Server Version#: 1.13.9.5456
Player Version#: N/A
Noticed over the last 2 weeks or so that there seems to be a potential memory leak with PMS. As you can see in the attached images there is a gradual climb in RAM utilization and after I manually restart the Plex Media server application RAM goes back down to normal levels. Before I noticed this, RAM would spike periodically during trasncoded sessions but would then return back to normal levels.
Any else seeing this on their Synology servers?
Nice graphic but I need more to go on than just that please.
What changed?
Please expound
Only thing that has changed is the PMS version, no updates to DSM, this happens irregardless of media being played or not, and no other apps were added.
At minimum, I need to see the log files, and know which version you were running.
Not sure which log is the right one, attached the whole download. Server Version#: 1.13.9.5456.Plex Media Server Logs_2018-11-15_12-13-02.zip (3.8 MB)
THanksā¦ The entire zip is best. I can also check all the subsystems at the same time
HOWEVER,
Would you please DISABLE Verbose logging until requested?
Verbose yields only 2 minutes of playback time.
All the data I need is long flushed from the buffer.
Thanks.
Sorry about that! Reattached following your instructions. I noted that this entire time the RAM load was as expected (total RAM utilization at around 30%) throughout entire playback. If this follows the same trend as previous weeks, Plexās RAM utilization will continue to rise until I restart Plex again at which point it comes back down to a reasonable level.
Iām also seeing this behavior on the latest official Docker image running on Fedora Core 28 (running in ESXI). I actually have re-set up my server a few times in the hope that it was something that I did in the docker setup, but it doesnāt seem to be the case.
I do not have the generation of preview or chapter thumbnails enabled. I also am not analyzing audio tracks (as I havenāt added a music library the last time I set up the container).
Currently, the container itself is holding onto just under 4GB of memory at 10 minutes (It was hitting 6GB, which in turn was maxing out the 8GB that I have assigned to the VM) after playing a movie with some transcoding. Itās currently not really causing any issues for me, but Iād rather it not be spilling over into SWAP if I can help it since itās running on a SSD.
Iāve tried hardware encoding off and on as well.
Iāll get some logs attached at some point if itād help.
Please do not confuse docker with native PMS. IT has different conditions and operational environment.
I understand that. Just stating my experience in case thereās a possible common issue.
I have these enabled but already had them set to run as a task. During those hours, PMS ramps up CPU usage (as expected) but does not have any noticable effect to RAM usage. Itās quick clear there are spikes in CPU load during maintenance hours but not for RAM (odd).
Either way, Iām not sure how this is relevant to a slowly ramping up memory load that is reset by restarting PMS.
Just wanted to provide an update: I did a fresh PMS install and it corrected my issue.
Could you clarify what yo mean with āfresh installā ? Did you remove entire library etc? Iām suspecting similar issues but would prefer to not have to spend hours of building my library up again. (It should of course be quick, but in practice that has unfortunately never been the case).
I also have this issue. Hasnāt been too much of an issue as the NAS has been restarted every few days, but today the app on the TV started buffering and when I tried to investigate I couldnāt access the NAS - it was to busy. Ended up having to shut it down (killed UPS power so it would shut down gracefully).
DS918+
Plex 1.14.0.5470
LG Smart Tv x 2 Samsung x 1
No Thumbnails or previews being generated.
Iāve enabled logging now.
Get back to us when you have logs.
FYI. If you press and hold the Blue power button until it beeps, DSM will perform an orderly shutdown. BE ADVISED If you continue to hold it, it will Power Off which is not desired.
Do you by chance have your APK for .5470? I would likee to upgrade to that over the newest build.
Here are the log files. Started from a complete new log through to the end off today. RAM usage went from 7% at start through to 87% tonight. When I stopped Plex RAM dropped back to 2% then 7% when Plax was restarted.
Logs.7z (3.0 MB)
Sorry, I donāt normally keep SPK files after theyāve been applied.
Thanks for the logs. Thereās a lot going on in there.
Plex Media Server.4.log:Dec 21, 2018 13:01:45.068 [0x7fde09847700] ERROR - [Transcoder] [mp4 @ 0x32a5c40] Application provided duration: -32 / timestamp: 83259304 is out of range for mov/mp4 format
Plex Media Server.4.log:Dec 21, 2018 13:01:45.205 [0x7fde09b35700] ERROR - [Transcoder] [mp4 @ 0x32a5c40] Application provided duration: -32 / timestamp: 86467480 is out of range for mov/mp4 format
Plex Media Server.4.log:Dec 21, 2018 13:02:03.284 [0x7fde2bbff700] WARN - LongPoll: Got error, closing.
Plex Media Server.4.log:Dec 21, 2018 13:02:03.286 [0x7fde2b911700] DEBUG - NotificationStream: Removing because of error
Plex Media Server.4.log:Dec 21, 2018 13:03:37.807 [0x7fde2b911700] WARN - LongPoll: Got error, closing.
Plex Media Server.4.log:Dec 21, 2018 13:03:41.910 [0x7fde221a1700] DEBUG - Job running: EAE_ROOT='/volume1/Plex/tmp_transcoding/pms-c8215932-5a9c-49f4-981a-8e4bdbb9ab20/EasyAudioEncoder' FFMPEG_EXTERNAL_LIBS='/volume1/Plex/Library/Application\ Support/Plex\ Media\ Server/Codecs/531e313-1328-linux-ubuntu-x86_64/' XDG_CACHE_HOME='/volume1/Plex/Library/Application Support/Plex Media Server/Cache' XDG_DATA_HOME='/volume1/@appstore/Plex Media Server/Resources' X_PLEX_TOKEN='xxxxxxxxxxxxxxxxxxxx' '/volume1/@appstore/Plex Media Server/Plex Transcoder' '-codec:0' 'h264' '-noaccurate_seek' '-probesize' '10000000' '-i' '/volume1/PlexMedia/Miscellaneous/The Magical World of J.K.Rowling (2000)/The Magical World of J.K.Rowling (2000).mkv' '-map' '0:0' '-metadata:s:0' 'language=eng' '-codec:0' 'copy' '-map' '0:1' '-metadata:s:1' 'language=eng' '-codec:1' 'copy' '-copypriorss:1' '0' '-f' 'dash' '-min_seg_duration' '5000000' '-skip_to_segment' '1' '-time_delta' '0.0625' '-manifest_name' 'http://127.0.0.1:32400/video/:/transcode/session/7bl97dmvl79obtl7d195txyd/8805ed98-9409-4c0a-adf1-2e4b1cdd2f08/manifest' '-avoid_negative_ts' 'disabled' '-map_metadata' '-1' '-map_chapters' '-1' 'dash' '-start_at_zero' '-copyts' '-vsync' 'cfr' '-y' '-nostats' '-loglevel' 'quiet' '-loglevel_plex' 'error' '-progressurl' 'http://127.0.0.1:32400/video/:/transcode/session/7bl97dmvl79obtl7d195txyd/8805ed98-9409-4c0a-adf1-2e4b1cdd2f08/progress'
Plex Media Server.4.log:Dec 21, 2018 13:03:42.395 [0x7fde09847700] ERROR - [Transcoder] [mp4 @ 0x1cdd140] Application provided duration: -32 / timestamp: 10216360 is out of range f
This looks like a bad installation. Something either damaged your PMS installation or that TV ID is completely unknown (which doesnāt make sense)
Plex Media Server.4.log:Dec 21, 2018 11:14:20.962 [0x7fde26623700] ERROR - Unable to find client profile for device; platform=webOS, platformVersion=4.1.0, device=, model=65UK6540PTD
It was released in May so there shouldnāt be any question about it falling into WebOS.
I will have to ask the TV App team to take a look at this part.
Plex Media Server.5.log:Dec 21, 2018 10:47:43.021 [0x7fde1f911700] ERROR - [MediaInspectorFreeImage] strptime on 0000:00:00 00:00:00 failed
Plex Media Server.5.log:Dec 21, 2018 10:47:43.024 [0x7fde1f911700] ERROR - [MediaInspectorFreeImage] strptime on 0000:00:00 00:00:00 failed
Are there attached/embedded images in some of your files?
Iām not trying to redirect here but somehow I think part of whatās happening is PMS is falling into a hole and most likely not cleaning up cleanly after an error of this type. I can do something about that.
Do you mind seeing if we can repeat this on just one file (this one),
Create a ācleaned upā version (saving your original)
Attempt to play it again
See what the logs show next time
Also, May I request you use the Download Logs button to retrieve Logs?
I"m not equiped to handle 7z (I work in Linux and 7zip is command line only but my editors are graphicalā¦ it gets VERY messy)
I have a couple of days next week when the system wonāt be in use at all. Would it be worth blowing it away and re-installing PMS from scratch?
I appreciate the analysis youāve done but if itās a corruption Iām probably going to end up doing a re-install anyway.
