PM: Version 4.14.2
Windows server 2012
duel cpu E5-2670 @ 2.60GHz
memory 64gig ( 10gig for ramdisk temp files for plex)
Gpu : vnidia s2000 card.
10G nic
1 Gig internet connection
Data base located on a NAS with a 10 gig nic Qnap
running PMS off the qnap and off my duel cpu
Test video Back to the future 1080p
This log is just after the server crash.
Any ideas? should I switch to ubuntu? or just build a new PMS server?
You media files can be on the QNAP but Plex should run off of local drives. Some of the crashes were because the Nvidia Card ran out of memory doing transcodes.
I have a qnap that is running plex off it AND i also have a duel cpu that has plex installed also. @JasonNalley
I got the Duel cpu since the qnap was limited on how many ppl can connect to it. I mapped the windows 2012 r2 server to the qnap so i can add the movie and tv show folders to the duel xeon.
Have you tried shutting down the Plex server on the Qnap and invited the users who use that one to the Windows Server? Running 2 instances of Plex just seems like a bad and confusing idea when they’re both accessing the same data…
But, on to your issue. I must then ask, are the crashes happening for the users connected to the Qnap server? Or the users connected to the Windows Server?
So now that I know which I am troubleshooting… According to Tiebierius your Nvidia card is running out of memory while transcoding and that’s causing the transcodes to fail and the users experience to crash.
Go to Settings>Transcoder
And can you paste me the screenshot of that page? Here’s mine for reference:
Hmm… Then I am not sure what’s going on to be honest. Tiebierius can read log files better than me (honestly a frog could read them better than me) so I trust his judgement that the card ran out of memory, and that would certainly cause the transcodes to crash.
What I don’t get is, I have a 1070 Ti that I unlocked to do unlimited transcoding, and I’ve tested 12 transcodes on it and it worked flawlessly. The 1070 Ti only has 8GB of Vram, so it’s not like I should be able to do more than double the amount that you’re reporting fails…
You may want to use something like GPU-Z or GPU Shark to see how much Vram your card is reporting, and how much you’re using… Because I see no reason this should be crashing…
Edit: This is what GPU Shark looks like, it’s quite handy
That is why i been scratching my head with this for a while. If i can get this to work then i will not have the same issue when i expand to a storintor q30 in the future. I can not see myself expanding when i have this issue. they ppl i reached out to can not figure out why either.
