linkwellfook 2 Posted July 8, 2025 Posted July 8, 2025 Hi, The docker version of Emby does not detct my RTX 3070 for h/w transcoding My setup Ubuntu 24.04 64 GB Ram Ryzen 9 CPU Running nvidia-smi +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 550.163.01 Driver Version: 550.163.01 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 3070 Off | 00000000:0D:00.0 On | N/A | | 45% 60C P2 49W / 220W | 362MiB / 8192MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 2449886 C /usr/lib/ffmpeg/7.0/bin/ffmpeg 175MiB | | 0 N/A N/A 2526651 C /usr/lib/ffmpeg/7.0/bin/ffmpeg 177MiB | +-----------------------------------------------------------------------------------------+ cat /sys/module/nvidia/version 550.163.01 cat /proc/driver/nvidia/version NVRM version: NVIDIA UNIX x86_64 Kernel Module 550.163.01 Tue Apr 8 12:41:17 UTC 2025 GCC version: gcc version 13.3.0 (Ubuntu 13.3.0-6ubuntu2~24.04) From within emby docker cat /sys/module/nvidia/version 550.163.01 cat /proc/driver/nvidia/version NVRM version: NVIDIA UNIX x86_64 Kernel Module 550.163.01 Tue Apr 8 12:41:17 UTC 2025 GCC version: gcc version 13.3.0 (Ubuntu 13.3.0-6ubuntu2~24.04) Files included Embyserver logs embyserver docker compose file embyserver.txt emby.yml
yocker 1247 Posted July 8, 2025 Posted July 8, 2025 Have you remembered the -runtime Nvidia ? I think that's how it's supposed to be
linkwellfook 2 Posted July 8, 2025 Author Posted July 8, 2025 Hi, Yeap. It's in the docker compose file. runtime: nvidia
Luke 42077 Posted July 8, 2025 Posted July 8, 2025 HI, are you able to install our native package directly on the host machine to see how that compares?
linkwellfook 2 Posted July 8, 2025 Author Posted July 8, 2025 Hi, Will do. I'm about to step out so it will be the following day before I get to do this. 1
linkwellfook 2 Posted July 9, 2025 Author Posted July 9, 2025 Hi, I've installed the native package onto the host machine (same machine hosts the docker version) and I see the hw encoders and decoders.
guunter 49 Posted July 9, 2025 Posted July 9, 2025 Btw you don't need these line in your compose if youre using an nvidia GPU. devices: - /dev/dri:/dev/dri #- /dev/dri/card0:/dev/dri/card0 - /dev/dri/renderD128:/dev/dri/renderD128
linkwellfook 2 Posted July 10, 2025 Author Posted July 10, 2025 Removing devices section did nothing. If anything it broke emby in docker [s6-init] making user provided files available at /var/run/s6/etc...exited 0. [s6-init] ensuring user provided files have correct perms...exited 0. [fix-attrs.d] applying ownership & permissions fixes... [fix-attrs.d] done. [cont-init.d] executing container initialization scripts... [cont-init.d] done. [services.d] starting services [services.d] done. /system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory [cont-finish.d] executing container finish scripts... [cont-finish.d] done. [s6-finish] waiting for services. [s6-finish] sending all processes the TERM signal. [s6-finish] sending all processes the KILL signal and exiting. Why do I get the feeling I should have adhered to the old mantera If its not broke don't fix it. when I run locate libpthread.so.0 from the host I get /snap/core20/2582/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core20/2582/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/core20/2599/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core20/2599/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/core22/1981/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core22/1981/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/core22/2010/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core22/2010/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/snapd/24718/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/snapd/24792/usr/lib/x86_64-linux-gnu/libpthread.so.0 /usr/lib/i386-linux-gnu/libpthread.so.0 /usr/lib/x86_64-linux-gnu/libpthread.so.0 /usr/lib32/libpthread.so.0 I can still launch native emby
guunter 49 Posted July 10, 2025 Posted July 10, 2025 5 hours ago, linkwellfook said: Removing devices section did nothing. If anything it broke emby in docker [s6-init] making user provided files available at /var/run/s6/etc...exited 0. [s6-init] ensuring user provided files have correct perms...exited 0. [fix-attrs.d] applying ownership & permissions fixes... [fix-attrs.d] done. [cont-init.d] executing container initialization scripts... [cont-init.d] done. [services.d] starting services [services.d] done. /system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory [cont-finish.d] executing container finish scripts... [cont-finish.d] done. [s6-finish] waiting for services. [s6-finish] sending all processes the TERM signal. [s6-finish] sending all processes the KILL signal and exiting. Why do I get the feeling I should have adhered to the old mantera If its not broke don't fix it. when I run locate libpthread.so.0 from the host I get /snap/core20/2582/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core20/2582/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/core20/2599/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core20/2599/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/core22/1981/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core22/1981/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/core22/2010/usr/lib/i386-linux-gnu/libpthread.so.0 /snap/core22/2010/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/snapd/24718/usr/lib/x86_64-linux-gnu/libpthread.so.0 /snap/snapd/24792/usr/lib/x86_64-linux-gnu/libpthread.so.0 /usr/lib/i386-linux-gnu/libpthread.so.0 /usr/lib/x86_64-linux-gnu/libpthread.so.0 /usr/lib32/libpthread.so.0 I can still launch native emby Did you install the nvidia toolkit? The runtime: nvidia does nothing without it.
linkwellfook 2 Posted July 10, 2025 Author Posted July 10, 2025 Yes, as shown when using the commands on the host: nvidia-smi cat /sys/module/nvidia/version cat /proc/driver/nvidia/version I've included the results when I created the post.
guunter 49 Posted July 10, 2025 Posted July 10, 2025 (edited) 26 minutes ago, linkwellfook said: Yes, as shown when using the commands on the host: nvidia-smi cat /sys/module/nvidia/version cat /proc/driver/nvidia/version I've included the results when I created the post. That is the nvidia driver not the toolkit. It's a separate install. https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html. If you do the command below do you get an output? which nvidia-container-toolkit Edited July 10, 2025 by guunter
linkwellfook 2 Posted July 10, 2025 Author Posted July 10, 2025 hi Results from which nvidia-container-toolkit from the host /usr/bin/nvidia-container-toolkit
guunter 49 Posted July 10, 2025 Posted July 10, 2025 17 minutes ago, linkwellfook said: hi Results from which nvidia-container-toolkit from the host /usr/bin/nvidia-container-toolkit Looks like you have everything in place. Alternatively did you try using linuxserver docker image? I use that with my emby. You should be able to just change the volumes and it'll start up or just point it to a new config volume for testing. services: emby: image: lscr.io/linuxserver/emby:latest container_name: emby runtime: nvidia environment: - PUID=1003 - PGID=1003 - TZ=America/Los_Angeles - NVIDIA_VISIBLE_DEVICES=all volumes: - /opt/appdata/emby:/config - /mnt/nas/media/:/mnt/nas/media - /dev/shm:/transcoding ports: - 8096:8096 - 8920:8920 #optional restart: unless-stopped
Neminem 1518 Posted July 10, 2025 Posted July 10, 2025 @guunterare you using RAM drive as your transcode drive ?
guunter 49 Posted July 10, 2025 Posted July 10, 2025 9 minutes ago, Neminem said: @guunterare you using RAM drive as your transcode drive ? Yes I am. 256gb of ram . No live tv. 1
Neminem 1518 Posted July 10, 2025 Posted July 10, 2025 (edited) Ok that's why . I tried once and that failed because of a run away live tv stream Someone watching le-man race. Only have 64GB RAM and shm taking 32GB but not enough in that use case Edited July 10, 2025 by Neminem
linkwellfook 2 Posted July 10, 2025 Author Posted July 10, 2025 Hi Here's the revised compose file services: emby: #image: ysx88/embyserver:latest image: lscr.io/linuxserver/emby:latest container_name: embyserver runtime: nvidia environment: - UID=998 # The UID to run emby as (default: 2) - GID=995 # The GID to run emby as (default 2) - GIDLIST=44,108 # A comma-separated list of additional GIDs to run emby as (default: 2) - NVIDIA_VISIBLE_DEVICES=all #- NVIDIA_DRIVER_CAPABILITIES=compute,utility,video #network_mode: "service:gluetun" volumes: - "$DOCKERDIRDATADIR/emby/programdata:/config" # Configuration directory - "$MEDIADIR:/media" # Media directory - "$MUSIC:/music" # Media directory - "/dev/shm:/transcoding" ports: - 8096:8096 # HTTP port - 8920:8920 # HTTPS port# shm_size: "6gb"# restart: on-failure I removed the lines - NVIDIA_DRIVER_CAPABILITIES=compute,utility,video deploy: resources: reservations: devices: - driver: nvidia count: 1 capabilities: [gpu] and devices: - /dev/dri:/dev/dri - /dev/dri/card1:/dev/dri/card1 - /dev/dri/renderD128:/dev/dri/renderD128 I still get **** adding /dev/dri/renderD128 to video group root with id 0 **** **** permissions for /dev/dri/card1 are good **** [custom-init] No custom files found, skipping... /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory Regards
guunter 49 Posted July 10, 2025 Posted July 10, 2025 (edited) 13 minutes ago, linkwellfook said: Hi Here's the revised compose file services: emby: #image: ysx88/embyserver:latest image: lscr.io/linuxserver/emby:latest container_name: embyserver runtime: nvidia environment: - UID=998 # The UID to run emby as (default: 2) - GID=995 # The GID to run emby as (default 2) - GIDLIST=44,108 # A comma-separated list of additional GIDs to run emby as (default: 2) - NVIDIA_VISIBLE_DEVICES=all #- NVIDIA_DRIVER_CAPABILITIES=compute,utility,video #network_mode: "service:gluetun" volumes: - "$DOCKERDIRDATADIR/emby/programdata:/config" # Configuration directory - "$MEDIADIR:/media" # Media directory - "$MUSIC:/music" # Media directory - "/dev/shm:/transcoding" ports: - 8096:8096 # HTTP port - 8920:8920 # HTTPS port# shm_size: "6gb"# restart: on-failure I removed the lines - NVIDIA_DRIVER_CAPABILITIES=compute,utility,video deploy: resources: reservations: devices: - driver: nvidia count: 1 capabilities: [gpu] and devices: - /dev/dri:/dev/dri - /dev/dri/card1:/dev/dri/card1 - /dev/dri/renderD128:/dev/dri/renderD128 I still get **** adding /dev/dri/renderD128 to video group root with id 0 **** **** permissions for /dev/dri/card1 are good **** [custom-init] No custom files found, skipping... /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory Regards I'm at a loss you seem to have it all set right. Did you install docker via apt or snap? Edited July 10, 2025 by guunter
linkwellfook 2 Posted July 10, 2025 Author Posted July 10, 2025 Hi A quick recap When I first opened this discussion emby was running however I could not use GPU transcoding. It works if I use the native app. No matter if I use my original compose modified or use a different image I get the above output. Regarding docker. It was installed using apt from dockers repo
guunter 49 Posted July 10, 2025 Posted July 10, 2025 (edited) 7 minutes ago, linkwellfook said: Hi A quick recap When I first opened this discussion emby was running however I could not use GPU transcoding. It works if I use the native app. No matter if I use my original compose modified or use a different image I get the above output. Regarding docker. It was installed using apt from dockers repo If you go to the emby docer config folder location on your host are the permissions set correctly? This error makes it seem like a permission issue . Just trying to throw out troubleshooting ideas. /app/emby/system/EmbyServer: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory Edited July 10, 2025 by guunter
linkwellfook 2 Posted July 10, 2025 Author Posted July 10, 2025 Hi, yeah the permissions are correct. My user name has rw access to emby config folder My user name is a member of docker and emby group. Please note I will be away for a couple of days with limited internet access.
guunter 49 Posted July 11, 2025 Posted July 11, 2025 Is your emby docker setup all the way? If its a brand new install i'd just blast the config folder and start new with the linuxserver.io image.The Linuxserver and official image aren't the same. Alternatively you can make a copy of your standard emby and point your docker config volume to the copy. That way you don't have to make a whole new server from scratch. Just be sure to turn off the standard install server first.
linkwellfook 2 Posted July 21, 2025 Author Posted July 21, 2025 Hi guunter, I'll nuke the config and try linuxserver image.
guunter 49 Posted July 21, 2025 Posted July 21, 2025 12 hours ago, linkwellfook said: Hi guunter, I'll nuke the config and try linuxserver image. Sounds good. I hope it works out for you! Linuxserver includes extra packages in their image to help with gpu compatibility.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now