Out of Memory: Kill Process (Plex Media Server)

Hey guys, first let me say thanks for all the help, I've been able to search through the forums for most of the other

 

issues I've encountered and found enough help to resolve my issue.

 

My setup is as follows

 

unRAID Version: unRAID Server Pro, Version 5.0

Motherboard: ASUSTeK COMPUTER INC. - P8Z77-V LK

Processor: Intel® Core i3-2120 CPU @ 3.30GHz - 3.3 GHz

Cache: L1-Cache = 128 kB (max. 128 kB)

L2-Cache = 512 kB (max. 512 kB)

L3-Cache = 3072 kB (max. 3072 kB)

Memory: 16384 MB (max. 32 GB)

4096 MB = BANK 0, 1333 MHz

1333 MHz = 4096 MB, BANK 1

1333 MHz = 1333 MHz, 4096 MB

BANK 2 = 1333 MHz, 1333 MHz

4096 MB = BANK 3, 1333 MHz

1333 MHz = ,

Network: eth0: 1000Mb/s - Full Duplex

Uptime: 4 days, 7 hrs, 5 mins, 15 secs

 

Running plugins

 

Native Plugins

OpenSSH

Plex Media Server

 

 

Unmenu Plugins

Plex

bwm-ng

htop

lsof

unraid status alert email

monthly parity check

 

 

As of recent (maybe over the past 2-3weeks) i'I've been having an issue while playing back media through Plex Media Server, where the content will start to stream, and then after a few minutes, the stream will stop, when i try to

 

restart it, Plex Media Server is no longer running.

 

I've checked my syslog, and it shows an 'out of memory' condition followed by it killing PMS

 

****

Dec  3 13:53:46 STORSERV kernel: Out of memory: Kill process 11279 (Plex Media Serv) score 2 or sacrifice child

Dec  3 13:53:46 STORSERV kernel: Killed process 11279 (Plex Media Serv) total-vm:316424kB, anon-rss:33892kB, file-rss:11112kB

****

 

I'm sure something is leaking memory but i'm not savvy enough to be able to identify which plugin it is?

 

I've tried the SwapFile plugin without success, and I've tried re installing some of plugins without any change either.

 

I can usually just restart the Plex Media Server and it'll come back online, but i'd prefer to find the plugin thats causing the problem, then treat the symptom.

 

I'll attach my full syslog as well, in case that will help.

 

 

vmstat shows

****

Linux 3.9.6p-unRAID.

root@STORSERV:~# vmstat

procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----

r  b  swpd  free  buff  cache  si  so    bi    bo  in  cs us sy id wa

2  1      0 3484316 145364 12294496    0    0    9    1  17  21 10  7 73 11

 

****

 

free -l shows

****

root@STORSERV:~# free -l

            total      used      free    shared    buffers    cached

Mem:      16542820  13750824    2791996          0    145800  12984332

Low:        768168    694676      73492

High:    15774652  13056148    2718504

-/+ buffers/cache:    620692  15922128

Swap:            0          0          0

 

****

 

any help would be greatly appreciated!

Thanks

rs

syslog-2013-12-03.zip

You actually state the best way to solve this.. Find the plugin !

 

Easiest thing is to one by one disable them and see when the issue goes away..

 

You could play with OOM_ADJ (search forums) to make some other process more likely to get killed.. Sabnzbd is a memory hog, you could set that a likely kill candidate.. You could even write a small script to see if it is running and starting it if it is not.. But thats all bandaids.. finding the culprit is better !

You could also check to see if your PMS plugin is up to date...in case the Plex folks have been plugging memory leaks.

 

 

 

I had similar issues. Im using PMS, crashplan and and a few other modules.

I have 6GB ram and using version 5.0.3

 

What stopped this happening for me was to add this to my GO file:

 

echo 65536 > /proc/sys/vm/min_free_kbytes

 

Worth a try?

 

Tony

Thanks guys for all the replies, i thought it might have been the 'Openssh' plugin since that was the most recent one that i installed, but after 4 days of running with Openssh disabled, the same Out of Memory errors popped back up.

 

I then tried Tony's suggestion (echo 65536 > /proc/sys/vm/min_free_kbytes), which seems to be limiting the maximum amount of memory that linux is using (before the change, unraid would hover around 12GB of memory usage, now its hovering around 7GB).

 

But the problems are still here, i'll try to disable more plugins to see which one is causing my pains.

 

unless anyone has any other ideas?

Don't mean to resurrect any dead threads, but I'm having this exact same issue.

 

My setup is considerably less powerful than the OP's, but I've seen smaller machines handle full HD movies:

Atom D510 2.6 GHz

4GB RAM

Unraid 5.0.5

Plex Media server v0.9.9.7.429-f80a8d6

Plex Connect cloned May 13th (2c6418e3fd6c9c9034b007bc99ec83ccf0152fa1)

Python 2.7.6

 

while writing this, I got two dumps to my syslog:

https://www.dropbox.com/s/o0betaut4iuayi1/dump1.txt

https://www.dropbox.com/s/f7capzpky6jb7e0/dump2.txt

 

I got these playing a full 1080p movie (33GB mkv), but I played a 1.3GB mkv all the way through just fine.

 

Hope these logs help to shed some light on the issue

Hey BDHarrington, in my attempt to troubleshoot my issue i ended up disabling all of my plugins, but i still had strange/random lockups about once a week (followed by a lengthy consistency check).

 

I considered dropping my system memory down to 4gb (From 16gb) and doing a fresh install without any plugins, but just as i was about to remove the dimm's version 6.0b3 got its release, so instead i installed 6.0b3 and its been working great, Previously my uptime wouldn't be longer than a week without a strange lockup; now its 64days of uptime.

 

on a side now, i have since moved my plex service to a vm on my hyper-v host. I couldn't be happier with v6.0b3!