For some reason, and with no system changes, Plex has suddenly stopped being able to serve movies. The only change to my system was adding gigabit Fios speeds, but I kept my same router, so none of my IP addresses or port forwarding rules, should have changed.
Over the weekend, Plex just started to take forever to pull up the “title card” page for a movie, and then would just show a spinning circle. Occasionally, after a few minutes, the movie would start to play, but would freeze after a minute, and would generally stutter.
In short, performance has taken a mysterious hit. I’ve restarted a few times, but that didn’t seem to help.
I’ve attached my logs from this morning, trying to start a couple of titles on a few different devices.
Any insight and advice would be tremendously appreciated.
Jun 19, 2017 11:07:47.185 [0x7fe46f7fd700] DEBUG - Setting container serialization range to [0, 15] (total=1657)
Jun 19, 2017 11:07:47.188 [0x7fe46a3fd700] WARN - SLOW QUERY: It took 220.000000 ms to retrieve 57 items.
Jun 19, 2017 11:07:47.189 [0x7fe45abfa700] DEBUG - Setting container serialization range to [0, 11] (total=-1)
Normal resolution of this is:
Hover over the “Libraries” pane in the Web UI to expose the ellipsis (...). Click it and “Optimize Database”. Allow it to complete
Settings - Server - Scheduled Tasks (the butler). make certain these 3 options are checked.
Per your playback issue, I am seeing a great deal of database activity going on in your logs.
Database forward migrations (as if installing a new version which skipped over several intermediate versio
A large amount of EPG change activity.
I see a lot of playback activity so will need you to be more specific on what will not play, which client(s) won’t it play on, and the conditions causing the failure. The XML of one item which won’t play and the client it won’t play on are of great help. If you can replicate this (attempt playback), wait 30 seconds, and capture the logs again, this is the most help in conjunction with the other information I’ve requested here.
A large amount of EPG change activity.
Ahhhh. I hooked up a SiliconDust Homerun yesterday and started using the Plex DVR. Could that be what you’re seeing re: EPG change activity?
I see a lot of playback activity so will need you to be more specific on what will not play, which client(s) won’t it play on, and the conditions causing the failure. The XML of one item which won’t play and the client it won’t play on are of great help. If you can replicate this (attempt playback), wait 30 seconds, and capture the logs again, this is the most help in conjunction with the other information I’ve requested here.
I was trying to play a few different movies that are 720x480 (ish), AAC stereo 320kbs audio, h264 encoded mp4s. They would not load/play both on my web browser client or on my Android app client.
Here’s the most recent one. New logs coming in a minute.
This XML file does not appear to have any style information associated with it. The document tree is shown below.
ad9e-4501-8b5a-626f16364688/progress?progress=14.6&size=-22&remaining=42427&speed=0.2 (18 live) 0ms 326 bytes
Jun 19, 2017 16:33:00.204 [0x7fe46e364700] ERROR - Error issuing curl_easy_perform(handle): 28
Jun 19, 2017 16:33:00.204 [0x7fe46e364700] DEBUG - HTTP simulating 408 after curl timeout
Jun 19, 2017 16:33:00.204 [0x7fe46e364700] ERROR - HTTP 408 downloading url http://127.0.0.1:32400/library/metadata/4322?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&X-Plex-Client-Capabilities=videoDecoders%3Dh264{profile%3Ahigh%26resolution%3A1080%26level%3D41};audioDecoders%3Daac{channels%3A2}%2Cac3{channels%3A6}&X-Plex-Session-Identifier=b728dc4ee928c9aafbb99e853cf81c31
Jun 19, 2017 16:33:00.204 [0x7fe46e364700] ERROR - TranscodeUniversalRequest: unable to get container: /library/metadata/4322?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx&includeBandwidths=1&X-Plex-Client-Capabilities=videoDecoders%3Dh264{profile%3Ahigh%26resolution%3A1080%26level%3D41};audioDecoders%3Daac{channels%3A2}%2Cac3{channels%3A6}&X-Plex-Session-Identifier=b728dc4ee928c9aafbb99e853cf81c31
Jun 19, 2017 16:33:00.204 [0x7fe46e364700] WARN - Failed to obtain a streaming resource for direct play of part 15040
Error 28 is a timeout. The specific event happening at the time is it’s attempting to report playback status and can’t because the server (or host) is that busy.
Can you check it’s loading? You don’t have a weak box by any means for this task. Still, the error you captured (screen cap) does support it’s “outta gas”
There is literally nothing on that box except Plex. Its an Ubuntu Server 17.04 that only runs Plex. There should be nothing else at all going on in the background.
In /var/lib/plexmediaserver, rename Library to Library.SAVE
“Spin up” a new server instance without including the EPG.
Benchmark its performance
Add the EPG and related DVR/LiveTV
Benchmark again
There is a mysterious issue with the EPG we’re trying to actually identify. It seems to manifest as ‘slowness’.
I would like to see if this is the situation with your use-case.
@ChuckPA said:
Would you be willing to conduct an experiment?
Of course!
If so, I would like you to
Stop PMS
In /var/lib/plexmediaserver, rename Library to Library.SAVE
“Spin up” a new server instance without including the EPG.
Benchmark its performance
Add the EPG and related DVR/LiveTV
Benchmark again
I was in the middle of doing this with a friend of mine (my linux guru) and he was stumped, too. And then he got booted off of my server and couldn’t event move an mp3 file. Thats when he suspected something was wrong with the network card or cable.
There is a mysterious issue with the EPG we’re trying to actually identify. It seems to manifest as ‘slowness’.
I would like to see if this is the situation with your use-case.
So it seems like the “slowless” and everything backing up was due to a failing ethernet cable. Guess I’ll be putting some new endcaps on those wires tomorrow.
I can’t thank you enough, though, for sticking through this and offering up your technical insight and support. Great forum. Great people. THanks, ChuckPA!
Bravo! Good find. Get some end caps or just spend the few bucks on machine-crimped cables. I find them better than anything I can do because the boot is perfectly fitted to further reduce strain.
