4k direct play over Ethernet crashing

[Hemz]

Last night I was trying to play a 66gb 4k hevc file over direct play.

I have jellyfin setup on my Linux server with 32gb of ram and a 4 core Intel i3 12100f CPU. 

It has an nvidia Quadro 2000 as the dedicated transcoder but the only thing being transcoded is the audio from 7.1 to stereo. (Haven't upgraded my sound system from a 5.1 soundbar yet)

The issue was the movie stopping and causing the whole jellyfin server to crash while on the exo player as the playback on my NVIDIA shield pro 2019 version.

The only way to fix the crash was to restart the Linux server and I changed the player from exo player to libvlc (experimental) which allowed us to get through the whole movie.

Has anyone experienced a similar issue and knows of a fix? 

This is all local network over Ethernet direct play and I've only had issues with 4k stuff before.

The jellyfin logs show nothing out of the ordinary due to it being direct play.

[bitmap]

This doesn't sound like a Jellyfin problem, but a server problem. I'd look at your server's logs and see what the last outputs prior to the crash are there.

[Hemz]

This doesn't sound like a Jellyfin problem, but a server problem. I'd look at your server's logs and see what the last outputs prior to the crash are there.

I can't replicate this issue playing the file on a web browser though, it only does it when using the nvidia shield and only for large files. 

What files on the server would I look at because i've not done it before.

[TheDreadPirate]

Probably /var/log/messages. That should include just about everything going on with the OS side of the server.

Also, which OS? If Linux, which kernel?

Also, which Quadro "2000"? There are several generations of Quadro with 2000 in the model number. An M2000 has very different capabilities from the P2000, which has different capabilities from the A2000. I'm going to assume P2000 for now.

[Hemz]

Probably /var/log/messages.  That should include just about everything going on with the OS side of the server.

Also, which OS?  If Linux, which kernel?

Also, which Quadro "2000"?  There are several generations of Quadro with 2000 in the model number.  An M2000 has very different capabilities from the P2000, which has different capabilities from the A2000.  I'm going to assume P2000 for now.

Ubuntu 22.04

Quadro p2000

Found the logs and going through now.

Nothing that's not consistent throughout the whole log..

Main thing is failed to query DNS server and UFW block (but it was doing that when the server was working fine too)

Aug 12 19:53:06 hemflix kernel: [34593.598844] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:57:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=52281 DF PROTO=TCP SPT=36582 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:53:07 hemflix kernel: [34594.604185] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:57:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=52282 DF PROTO=TCP SPT=36582 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:53:09 hemflix kernel: [34596.624186] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=0:57:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=52283 DF PROTO=TCP SPT=36582 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:53:13 hemflix kernel: [34600.880191] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=0257:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=52284 DF PROTO=TCP SPT=36582 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:55:06 hemflix kernel: [34713.599434] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:7:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=31064 DF PROTO=TCP SPT=55714 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:55:07 hemflix kernel: [34714.604204] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:57:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=31065 DF PROTO=TCP SPT=55714 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:55:09 hemflix kernel: [34716.620213] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=0:57:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=31066 DF PROTO=TCP SPT=55714 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:55:13 hemflix kernel: [34720.684221] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:57:7e:02:42:ac:13:00:09:08:00 SRC=172.19.0.9 DST=192.168.0.8 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=31067 DF PROTO=TCP SPT=55714 DPT=8787 WINDOW=64240 RES=0x00 SYN URGP=0
Aug 12 19:55:25 hemflix kernel: [34732.458982] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:7:7e:02:42:ac:13:00:0d:08:00 SRC=172.19.0.13 DST=172.19.0.1 LEN=30 TOS=0x00 PREC=0x00 TTL=64 ID=64448 DF PROTO=UDP SPT=33873 DPT=5351 LEN=10
Aug 12 19:55:25 hemflix kernel: [34732.709191] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:7:7e:02:42:ac:13:00:0d:08:00 SRC=172.19.0.13 DST=172.19.0.1 LEN=30 TOS=0x00 PREC=0x00 TTL=64 ID=64452 DF PROTO=UDP SPT=33873 DPT=5351 LEN=10
Aug 12 19:55:25 hemflix kernel: [34733.208717] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:42:7:7e:02:42:ac:13:00:0d:08:00 SRC=172.19.0.13 DST=172.19.0.1 LEN=30 TOS=0x00 PREC=0x00 TTL=64 ID=64497 DF PROTO=UDP SPT=33873 DPT=5351 LEN=10
Aug 12 19:55:26 hemflix kernel: [34734.211035] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:42::7e:02:42:ac:13:00:0d:08:00 SRC=172.19.0.13 DST=172.19.0.1 LEN=30 TOS=0x00 PREC=0x00 TTL=64 ID=64551 DF PROTO=UDP SPT=33873 DPT=5351 LEN=10
Aug 12 19:55:28 hemflix kernel: [34736.210783] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:42::02:42:ac:13:00:0d:08:00 SRC=172.19.0.13 DST=172.19.0.1 LEN=30 TOS=0x00 PREC=0x00 TTL=64 ID=64736 DF PROTO=UDP SPT=33873 DPT=5351 LEN=10
Aug 12 19:55:56 hemflix kernel: [34764.209243] [UFW BLOCK] IN=br-82a708e84643 OUT= MAC=02:42:42:ac:13:00:0d:08:00 SRC=172.19.0.13 DST=172.19.0.1 LEN=30 TOS=0x00 PREC=0x00 TTL=64 ID=3549 DF PROTO=UDP SPT=33873 DPT=5351 LEN=10
Aug 12 19:56:45 hemflix dockerd[1069]: time="2023-08-12T19:56:45.002467043Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;cyteam.ru.\tIN\t AAAA" error="read udp 127.0.0.1:57811->127.0.0.53:53: i/o timeout"
Aug 12 19:56:47 hemflix dockerd[1069]: time="2023-08-12T19:56:47.505110010Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;cyteam.ru.\tIN\t AAAA" error="read udp 127.0.0.1:59923->127.0.0.53:53: i/o timeout"
Aug 12 19:56:49 hemflix dockerd[1069]: time="2023-08-12T19:56:49.003553116Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;cyteam.ru.\tIN\t AAAA" error="read udp 127.0.0.1:35872->127.0.0.53:53: i/o timeout"
Aug 12 19:56:55 hemflix dockerd[1069]: time="2023-08-12T19:56:55.001761900Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;chihaya.toss.li.\tIN\t A" error="read udp 127.0.0.1:51666->127.0.0.53:53: i/o timeout"
Aug 12 19:56:55 hemflix dockerd[1069]: time="2023-08-12T19:56:55.001767339Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;chihaya.toss.li.\tIN\t AAAA" error="read udp 127.0.0.1:34451->127.0.0.53:53: i/o timeout"
Aug 12 19:56:57 hemflix dockerd[1069]: time="2023-08-12T19:56:57.501922077Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;chihaya.toss.li.\tIN\t A" error="read udp 127.0.0.1:41757->127.0.0.53:53: i/o timeout"
Aug 12 19:56:57 hemflix dockerd[1069]: time="2023-08-12T19:56:57.501930995Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;chihaya.toss.li.\tIN\t AAAA" error="read udp 127.0.0.1:50438->127.0.0.53:53: i/o timeout"
Aug 12 19:56:59 hemflix dockerd[1069]: time="2023-08-12T19:56:59.002441571Z" level=error msg="[resolver] failed to query DNS server: 127.0.0.53:53, query: ;chihaya.toss.li.\tIN\t AAAA" error="read udp 127.0.0.1:55918->127.0.0.53:53: i/o timeout"

[Venson]

Hey @Hemz,

If the whole server crashes this sounds like a bigger problem then just with jellyfin but could you share your server jellyfin logfiles? You can download them via the admin dashboard -> logs -> top most element.
In your case please produce a crash, then (after a restart) select the 2nd top most one as each restart creates a new logfile. It should start with log_ not with ffmpeg_

[Hemz]

Hey @Hemz,

If the whole server crashes this sounds like a bigger problem then just with jellyfin but could you share your server jellyfin logfiles? You can download them via the admin dashboard -> logs -> top most element.
In your case please produce a crash, then (after a restart) select the 2nd top most one as each restart creates a new logfile. It should start with log_ not with ffmpeg_

I'll replicate it tomorrow and grab the logs. 

Just for reference, only my jellyfin docker compose image crashes not the whole server. Doing a restat of the compose fixes it.

[Hemz]

Same issue as before but happening for transcodes and direct ay intermittently.

Getting this line until I restart like the transcoding drops my gpu!

cu->cuInit(0) failed -> CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected