Dockerfiles

This repository hosts docker build files of software stacks and services, designed to enable Open Visual Cloud prioritized use cases such as media delivery, media analytics, cloud gaming and cloud graphics, and immersive media.

Software Stack Images:

The software stack images provide ready to use software stacks for application deployment. You can call the software executables or link with the software libraries.

Media Delivery

Image	Description
ffmpeg	Image optimized for media creation and delivery. Included codecs: aac, mp3, opus, ogg, vorbis, x264, x265, vp8/9, av1 and SVT-HEVC. The GPU images are accelerated with vaapi and qsv. See ffmpeg.md for details.
gst	Image optimized for media creation and delivery. Included the base, good, bad, ugly and libav set of plugins. The GPU images are accelerated with vaapi. See gst.md for details.
nginx	Image optimized for web hosting and caching. Based on FFmpeg, included NGINX the web server and RTMP the RTMP, DASH and HLS streaming module. See nginx.md for details.
svt	Image with SVT (Scalable Video Technology) Encoder and decoders. Ready to use SVT apps to try AV1, HEVC, VP9 transcoders. See svt.md for details.

Media Analytics

Image	Description
ffmpeg	Image optimized for media analytics. Included what are in the FFmpeg image. Inferencing engine and tracking plugins to be included. See ffmpeg.md for details.
gst	Image optimized for media analytics. Included what are in the GStreamer image. Inferencing engine and tracking plugins to be included. See gst.md for details.

Cloud Gaming and Graphics

Image	Description
ospray	Image optimized for intel ray tracing api. Based on embree, included ospray Ray Tracing engine and examples. See ospray.md for details.
ospray-mpi	Image optimized for intel ray tracing api. Based on embree, included ospray Ray Tracing engine with examples(which require OpenImageIO) and multi-host connection via MPI. See ospray-mpi.md for details.

Development Images:

The development images enable application compilation, debugging (with the debugging, profiling tools) and optimization (with the optimization tools.) You can compile C++ applications with these images and then copy the applications to the corresponding deployment images for deployment.

Image	Description
dev	Image containing the FFmpeg and GStreamer C++ development files. The OpenViNO model optimizer is to be included. See ffmpeg.md and gst.md for details.

Service Images:

The service images provides ready to use services. See their image descriptions for exposed service interfaces.

Image	Description
owt	Image optimized for video conferencing service based on the WebRTC technology and Open WebRTC Toolkit (OWT). Included conferencing modes: 1:N, N:N with video and audio processing nodes. see owt.md for details.

Support Matrix:

The project supports the following platforms and OS'es:

Supported Platforms	Supported OS'es
Xeon	Ubuntu 16.04 LTS, Ubuntu 18.04 LTS, CentOS 7.4, 7.5, and 7.6
Xeon E3	Ubuntu 16.04 LTS, Ubuntu 18.04 LTS, CentOS 7.4, 7.5, and 7.6
VCA2	Ubuntu 16.04 LTS, Ubuntu 18.04 LTS, CentOS 7.4, 7.5, and 7.6

Please see development and test statuses for the latest development statuses.

Host Platform Setup:

• Update kernels and firmwares: Please see each platform folder README for instructions.
• Follow the instructions to setup host date and time.
• Follow the instructions to install docker.ce.
• If you are behind a firewall, setup proxy as follows:

(1) sudo mkdir -p /etc/systemd/system/docker.service.d    
(2) printf "[Service]\nEnvironment=\"HTTPS_PROXY=$https_proxy\" \"NO_PROXY=$no_proxy\"\n" | sudo tee /etc/systemd/system/docker.service.d/proxy.conf    
(3) sudo systemctl daemon-reload     
(4) sudo systemctl restart docker     

Build docker image:

(1) mkdir build    
(2) cd build     
(3) cmake ..    
(4) cd Xeon/ubuntu-16.04/media/ffmpeg # please build your specific <_platform_>/<_OS_>/<_usage_>/<_image_> only as a full build takes a long time.     
(5) make # build on the target processor for best performance.    
(6) ctest   

Run shell:

Xeon/ubuntu-16.04/media/ffmpeg/shell.sh #<_platform_>/<_OS_>/<_usage_>/<_image_>

Customize:

You can modify any Dockerfile.m4 template for customization.
For example, uncomment #include(transform360.m4) in Xeon/ubuntu-16.04/media/ffmpeg/Dockerfile.m4 to add essential 360 video transformation in the FFmpeg build.
After modification, please rerun cmake and make.

Use alternative repo:

Certain source repo might be blocked in certain network. You can specify alternative repos before the build command as follows:

export AOM_REPO=...       
export VPX_REPO=...     
make

For a list of all REPOs and their versions, run the following command:

grep -E '_(REPO|VER)=' template/*.m4         

Use Dockerfile in other project:

It is recommended that you copy the Dockerfile(s) of your platform, OS and image directly into your other project. The following shell scripts show how to sync (if needed) and build the NGINX Dockerfile (and its dependency FFmpeg):

update.sh:

DOCKER_REPO=${DOCKER_REPO="https://raw.githubusercontent.com/OpenVisualCloud/Dockerfiles/master/Xeon/ubuntu-18.04/media"}    
(echo "# xeon-ubuntu1804-media-ffmpeg" && curl ${DOCKER_REPO}/ffmpeg/Dockerfile) > Dockerfile.2    
(echo "# xeon-ubuntu1804-media-nginx" && curl ${DOCKER_REPO}/nginx+rtmp/Dockerfile) > Dockerfile.1    

build.sh:

for dep in .2 .1; do   
    image=$(grep -m1 '#' "Dockerfile$dep" | cut -d' ' -f2)   
    sudo docker build --network=host --file="Dockerfile$dep" -t "$image:latest" . $(env | grep -E '_(proxy)=' | sed 's/^/--build-arg /')   
done