sittingmongoose

Hmmm still didnt work...weird. Something is funky, nvidia isnt showing up either and that works perfectly in plex and tdarr. Odd.

I thought you shouldnt use modprobe if you are on 6.9.2?

Thanks, didnt work for me still. I am actually trying with emby not Plex. But I also cant get emby working with nvidia either...using settings from tdarr that work and trying my plex docker settings. running 6.9.2 so not sure whats going on. Plex works with nvidia currently.

So I removed the lines from my go file. How did you go about creating the empty file? I am a complete noob so I am struggling here.

So I got v2 setup and the node. Added the nvidia parameters to my node docker(plex works with my p2000). Selected the Winsome H265 NVENC plugin, turned on a gpu worker is its "working" but unraid isnt showing any processes on nvenc. So i am not quite sure if its actually working on GPU. I have a cpu worker disabled currently. To add to it, if I increase GPU workers, it works and my cpu usage doesnt go up.

Just an update for everyone. I reset my bios settings, pulled the battery and no more crashes. So I guess I’m good!

Anyone update to RC1 and see if this is resolved?

Any idea if @limetech has looked at this? Its been quite a while and I don't see any attempts to look at this problem. Its affecting beta 35 as well. Presumably this issue would follow forward into the next release as well...

Ah you are right, sorry. Either way, server passes 4 days on memtest with no issues so certainly not a problem with my ram. Also, back on beta 25 it’s not crashing at all.

Is there an easy to follow how to for logging to another computer? I might be able to use my dads windows computer as that’s on the same network.

I did post my diagnostics and enhanced syslog but I don’t see anything in there at all telling. And it’s random, it probably won’t crash overnight at all. But randomly during the day it will. I can’t even pin down a thing it’s doing to trigger it. don’t have an option to run a remote syslog as it’s kinda remote. Nor do I even know how to do that.

I am going to have to revert back to beta 25 tomorrow. I was trying to stick with it for a few days so @limetech could look at it and fix it for the future but I can’t wait any longer. It’s crashing every .5 to 2 hours. Besides rendering my server pretty much useless, it’s not shutting down gracefully so it’s only a matter of time till something dies. I’m worried this issue will continue to carry forward into future releases because it started with beta 29 for me.

I am also within spec, 4x16gb of ddr4 3000mhz. Memtest not showing any errors.

Another update...crashing about every hour randomly. I tried seeing if Plex might be the cause either through disk access or the gpu and nothing. I tried like 10 streams, direct play, transcoded and hardware transcoded and that didn't trigger it. Tried them all at the same time, tried mixing disks, tried only accessing movies on the one sas card versus the other, nothing.

asrock z390, 9900k, supermicro 846, 2x lsi 9207(ones an external model and the other is internal), p2000, 3x 1tb 970 evo plus ssds, dual 1200watt server psus. using 1 windows 10 vm but its just idling, I haven't touched the VM since I upgraded at all. I will try to disable the vm to see if that's it. I can't pin a rime or reason to it. No errors in logs, and it seems random, it seems to stay alive through the night and die a few times during the day.....and again, beta 25 is stable. I am going to roll back but I figured I would stay on this for a few days to let @limetech gather some data to fix it in the next release. Edit: VMs being disabled didn't change anything. Just crashed again. Going to try disabling hard accel aka not using the p2000. Edit 2: Crashed again with no P2000...So its not gpu related....

I am having a similar issue. Started with Beta 29, beta 25 was good. Beta 25 was the Nvidia build with a quadro p2000 and using hardware accel in my dockers. Never crashed. Now on beta 29 and higher, I get crashes every few hours. Ive been running for like 3 days now and I have had 12 crashes already. Completely unresponsive, can't access anything, I need to do a power cycle. Using the Nvidia Driver on beta 35 but on beta 29 I was using the Nvidia build. Absolutely nothing in my logs....Server is pretty much useless now, I have no idea when it dies, don't get any notifications or anything. So only way I know its down is someone messages me saying they can't access Plex. Attaching Diagnostics and Enhanced Syslogs Archive.zip

Awesome! Let me give unraid a day or so to try to figure it out before I roll it back. Its crashed 4 times today so I am happy there is an option.

Unfortunately I cant see anything happening. I cant access my server at all when it crashes. I don't have a monitor on it either. But yea I am running the nvidia driver and a p2000. But on beta 29, I wasnt running the nvidia plugin, I was using the nvidia build.

Getting complete system crashes with this build. I was previously running beta 25 which was stable for me. I upgraded to beta 29 and I saw complete system crashes(I would need to power cycle the server) so I rolled back to 25. Now I upgraded to 35 because of all the changes to community related stuff, everything works and I am not getting any errors...But it's still crashing like once every 24 hours. Looking in logs shows nothing at all. System is completely stable on beta 25...but obviously I can't roll back now because I can't easily get the Nvidia build for beta 25. I am attaching my diagnostics. Any help would be greatly appreciated. I don't even see errors to see where the problem could be. It crashed today at roughly 5:18pm if that helps. 9900k-diagnostics-20201127-1951.zip Edit....Literally just crashed again right after I submitted this post...

Ah, I only have beta 25 installed. That’s why I can’t see that plug-in in CA, gotcha. Thanks!

I am a little confused, I am not seeing "Nvidia Drivers" in Community applications? I am afraid to upgrade from beta 25 as I obviously now can not easily roll back without the Nvidia build plugin functioning anymore

Unfortunately I already reverted. Long story short, my server is temporarily not accessible and power and internet are going out often. So it was too hard to manage, Is this a crash? Is the internet out? Is the power out?(Yes, I have them on battery backup so when power goes out it would safely shutdown).

Just got another complete system crash. This is the second time in a day on beta 29 coming from beta 25. I never had crashes prior to this. I am attaching my diagnostics, syslog and enhanced sys log. I don't see any errors on my end. Rolling back to beta 25. 9900k-diagnostics-20201001-1740.zip 9900k-syslog-20201001-2140.zip 9900k-syslog-20201001-1739.zip

Which part happened to you before?

Seeing some weird things now. Upgraded from beta 25. My server completely crashed, hasn't done that in many months. Not seeing any errors anywhere. Had to pull the power to reboot it. Then there are some issues with docker. A bunch of the dockers say "not available" under the version. Andddd NZBget no longer works. But it did before the crash. Attaching diagnostics and Syslog. syslog.txt 9900k-diagnostics-20201001-0016.zip