Using an nvidia shield as the player and a WD pr2100 as the server. HDhomerun quatro as the tuner. When pausing live for a long period of time, resuming the paused stream it locks up and crashes. Basically losing the spot of the paused content. I’ve had this happen multiple times. I have been testing live tv for a while and come back and do testing and reviewing of the new player and server for improvements.
Server Version#: 1.16.6.1592
Player Version#: 7.20.0.11992
Still having this issue… Anyone else?
Time shifting live tv is a bit of a hot mess for quite a while now. Hopefully it’s eventually addressed.
How long are you pausing it?
There are a few things that can impact this:
Stream timeout setting
Available space for transcoding
Low Ram/CPU
The WD is definitely on the low end of the cpu side and unless direct playing the stream will struggle
I usually test between 3 to 5 minutes using direct play. I have also monitored the resources and nothing seems to be getting high usage. I have also had the same issues testing on a Windows 10 desktop with Ryzen 2700x with 16 gigs of ram, gtx 1660 for video. Not sure what’s going on with the resume crashing.
If the same symptoms with two different servers then it is either the client or network. You can try setting the “Terminate Sessions Paused for Longer Than” to 0 in server >settings > network (Advanced). Some clients would crash when this was > 0.
The Shield has had its share of problem lately but it is a focus for both Nvidia and Plex. With the new models coming out soon this should be resolved soon.
Same thing after changing the terminate paused sessions to 3600.
This seems to be the error that I see in the logs when the live session crashes.
TimeNov 05, 2019 15:09:55.431
LevelError
Thread0x7f1f97735700
MessagedownloadContainer: expected MediaContainer element, found html
(10 live) TLS GZIP 8ms 468 bytes (pipelined: 5)
Nov 05, 2019 15:10:37.007 [0x7f1f95cd7700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.079 [0x7f1f9540d700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.132 [0x7f1f95cd7700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.157 [0x7f1f962b3700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.218 [0x7f1fe6266700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.298 [0x7f1f97735700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.324 [0x7f1f9540d700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.387 [0x7f1f95fc5700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:37.415 [0x7f1f9688f700] Error — downloadContainer: expected MediaContainer element, found html
Nov 05, 2019 15:10:42.315 [0x7f1fe710c700] Debug — Request: [192.168.2.152:34472 (Allowed Network (Subnet))] GET /:/timeline?
Hopefully someone reads this.
Thanks!
I’ve been testing this and trying to get it to work and it seems if I force the server to transcode the stream it doesn’t crash. Example, a 1080 stream forced down to a 720p stream. Still no idea what’s up, but this is the only work around so far.
Just to point out this is still an issue, with the exact same error. Latest PMS on Debian with the Plex Shield client. The web client seems to work fine.
It’s pretty clear why this is happening – on the Shield when the screensaver kicks in, the Plex client does something where it causes the Plex server to stop recording live TV. Probably something to do with some sort of keep alive dying. You can see the effect by looking at the activity monitor of the server and watching when the Shield’s screensaver activates while the Plex client is paused. When it does, the activity monitor shows the live tv recording ends.
That’s interesting…assuming you tested with the screensaver deactivated and solved the problem?
With a longer delay from the default five minutes the recording still stops after awhile, but the five minute screen saver causes it to stop as well.
It still crashes with the screen saver set at 3 hours and me pausing it for a few minutes.
So unfortunately it doesn’t seem to be related to the screensaver… 
I am seeing this issue with server version Version 1.18.5.2260 (and shield client with 3 hour screen saver setting, roku, or desktop). I can see usually around 220-230 files in the transcode directory before it gets killed. In the debug log I found these entries (bolded), and I certainly did not initiate the termination…
Jan 25, 2020 10:40:40.152 [0x7f4582ffd700] VERBOSE - It took 0.0 sec to serialize a list with 0 elements.
Jan 25, 2020 10:40:40.152 [0x7f45837fe700] DEBUG - Completed: [127.0.0.1:58934] 206 PUT /video/:/transcode/session/e9da9ab2-1049-4300-bd9e-412f4272f90b/35bc5a65-7742-474b-83e6-9fd9f5eab0f8/progress?progress=-1.0&size=-22&remaining=-1&speed=1.0 (12 live) 0ms 355 bytes (pipelined: 469) (range: bytes=0-)
Jan 25, 2020 10:40:40.346 [0x7f4581ffb700] DEBUG - Shutting down idle session e9da9ab2-1049-4300-bd9e-412f4272f90b (idle time is 180 seconds)
Jan 25, 2020 10:40:40.346 [0x7f4581ffb700] DEBUG - Killing job.
Jan 25, 2020 10:40:40.346 [0x7f4581ffb700] DEBUG - Signalling job ID 20116 with 9
Jan 25, 2020 10:40:40.346 [0x7f4581ffb700] DEBUG - Job was already killed, not killing again.
Jan 25, 2020 10:40:40.346 [0x7f4581ffb700] DEBUG - Stopping transcode session e9da9ab2-1049-4300-bd9e-412f4272f90b
Jan 25, 2020 10:40:40.346 [0x7f4588d6c700] DEBUG - Cleaning directory for session e9da9ab2-1049-4300-bd9e-412f4272f90b (/transcode/Transcode/Sessions/plex-transcode-e9da9ab2-1049-4300-bd9e-412f4272f90b)
Jan 25, 2020 10:40:40.350 [0x7f458beec700] VERBOSE - JobManager: child process with handle 20116 exited
Jan 25, 2020 10:40:40.351 [0x7f458beec700] DEBUG - Jobs: ‘/usr/lib/plexmediaserver/Plex Transcoder’ exit code for process 20116 is -9 (signal: Killed)
Jan 25, 2020 10:40:40.351 [0x7f45377fe700] DEBUG - Grabber: Freed a tuner on device://tv.plex.grabbers.hdhomerun/107563B4 (now 4 available)
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Grabber: Recording for ‘Live TV - Session 377287fd085ed765-com-plexapp-android’ on channel 3.1 stopped with status complete.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Grabber: Using recording status.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Grabber: This was a rolling sub, not assimilating.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Grabber: Operation for ‘Live TV - Session 377287fd085ed765-com-plexapp-android’ on channel 3.1 completed with status complete (Success)
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Activity: Ended activity f8b18359-978d-4a97-9796-71616c6a8080.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Activity: updated activity 1067b234-0cff-4f36-b484-21bc28226f4d - completed 100.0% - Refreshing Sub
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Activity: Ended activity 1067b234-0cff-4f36-b484-21bc28226f4d.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Grabber: Cleaning up rolling grab for channel 3.1.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Recorder: No more consumers, stopping.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Recorder: Asked to stop recording ‘Live TV - Session 377287fd085ed765-com-plexapp-android’ on channel 3.1.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Recorder: Stopping transcode session.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Job was already killed, not killing again.
Jan 25, 2020 10:40:40.351 [0x7f4536ffd700] DEBUG - Job was already killed, not killing again.
Jan 25, 2020 10:40:40.351 [0x7f4564ff9700] DEBUG - HTTP requesting GET http://172.16.0.121:80/discover.json
Jan 25, 2020 10:40:40.352 [0x7f4581ffb700] DEBUG - Transcoder: Cleaning old transcode directories.
Jan 25, 2020 10:40:40.352 [0x7f4536ffd700] DEBUG - Subscription: Cancelling grab for subscription 67: ‘Live TV - Session 377287fd085ed765-com-plexapp-android’ on channel 3.1
So, this is weird, but it seems that Plex thinks it has an “idle” session on its hands, and then just goes and terminates it. and when restarting the stream on the front-end there is a little bit of playing going on (approx. 30-60 seconds) and then the picture freezes. In the log the other messages show up, where I think it’s looking for the “container” and can’t find it, because Plex cleaned everything out.
I am having the same issue. I’ve tried this several ways. I have the most recent Plex server released on 3 machines: A Mint Linux i7 with 8 gigs of ram, Windows 10 i7 with 16 gigs of ram, and the new Shield Pro. I have tried this in the web client and the native client on windows 10, the Plex app on Shield, the Plex app on Roku, the Plex app on my Galaxy S10, and the Plex app on Amazon TV. TV source is a HD-Homerun Quatro. All of them act exactly the same.
I just tested this as I am typing with the shield client connected to the Windows 10 server. I can pause for very short periods of time without issue. But If I pause for just 5 minutes I have about 30 seconds of buffered video and then it freezes and the only way to resume is to back out to the guide and tune the channel again, losing everything in between that point and whatever is live now.
I am not short of disk space on this server. I’ve got an entirely unused 2tb drive I can dedicate to this purpose, most of my media is on a NAS. There is no reason why I can only pause for a minute at a time without causing issues. Is there a setting where I can tell Plex to use as much space as it needs, or set the timeout for clients to something much more reasonable, 10 minutes at least so I can take a work phone call without losing what I was watching?
Just to chime in and keep this issue open, I’m having the same issue with shield client and windows PMS. Resources are not an issue and internal network is fine. Live play is no issue, just resuming after pause.
I’ve been running into the exact same issue. Windows 10 server with a core i5, 16GB RAM, 512GB SSD for the OS, dedicated 128GB SSD just for recordings/transcoding (so resources are not an issue). I’ve confirmed the exact same behavior when streaming to an Nvidia Shield, 4K Firestick, and an older Xiaomi Mi box. The only way it seems to work properly is if I watch through a web browser where the stream is transcoded, but that’s not really a solution.
One thing I can confirm though is it’s not the Shield’s screensaver that causes the issue. I’m able to watch the Transcode/Sessions folder on my server and see the exact moment when it kills the session and deletes all the files. I just saw this happen after a show had been paused for 3-4 minutes, but it was about another 2 minutes before the Shield’s screensaver came on.
I’ve tried changing the “Terminate Sessions Paused for Longer Than” setting to 21600, toggled between the new/old video player, along with several other things.
Plex server version: 1.19.2.2737
Plex client version: 8.0.0.17276
Steps to reproduce the logs attached below:
plex_log.txt (394.5 KB)
Id just post up all the logs and exact steps to reproduce plus all the app version numbers. Logs from client and server. Give the time you started watching programme, the channel, and the time you paused and the time you unpaused.
Everything so Plex can reproduce.
Thanks, I’ve updated my post above to include steps to reproduce and the logs.
