I started trying to sort out a transcoding issue. Basics:

• I have Plex Pass
• Intel® Xeon® CPU E3-1220 v3 @ 3.10GHz
• NVIDIA Quadro P4000

transcoding fails and I find this in the logs:
 

Jul 14, 2023 07:27:26.000 [22714707979064] ERROR - [Req#18400/Transcode/bcc643af-2bf9-44d9-aa3a-acd90a145758/1c99bc4a-7c0a-4946-9901-dde19247fa87] [eac3_eae @ 0x14b17a841040] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-a106105e-84f9-4e1e-ba05-48254908b13e/EasyAudioEncoder/Convert to WAV (to 8ch or less)/bcc643af-2bf9-44d9-aa3a-acd90a145758_16882-0-2282.wav'
Jul 14, 2023 07:27:26.000 [22714679372600] ERROR - [Req#183ff/Transcode/978fbdf6-addd-40d2-aefc-b6b491ff9e44/7cbad210-1e84-4939-9d01-0196d1ffb394] [eac3_eae @ 0x14ec12450600] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-a106105e-84f9-4e1e-ba05-48254908b13e/EasyAudioEncoder/Convert to WAV (to 8ch or less)/978fbdf6-addd-40d2-aefc-b6b491ff9e44_20019-0-1712.wav'

 

I've tried restarting the container. Updating NVIDA drivers and checking other variables. All appear good. Everything else works fine.

Thoughts?

 

Actually, I think I solved it with this: DELETE CODEC FOLDER

Edited July 14, 20232 yr by The_Holocron
UPDATE

So twice now in 7 days, I've had my unraid server suddenly lock up (where the GUI takes 1-2 minutes to load a page, and I can see the CPU and RAM usage are at 100%). Both times it has happened around ~6am or ~7am in the morning, and when I started turning off docker containers one at a time - it was the Plex Media Server docker (repo: plexinc/pms-docker) that allowed my CPU to drop back to ~20% when I stopped the docker container. That makes me think that the Plex container is the culprit of my lockups.

 

Does anything stand out to anyone in my attached diagnostics file? I do have my Plex transcodes sent to /dev/smh - not sure if that is related? Another theory is that some recurring task in Plex (maybe end credit detection?) is consuming all of my CPU or something in that 6am or 7am window? Regardless, I'm curious to see if those better than I at log-reading might see anything of note.

 

Appreciate any help and insight you can give!

magicserver-diagnostics-20230805-0838.zip

Did I do something wrong ? I set up the plex server using the official App. I was able to add the movies directory. No problem at all. I though I would try to add my 'music' share. Every time I try the Docker is orphaned. I have to reinstall on the App/Docker reinstall page.  I included some pictures. Maybe I set it up wrong from the start?  I simulated how I was adding the music share without adding this time just to show in the picture. Any Ideas??

@fschris

You are mounting two different source paths to the same destination path /data which can't work:

 

My suggestion:

Set source and destination to the same path:

1.) /mnt/user/music > /mnt/user/music

2.) /mnt/user/Media > /mnt/user/Media

 

By that you will find the content inside of the Plex Container in the same paths as you know them from unRAID.

thanks just to clarify is this what you are suggesting:

3 hours ago, fschris said:

what you are suggesting:

Yes. That's how I'm doing it with all containers as long they don't require specific paths, but Plex is absolutely flexible regarding this. Plex even allows changing paths of already indexed files and automatically updates its database if the location of a file changes.

Thanks mgutt - I will give it a go. Appreciates your feedback.

This was the fix - thank you!  I still don't understand the Host container vs path container relationship!

Hi,

 

In the last two weeks, Plex has stopped working for some files. I think its a transcoding issue.

 

I get these errors:

 

[Req#5e2b/Transcode/r7pw9nhy59asswlpgjm37s5q] IsFileWritable: failed to create file '"/transcode/c3a08e2f-6ae3-43cd-ad23-a246f5f31505"'

 

I have attached my settings. I wonder if this is to do with the Host Path 2 I have set up. I thought this was to make it use ram to transcode but perhaps I have done something wrong here.

 

The extra parameters I have are:

--runtime=nvidia --no-healthcheck

 

I did get an issue with the nvidia driver when I updated to lasted Unraid OS but this seems to be resolved.

 

Any help would be greatly appreciated!

As Plex is writing to /tmp and this dir is in unRAID a RAM Disk, you RAM become probably full?!

 

Another possibility is that /tmp/transcode has wrong permissions. That's why I added this to my /boot/config/go file:

 

mkdir -m 777 /tmp/transcode

 

 

Hi all, can anyone help me with a permissions issue I appear to have for Plex-created Optimised versions?

 

I want to move all of my optimised versions of films out of their individual folders and into a different library to enable easier selection when remotely streaming.

 

The problem I have is that it appears that when Plex creates an optimised version that it permissions the "Plex Versions" folder as 755:

 

 

This all works just fine inside of Plex, but when I'm trying to move the optimised versions via SMB on my Windows PC I don't have permission to the files to modify them as the users group only has Read and Execute permissions

 

The Docker container is configured to use PUID 99 and PGID 100 which I believe is correct, but there doesn't seem to be an option to specify the UMASK for created files.

 

Is there a permanent fix for this?

 

I know I could just CHMOD all of the folders, but the next time a file is written by Plex it'll be wrong again  

 

Thanks!

Edited September 7, 20232 yr by CodeThief
Wrong umask :)

8 minutes ago, CodeThief said:

I know I could just CHMOD all of the folders, but the next time a file is written by Plex it'll be wrong again

But this would be the easiest fix. Making a script with "chmod -r" which gets executed on daily basis. The different solution is to create a custom SMB share which forces the user nobody.

 

Here is an example:

[plex-optimized]
  path = /mnt/user/plex/bla
  valid users = john
  read list =
  write list = john
  force user = nobody

 

Wow, thanks for the super-quick reply!

 

I guess this is potentially a one-time problem in reality, as I could just CHMOD everything, move the files to the new share and then I won't need to move them again so it's irrelevant what they're permissioned as...

 

Thanks

Greetings... I had an power outage yesterday, and realized that during some troubleshooting, I had disabled my battery backup on Unraid and forgotten to enable it again... Needless to say, my Unraid server powered off, and now I'm having issues with Plex.

 

I'm using the Official Plex Docker Image, and although I'm able to start Plex, I'm getting the following message in the logs.

 

/config/Library/Application Support/Plex Media Server/Preferences.xml:1.1: Document is empty

^
Failed to load preferences at /config/Library/Application Support/Plex Media Server/Preferences.xml
Failed to load preferences at /config/Library/Application Support/Plex Media Server/Preferences.xml
/config/Library/Application Support/Plex Media Server/Preferences.xml:1.1: Document is empty

^
Failed to load preferences at /config/Library/Application Support/Plex Media Server/Preferences.xml
Failed to load preferences at /config/Library/Application Support/Plex Media Server/Preferences.xml

 

I've googled around, and it seems like some of my options are deleting or renaming the Preferences.xml so it gets rebuilt. Is that the best course of action?

42 minutes ago, markc99 said:

I've googled around, and it seems like some of my options are deleting or renaming the Preferences.xml so it gets rebuilt. Is that the best course of action?

Maybe. But main question: Where is your backup?!

11 minutes ago, mgutt said:

Maybe. But main question: Where is your backup?!

So, I have Unraid USB stick backups... Is that what you are referring to, or Plex container specifically?

Edited September 8, 20232 yr by markc99

Hi, Folks

 

I got QNAP ts464 with Celeron 5095 and Unraid 6.11.5. I have Plex docker installed (version 4.108.0) and joined the Plex premium membership. I enable the hardware acceleration in Plex. However when I stop watch video via Plex, I found it is still using CPU instead of GPU to transcode. 

 

I read the log from Unraid and found the following error telling me cannot can find the hw device. 

 Sep 08, 2023 17:13:45.420 [23103432821560] Debug — [Req#1659/Transcode] Codecs: testing h264_nvenc (encoder)
 Sep 08, 2023 17:13:45.420 [23103432821560] Debug — [Req#1659/Transcode] Codecs: hardware transcoding: testing  API nvenc for device '' ()
 Sep 08, 2023 17:13:45.420 [23103432821560] Error — [Req#1659/Transcode] [FFMPEG] - Cannot load libcuda.so.1

 Sep 08, 2023 17:13:45.420 [23103432821560] Error — [Req#1659/Transcode] [FFMPEG] - Could not dynamically load CUDA

 Sep 08, 2023 17:13:45.420 [23103432821560] Debug — [Req#1659/Transcode] Codecs: hardware transcoding: opening hw device failed - probably not supported by this system, error: Operation not permitted
 Sep 08, 2023 17:13:45.420 [23103432821560] Debug — [Req#1659/Transcode] Could not create hardware context for h264_nvenc
 Sep 08, 2023 17:13:45.421 [23103432821560] Debug — [Req#1659/Transcode] Codecs: testing hevc (decoder) with hwdevice vaapi
 Sep 08, 2023 17:13:45.422 [23103432821560] Debug — [Req#1659/Transcode] Codecs: hardware transcoding: testing API vaapi for device '' ()
 Sep 08, 2023 17:13:45.422 [23103432821560] Debug — [Req#1659/Transcode] Codecs: hardware transcoding: opening hw device failed - probably not supported by this system, error: Generic error in an external library
 Sep 08, 2023 17:13:45.422 [23103432821560] Debug — [Req#1659/Transcode] Could not create hardware context for hevc
 Sep 08, 2023 17:13:45.422 [23103432821560] Debug — [Req#1659/Transcode] Codecs: testing hevc (decoder) with hwdevice nvdec
 Sep 08, 2023 17:13:45.423 [23103432821560] Debug — [Req#1659/Transcode] Codecs: hardware transcoding: testing API nvdec for device '' ()
 Sep 08, 2023 17:13:45.423 [23103432821560] Error — [Req#1659/Transcode] [FFMPEG] - Cannot load libcuda.so.1

 Sep 08, 2023 17:13:45.423 [23103432821560] Error — [Req#1659/Transcode] [FFMPEG] - Could not dynamically load CUDA

 Sep 08, 2023 17:13:45.423 [23103432821560] Debug — [Req#1659/Transcode] Codecs: hardware transcoding: opening hw device failed - probably not supported by this system, error: Operation not permitted
 Sep 08, 2023 17:13:45.423 [23103432821560] Debug — [Req#1659/Transcode] Could not create hardware context for hevc

 

I checked the device and found the card should be renderD128. 

root@Eddie-NAS3:/dev/dri# ls -al
total 0
drwxrwxrwx  3 root root       100 Jun 27 04:51 ./
drwxr-xr-x 16 root root      3360 Jun 27 04:51 ../
drwxrwxrwx  2 root root        80 Jun 27 04:51 by-path/
crwxrwxrwx  1 root video 226,   0 Jun 27 04:51 card0
crwxrwxrwx  1 root video 226, 128 Jun 27 04:51 renderD128

 

I edit the Preferences.xml by adding the following line and saved it. However it still has the same problem and same error in log. 

HardwareDevicePath="/dev/dri/renderD128"

 

Not sure what I've missed and also not sure if this thread is the right one to ask this question. Can you give me a hint what could be the next step to fix it? Thanks a lot. 

4 hours ago, eddie_bear said:

I edit the Preferences.xml

This step is not necessary. You only need to pass /dev/dri through the container settings

 

I'd say the CPU is not fully supported by the Container or unRAID Linux Kernel:

https://www.reddit.com/r/PleX/comments/r1g3wl/plex_hardware_transcoding_issue_with_pms_on_intel/

 

14 hours ago, markc99 said:

or Plex container specifically?

You MUST create backups of the whole appdata share. And yes this contains Plex, too.

 

 

4 hours ago, mgutt said:

This step is not necessary. You only need to pass /dev/dri through the container settings

 

I'd say the CPU is not fully supported by the Container or unRAID Linux Kernel:

https://www.reddit.com/r/PleX/comments/r1g3wl/plex_hardware_transcoding_issue_with_pms_on_intel/

 

 

Hi, 

Thanks for the hint. Just wondering which part of the container settings to pass /dev/dri through? I cannot see any hint in the container settings to do that. 

In the thread you provided, I can see 4K is not fully capable but 1080P and 720P should be supported to transcode. However I've tested 1080P and 720P both, and did not see that happen. 

1 hour ago, eddie_bear said:

Just wondering which part of the container settings to pass /dev/dri through? I cannot see any hint in the container settings to do that. 

Install the Intel GPU Top Plugin.

 

Then edit the Plex Container, scroll to the bottom, Add device and set both path to /dev/dri

 

I thought you already did those steps. 😅

6 hours ago, mgutt said:

This step is not necessary. You only need to pass /dev/dri through the container settings

 

I'd say the CPU is not fully supported by the Container or unRAID Linux Kernel:

https://www.reddit.com/r/PleX/comments/r1g3wl/plex_hardware_transcoding_issue_with_pms_on_intel/

Thanks for the hint. I did some research and I can pass /dev/dri through the container settings. Now in the Plex transcode setting, I can see the hardware transcode device is JasperLake [UHD Graphics]. That means the iGPU is detected by the Plex. However when I play the movie again, I still don't see the [hw] shown in playback view as below. 

Checked CPU usage it does not seem the Plex use iGPU to transcode the movie.   

 

Do you know why with iGPU already detected by Plex, it still does not do the hardware transcode? 

 

 

Thanks.  

1 hour ago, eddie_bear said:

Do you know why with iGPU already detected by Plex, it still does not do the hardware transcode? 

Plex pass?

https://support.plex.tv/articles/115002178853-using-hardware-accelerated-streaming/

I have hardware transcoding working well in Plex Media Server on two Unraid systems:

• Build One: AMD Ryzen 1700X CPU and NVIDIA 1050 Ti hardware for video hardware transcoding. Unraid's NVIDIA Driver plug-in was straightforward to install and configure (make sure it is configured with ID matching the NVIDIA card's unique ID).
• Build Two: Intel Core i7-7700 CPU, leveraging the Core i7's GPU (Intel Graphics HD 630) for video hardware transcoding (Intel Quick Sync Video).

 

Both systems use the "official" docker container provided by Plex (sourced via Unraid Apps). Be sure to configure docker container to reference the Plex Pass version container (see next paragraph). Both Unraid builds have "GPU Statistics" plug-in installed. "Intel GPU TOP" plugin installed on Intel build (helpful to reveal/make-available the GPU to video applications).

 

I discovered (by error) it is essential to not overlook including the "plexpass" bit of the repository name ("plexinc/pms-docker:plexpass") AND ensure the directory used for transcoding has read/write permissions.

 

Video hardware transcoding on both systems has delivered excellent results, been been easy to monitor/check and maintain across several Unraid updates/upgrades and Plex Media Server updates/upgrades.

 

I hope this helps your installation and troubleshooting.

 

 

1 hour ago, Maxrad said:

I discovered (by error) it is essential to not overlook including the "plexpass" bit of the repository name ("plexinc/pms-docker:plexpass")

😛lexpass is a beta / early access channel. Which means you get the most recent features, yes, but all the most recent bugs, too. Here is an example of a newest beta version, which is only available through the :beta or :plexpass tag:

 

 

Ah, yes. Good point. Thanks for the reminder.