues with my docker.img is filling up, up to 35gb now, and still ends up at 100% after a couple of days. How do i check what docker that is actually filling up?

Hi, Is there a way to force ONLY a Intel Arc to be passed through to a docker container and NOT the igpu? Got a couple of dockers that seems to use the igpu instead of the Arc

Hi, Was just wondering if there is anybody that have managed to get av1 encoding work in tdarr docker? I have the custom kernel installed, works fine in the rest of my dockers, but when it comes to encoding av1 in tdarr it failes badly! Just getting these errors: 2023-11-01T22:59:43.654Z [av1_qsv @ 0x56477ed80500] Selected ratecontrol mode is unsupported 72 2023-11-01T22:59:43.654Z [av1_qsv @ 0x56477ed80500] Current frame rate is unsupported 73 2023-11-01T22:59:43.654Z [av1_qsv @ 0x56477ed80500] Current picture structure is unsupported 74 2023-11-01T22:59:43.654Z [av1_qsv @ 0x56477ed80500] Current resolution is unsupported 75 2023-11-01T22:59:43.654Z [av1_qsv @ 0x56477ed80500] Current pixel format is unsupported 76 2023-11-01T22:59:43.654Z [av1_qsv @ 0x56477ed80500] some encoding parameters are not supported by the QSV runtime. Please double check the input parameters. 77 2023-11-01T22:59:43.654Z [vost#0:0/av1_qsv @ 0x56477ed7db40] Error initializing output stream: Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height 78 2023-11-01T22:59:43.654Z Conversion failed! 79 2023-11-01T22:59:43.654Z I do see that both vaapi and intel driver is coming up older than the VM i've gotten it to run on, so might be that ofcourse. So my question realy is, do anybody know if its something special you need to do with the container to get av1 encoding to work in an unraid docker? besides from useing thor's kernel.

Nice one, thanks! Ive added it in and rebootet crossing my fingers!

Damn, i spooke to soon, same issue with Samsung 990 PRO aswell, so i guess its not the SSD's fault! it just took longer before it happend. Unfortunally these diags are from after the rebooot, but atleast its a lot of errors surrounding the cache disk there aswell, and will remember doing it if neede before rebooting the next time enterprise-diagnostics-20231022-1302.zip

Hi, Not 100% unraid related, but giving it a shoot: Ended up with alittle snag, i've installed a ubuntu 23.10 vm with an arc 380 passthrough. However, it seems it finds the cards, outputs to the screen, and it sees the A380, however, when checking ffmpeg hwaccels, only vdpau is there, no quicksync or anything. From what i understand from intels side everything should have been backed into the kernel from ubuntu 23.04 and forward. What am i missing here?

Ended up returning the second disk aswell before i was able to check diags. Ended up with getting a samsung 990 pro instead, and that seems to work much better, so guess unraid just didnt like the SN 770's

Hi, I've couple of days ago i added another cache pool with a brand new WD BLACK SN770 2TB disk in it, but struggling to keep it up and running. After a while, it can be 5 mins, or a ocuple of hours, the device gets thrown out of the cachepool with the message "Invalid Path" ... looking at the log of the disk IO error. I then concluded with a faulty disk, so went back to the store exchanged it for another one, came home, swapped in the new disk again, and 5 minuttes later. Invalid Path and IO error. I have tried all the M2 slots now, so its not that, I have no issues with the other SSDS in the server. And the disk always "works" again for a limited period of time after a reboot. Its a XFS disk btw. Anyone with the smallest of ideas of what is happening here ? Or is it just that unraid dosnt like WD BLACK SN 770 disks?