Server Version#: 1.42.2.10156 (unraid, official docker image, newly installed and setup)
Hello
I recently switched to the official Plex Docker image to solve another issue on unraid. I have noticed a new issue though! 
The server is constantly getting stuck detecting new intros. A restart solves it but then as soon as new media is detected it gets stuck.
Any ideas please?
Is it always gettin stuck on the same items?
no. it’s whatever the latest media is it needs to scan. It’s got stuck on everything new each time after a restart.
Likely going to need logs so i can show it to devs
Jan 27, 2026 16:42:53.000 [22746522446648] ERROR - [Req#15235/Transcode/84ed05b2-3458-4395-93eb-ab508ee24e85/e32355bf-cd47-4c1d-b529-ca64378e2d8c] [eac3_eae @ 0x14855f303e00] EAE timeout! EAE not running, or wrong folder? Could not read '/tmp/pms-44d64533-6cab-401d-a6e9-c2f9615e9e40/EasyAudioEncoder/Convert to WAV (to 8ch or less)/84ed05b2-3458-4395-93eb-ab508ee24e85_40427-0-0.wav'
Jan 27, 2026 16:38:08.002 [22746539297592] ERROR - [Req#14363/Transcode/15f4dfe3-898a-4168-b8bf-191a3695c62d/a3062b24-def6-4d6e-bbde-e75324b09589] Error while decoding stream #0:1: No space left on device
Your /tmp path is either out of space, becoming inaccessible or you have reached the inotify limit for the system.
Check that this path is writable and has enough space then restart PMS and try again.
Thank you, I have my transcode directory mapped to /transcode which maps to /dev/shm/ on the host - which I believe means it uses RAM. This is to reduce wear on disk drives. I have monitored RAM usage and it never gets anywhere close to full when the issue happens. I’ve had a google around and some people recommend creating a temp folder like this on ram…
mkdir -p /tmp/PlexRamScratch && mount -t tmpfs -o size=8g tmpfs /tmp/PlexRamScratch
So I’ve now mapped my transcode folder to /tmp/PlexRamScratch and set the above command to run on server startup.
I’m not familiar with inotify, is this something I should research and consider?
Do you think this is a good idea. I’ll try it see how it goes.
There is zero practical benefit using ram for this and only causes problems like this.
Just map it to any physical drive with sufficient space.
ok thank you. I have updated transcode directory to a folder on an SSD. I’ll see how it goes. Many guides seem to recommend RAM, which does make sense in theory, but if it causes problems or isn’t stable like you say then whats the point
I really wish people online would stop recommending this practice 
I’m not sure what you think to this but now that there is an official server media docker template on Unraid you could create a quick FAQ with a few of these recommendations in. Most people (myself included) have to instantly look up help to get the intel hardware decoding configured correctly - needing to map /dev/dri/ (or whatever hardware you have), and then learning how transcoding works quickly follows. The recommendations could even be built in to the docker template help bit when setting it up. Or if there is this help already I missed it.
