Okay was having the crashing problem with 0.9.5.2 in a hidden folder called .PMS on my cache drive.  It would crash after approx. 12 hours.  I created a Cache Only share called PMS, moved my library there, and it's been up for almost (2) days straight.  Just wanted to add my experience to the thread.

 

I had similar issue .. also moved it outside of the hidden folder and it is also running without any issues since then

 

 

I never seemed to have a problem even inside a hidden /.data folder but just in case I moved it to /data (cache only) and still seem to have no problems for the last few days.  And for me, Plex is new so I'm playing about with it all the time running test streams to my roku, updating the library, tweaking metadata, etc.

glad it works for the most of you now.

 

n8 

Hey Spidi, so today marks the longest I've ran your PMS plugin without any reboot or stopping of the PMS app.  As I've previously noticed, the growing virtual memory has been pretty slow.  BUT! for the last day or so, the virtual memory has held steady at 517m.  I started using the web manager--switching library sections--and there was zero bump in memory usage.   This is fantastic!  I'll do my best to keep the uptime increasing and post back.  Definitely seems a bit higher than most, but I have no problem with that size if it doesn't continue to grow every day!

 

Thanks again for all your hard work!

 

Edited: Looks like it went down a few megs today to 514m.  Not to jinx things, but I think I'd call this stable.

Same here.  Looks like I have stabalized at:

 PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
3114 unraid-p  20   0  302m  44m 9252 S    0  1.1  23:09.85 Plex Media Serv  

 

 

And today I was particularly harsh on the server trying to stream a recent blueray rip converted via makemkv and a DTS-HD audio stream.  Starting and stopping it several times.  I think I even got a transcoding pid to not close as I saw two operating at a time, but the "ghost" closed eventually.  Coupled with a few cycles of the web console to boot.

 

Unraid uptime is a little over 5 days with plex server running the whole time.

I have a question to the community here to see if others have experienced similar issues as me with transcoding with Plex + Unraid + iOS.

I've found two seperate issues, one has already been confirmed by many users on the plex forum, the other i'm a bit unclear yet on what the cause is.

 

On both issues i need to restart PMS to get it working again, though others have had the same issue but can try to re-play the file without restarting PMS.

 

1. Subtitle Crash --> Plex + PMS Linux + iOS/Transcoding + Subtitles -->Transcoding crash <Already confirmed by many users on plexapp forum>

Issue where, on Linux versions of PMS together with a client requiring transcoding (Especially/particularly iOS / iphone) , a video will crash if subtitles are used.

Wether they are internal or external makes no differences. This issue has been found by many users so far on Synology, Debian and Unraid installs:

 

http://forums.plexapp.com/index.php/topic/32049-subtitles-on-pms-095-on-unraid/

 

Did anyone experience this issue as well?

2. Error in transcoding: Cannot get duration line --> Plex + PMS Linux(May be unraid specific) + Transcoding --> Random crash of transcoding <Only found this a few days ago>

 

At seemingly random moments, transcode dies with the following error, this happens on files that are normally perfectly fine.

Dec 09, 2011 15:01:20 [0xafe23b70] DEBUG - [Transcoder] Starting transcoder v306 (t=8s, q=2)
Dec 09, 2011 15:01:20 [0xb13c5b70] DEBUG - [Transcoder] Registered components, about to parse arguments.
Dec 09, 2011 15:01:40 [0xafb35b70] ERROR - We did not get a duration line from the transcoder in time (which may indicate badly muxed AVI), so we're aborting.

 

Did anyone experience this issue?

 

I have Plex 0.9.5.1 installed via the plugin, but when I start the Plex Media Manager, I can't configure the Library Sections to point to my media.  It's as if the page is read-only.

 

Try a different browser. A script or pop-up is probably being blocked. A window should come up where you can add you media folders.

 

i guess it has something to do with the browser you are using or some extensions...

 

@SCSI,

@Benni-chan,

A browser update to Firefox 8.0 did the trick.  Thanks for your suggestions!

 

Kevin

 

Thank you all  I was having this exact same problem with my library.

Updated Firefox and problem gone!

 

Plex 0.9.5.2 plugin

unRAID 5b14

I am in the process of building my unRaid server which I plan to run Plex on.  This memory leak thing has me worried.  I'm just now buying the hardware for my build.  Is there anything I can do now to give myself a leg up on this problem.  Would having more memory, say 8GB instead of 2GB give me a longer up time between crashes?  What about processor choice?  I read through the thread that polled users what processor they had and if they had the leak.  Anything else that I could do now to give me the best unRaid/Plex setup?

 

the only think you have to do at the moment is to avoid hidden folders (dot folders/ ".plex")

 

IMHO 4GB of RAM should be enough.   

plex seems to randomly not be able to connect sometimes.. anyone know why?  is it cause my drives aren't spinning?

 

seems to happen more on android then on ios though

So avoiding dot/hidden folders and having 4GB of RAM should keep Plex up and running for several days at least?  Seems like I read somewhere earlier in this thread that the leak was so slow it would take ~20 days to crash Plex. If this is the case I can definitely live with it.

I have 8 GB of RAM and I haven't seen the process grow over 550 MB yet. 

FWIW the process growth/crashes I was seeing were not in the main Plex process itself, but in the transcoding process.  Perhaps completely unrelated to the issue mostly discussed here, but the end result is that Plex lost not only all GAF (Girlfriend Approval Factor) ... frankly most of mine.  The last thing I want to deal with when sitting down with fresh popcorn and the lights dimmed is running over to my PC to stop/start Plex and kill any runaway transcode processes.

 

I've moved on to other solutions, but I am keeping my eye on Plex for future use.

That's interesting.  Even when the leak was at its worst, I never saw this issue related to transcoding.  I assume you've had a telnet session open and ran 'top -p' with the various PIDs to see the growth happen?  What are you transcoding to?  ATV2?

 

I've actually found this to be so stable that I've completely removed my Mac Mini-based PMS media library and now purely use the unRAID version.  Sorry to hear your experience is not the same.

That's interesting.  Even when the leak was at its worst, I never saw this issue related to transcoding.  I assume you've had a telnet session open and ran 'top -p' with the various PIDs to see the growth happen?  What are you transcoding to?  ATV2?

 

I've actually found this to be so stable that I've completely removed my Mac Mini-based PMS media library and now purely use the unRAID version.  Sorry to hear your experience is not the same.

 

I did notice some growth during a telnet session once, but my first clues were a failed stream to my Roku2 followed by a second attempt to stream, maybe even a third or fourth trying a different file.  Open //tower and see a transcode process crash in the log ... open Top and see an active transcode process (when no stream is happening).  Kill the process, stop and start plex, and then try to restream the first file with success.

 

Note that this is pushing to a Roku2 ... so for all I know, the issue is with the Roku2 Plex app not cancelling the failed stream correctly, but it begs the question why the stream failed in the first place.  Especially since a newly restarted Plex server was able to stream with no trouble. 

 

Anyway I didn't dig too far into it because I was getting frustrated and being xmas time I decided to throw money at it.  That way unRaid could go back to doing what it does best: store and serve data.  Let something else actually process video stream.

First wanted to say thanks for the great plugin! I have a problem that I am stumped on. I built an unraid server for plex recently and when I initially set it up I placed the library on disk1 as I didn't have a cache drive at the time. I recently added a cache drive and I moved my library to that and everything appeared to be fine until I tried to add a couple new movies to my library. Now whenever I add movies the scanner won't pick up any metadata it just shows the title and a frame from the movie. I have tried rescanning, deep scanning, fixing incorrect match but it never seems to find anything. Did I screw up something when I moved from Disk1 to the cache drive?  Let me know what other info you'd like (i am a beginner with Linux so I might need some help finding it )

 

Thanks in advance,

Jens

That's interesting.  Even when the leak was at its worst, I never saw this issue related to transcoding.  I assume you've had a telnet session open and ran 'top -p' with the various PIDs to see the growth happen?  What are you transcoding to?  ATV2?

 

I've actually found this to be so stable that I've completely removed my Mac Mini-based PMS media library and now purely use the unRAID version.  Sorry to hear your experience is not the same.

, but my first clues were a failed stream to my Roku2 followed by a second attempt to stream, maybe even a third or fourth trying a different file.  Open //tower and see a transcode process crash in the log ... open Top and see an active transcode process (when no stream is happening).  Kill the process, stop and start plex, and then try to restream the first file with success.

 

The effects your seeing seem extremely similar to my problems with iOS + failed transcode streams.(http://lime-technology.com/forum/index.php?topic=14803.msg157338#msg157338 ) My plex continues to work fine, even after this occurs, on clients that do not use transcode. To get transcode running, i need to restart PMS. Scroll up a few posts to see my problems

 

Would help if you post some logs.

 

Everything works fine but getting metadata? Sounds like an issue with the library. You may want to check file permissions. Perhaps when you moved the library to your cache, root become the owner. You can do a "chown -R unraid-plex:users <path to library folder>" and see if that fixes your issue. Or, while not ideal, you can always just whack your library and start fresh... If it worked before, something got mucked up along the way.

Like kaiguy said, it might be permissions.

 

- Also, how did you point the plugin to the new cache drive? /mnt/cache ?

 

- Are you using hidden folders? (e.g. .plex?) If not, your files may be moved to the array daily with an automatic filemove

 

- Using hidden folders may not be recommended so you can look at using one of the unmenu scripts to also exclude folders starting with underscore (_) from the file move.

You should then rename your plex folder to _plex.

 

Script description: " Exclude directories beginning with "_" from being moved from the cache drive."

Everything works fine but getting metadata? Sounds like an issue with the library. You may want to check file permissions. Perhaps when you moved the library to your cache, root become the owner. You can do a "chown -R unraid-plex:users <path to library folder>" and see if that fixes your issue. Or, while not ideal, you can always just whack your library and start fresh... If it worked before, something got mucked up along the way.

 

Everything appears to be working fine, I am streaming a blu ray rip right now that I cannot collect metadata on. All the metadata from before the move is intact but if I remove a file with metadata and re add it no metadata is able to be found.

 

I tried the chown -R unraid-plex:users /mnt/cache/.data/plex but it didn't seem to fix the problem. I have no problem whacking my library if necessary, is the best way to do this just "rm -r /mnt/cache/.data/plex" or is there something else I should be deleting? Just want to make sure I don't screw anything else up. See below for directory strucures.

 

 

Like kaiguy said, it might be permissions.

 

- Also, how did you point the plugin to the new cache drive? /mnt/cache ?

 

- Are you using hidden folders? (e.g. .plex?) If not, your files may be moved to the array daily with an automatic filemove

 

- Using hidden folders may not be recommended so you can look at using one of the unmenu scripts to also exclude folders starting with underscore (_) from the file move.

You should then rename your plex folder to _plex.

 

Script description: " Exclude directories beginning with "_" from being moved from the cache drive."

 

My directories look as follows on the Plex page in the settings tab

Install Directory  /usr/local/plex

Library Directory  /mnt/cache/.data/plex

Temp Directory    /tmp

Run As User        unraid-plex

Use Soft Links is checked

 

I am under the impression that directories beginning with "." are hidden so I should change .data to _data and then use unmenu to to exclude "_" directories from the mover script?

 

Thanks again for your help!

Hi guys

 

Want to add this to my VM unraid.

 

I have unraid booting from /mnt/vmdk isntead of the usb drive (faster boot), and my preference is to use this drive as the location for all packages.  Am i flawed in this thinking?

 

What changes should be made to the plugin script to accomodate this change that I want to make.

 

Thanks

 

Mick

off topic, but what solution did you move to?

 

FWIW the process growth/crashes I was seeing were not in the main Plex process itself, but in the transcoding process.  Perhaps completely unrelated to the issue mostly discussed here, but the end result is that Plex lost not only all GAF (Girlfriend Approval Factor) ... frankly most of mine.  The last thing I want to deal with when sitting down with fresh popcorn and the lights dimmed is running over to my PC to stop/start Plex and kill any runaway transcode processes.

 

I've moved on to other solutions, but I am keeping my eye on Plex for future use.

Hi,

 

has somebody an idea how to get this working with the S3 sleep script? If I run the PMS my cache drive never spins down and the server is unable to go to sleep.

 

My PMS config:

install dir: /usr/local/plex

library dir: /mnt/user/plexlib/library

temp dir: /mnt/user/plexlib/tmp

 

plexlib is a cache disk only user share.

 

Thanks

pan!c

 

Hi,

 

has somebody an idea how to get this working with the S3 sleep script? If I run the PMS my cache drive never spins down and the server is unable to go to sleep.

 

My PMS config:

install dir: /usr/local/plex

library dir: /mnt/user/plexlib/library

temp dir: /mnt/user/plexlib/tmp

 

plexlib is a cache disk only user share.

 

Thanks

pan!c

The library and tmp directory are probably being scanned by Plex regularly.  Unless you can stop plex before invoking the sleep you are probably out of luck.

curious whether anyone is using plex to stream to remote clients (http://wiki.plexapp.com/index.php/ConfigurePortForward)? I'm sure i've seen talk about unraid not really being designed to be open to the internet... but still, being able to stream stuff during my lunchbreak at work would be pretty cool!