[Plugin] Linuxserver.io - Unraid Nvidia


2468 posts in this topic Last Reply

Recommended Posts

1 hour ago, Clayton said:

Hello,

I've noticed since adding a P2000 GPU to my UnRaid server (using it for Plex and handbrake) that I have a near 50/50 chance of my server booting if it's headless. I can plug in a monitor to the server and see where it stuck at if the WebGUI hasn't responded in some time after boot and it looks to always be stuck here. Not sure if there is a log or something else I can produce that gives better insight. I didn't have any issue until enabling this NVidia plugin and plex use of it. would anyone have any thoughts? im using unraid 6.8.2

 

20200215_154731.jpg

I'm in the exact same boat with a P2000.

Link to post
  • Replies 2.5k
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

DEPRECATED

v6.8.2 uploaded.   Delayed for a few reasons, had problems (and still do) with the nvidia container runtime, worked around it in the end, but not a long term solution looking forward, I'm wo

DEPRECATED

Posted Images

7 hours ago, j0nnymoe said:

Only thing we can suggest is, trying another usb stick? reseat the GPU or use a different PCI-e slot?

It is directly tied to the nvidia install though. It boots first time, every time if it's not installed. Install it and I can't boot. Period. 

 

That, to me, seems like it's not a USB problem. I will try to reseat though.

Link to post

Not sure if I have been lucky so far but after posting my question about being able to boot headless with my P2000 GPU, I have been lucky 4 times in a row starting last night. I mentioned that I was using unraid 6.8.2  but I didn't say that I just updated to it. since the update I re-installed the plugin and so far it has been working. nock on wood but maybe the update fixed it... last night a parity check kicked off so im waiting for that to finish before I attempt another headless reboot. will update you if it continues to behave. 

Link to post
5 hours ago, rmeaux said:

piHole on your network?

I do have a piHole, and I did see some others having issues with that, but...

  1. I checked the query log, and refreshed after clicking on the plugin, and don't see the piHole blocking anything from "*digitalopenspaces.com".
  2. My unraid network settings, are not set to use the piHole for DNS. I have unRaid set to use 1.1.1.1
  3. I disabled the piHole for 5 minutes, and tried, and had the same issue. 
Link to post
5 hours ago, guyonphone said:

I do have a piHole, and I did see some others having issues with that, but...

  1. I checked the query log, and refreshed after clicking on the plugin, and don't see the piHole blocking anything from "*digitalopenspaces.com".
  2. My unraid network settings, are not set to use the piHole for DNS. I have unRaid set to use 1.1.1.1
  3. I disabled the piHole for 5 minutes, and tried, and had the same issue. 

Have you changed anything else on your network?

Link to post
45 minutes ago, guyonphone said:

Not that I can tell, or am aware of. I can browse to https://lsio.ams3.digitaloceanspaces.com/ in a browser and I get an XML page. So I don't believe I am being blocked. 

Have you had to change any MTU settings or anything? there's not really much else I can suggest looking at.

All previous cases of this is there had been some change to a users network.

  • Thanks 1
Link to post
7 hours ago, j0nnymoe said:

Have you had to change any MTU settings or anything? there's not really much else I can suggest looking at.

All previous cases of this is there had been some change to a users network.

Ah! Shoot, yes, you are right. I did change my MTU setting, to Jumbo frames a while back! I am using 10 gigabit so it's really been helping the speeds, and so far there hadn't been any negative consequences. I take it that's the cause? 

 

The thing that get's me here, MTU is Layer 2, I have a firewall. So when unRaid sends packets to my firewall, the MTU size is 9000, but the firewall should be tearing down those ethernet frames and rebuilding them to send to the cable modem at 1500 MTU with it's mac address, and in return the MTU size should be 1500 from the modem to firewall to unRAID on the return path. 

 

I guess I don't exactly get why this is hanging up this plugin. unRAID otherwise has perfectly fine access to the internet. 

 

Thanks!

Edited by guyonphone
Link to post
34 minutes ago, surfshack66 said:

Hello - So I have 2 GPUs in my server. 1 is dedicated to plex transcoding. I tried assigning the other to a VM but it crashed KVM. Has anyone else experienced this issue?

i have 3 GPU and no problems at all. 1 for GUI boot (gt710), 1 for Plex (1050ti) and a Vega64 for the VM's (reset bug is still there).
are you sure you didn't assigned the wrong one?

Link to post
On 2/17/2020 at 1:19 PM, sjaak said:

i have 3 GPU and no problems at all. 1 for GUI boot (gt710), 1 for Plex (1050ti) and a Vega64 for the VM's (reset bug is still there).
are you sure you didn't assigned the wrong one?

Interesting. Also, I'm sure I didn't assign the wrong one.

 

EDIT: Both cards the same. Not sure if that matters.

Edited by surfshack66
Link to post

I am implementing Nvidia GPU support into my Zoneminder docker and having an issue with the driver version.  I am installing cuda and cudnn version 10.2 and compiling support for the GPU in a module called 'opencv'.  I do not have a Nvidia GPU so I depend on users to provide feedback.  ndivia-smi reports driver version 440.40, but cuda wants to install nvidia utils for 440.33.  I can get the compile to work by forcing cuda-10.2 to be installed.  This is fine for the moment, but when I do an apt-get update, cuda wants to update the 440.33 driver which isn't installed, so it fails.

 

I have no idea what triggers the 440.33 driver, when the 440.40 driver is installed.  Any ideas on where I can go to get an answer?  I can carry on this conversation via PM if desired.

Link to post

 

On 2/15/2020 at 8:54 PM, jcarroll said:

I'm in the exact same boat with a P2000.

Similarly, I'm running an Xeon E5-2648Lv2 on a Supermicro X9SRH-7F board and no matter what NVIDIA Unraid build I try, it will kernel panic with my Quadro P400 inserted. Also, it will kernel panic MOST times (90%+) on any build when the card is not even inserted. My processor has the VT-d instruction set and by all other metrics it looks like the Unraid Nvidia builds should work, but they won't. I always get one of two kernel panics after unpacking initramfs.

Link to post
53 minutes ago, Distinguished said:

 

Similarly, I'm running an Xeon E5-2648Lv2 on a Supermicro X9SRH-7F board and no matter what NVIDIA Unraid build I try, it will kernel panic with my Quadro P400 inserted. Also, it will kernel panic MOST times (90%+) on any build when the card is not even inserted. My processor has the VT-d instruction set and by all other metrics it looks like the Unraid Nvidia builds should work, but they won't. I always get one of two kernel panics after unpacking initramfs.

6.8.1 worked flawlessly. Is there a way to downgrade?

Link to post
7 minutes ago, Stevefo said:

How do I properly update Nvidia Unraid? I assume I should upgrade stock Unraid 6.8.1 to 6.8.2, then upgrade Nvidia Unraid from 6.8.1 to 6.8.2

 

No.  Just go to the plugin and install Nvidia 6.8.2 - a full version of Unraid with extra drivers.

  • Like 1
Link to post
  • trurl locked this topic
Guest
This topic is now closed to further replies.