-
Notifications
You must be signed in to change notification settings - Fork 639
Docker GPU Installation
Asharam Seervi edited this page Apr 10, 2023
·
6 revisions
Attention: We have migrated our documentation to our new platform, Ant Media Resources. Please follow <a href="https://antmedia.io/docs/guides/advanced-usage/Using-NVIDIA-Hardware-based-Encoder-on-Docker?utm_source=github&utm_medium=wiki" for the latest and up-to-date documentation.
On host(Ubuntu 16.04, 18.04 and 20.04)
-
Install CUDA Drivers
-
Install docker-ce according to the link - https://docs.docker.com/install/
sudo apt-get install apt-transport-https ca-certificates curl gnupg-agent software-properties-common
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"
sudo apt-get update
sudo apt-get install docker-ce docker-ce-cli containerd.io
- Add Repos for nvidia-docker2
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | \
sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update
- Install nvidia-docker2 for Ubuntu 16.04, 18.04 and 20.04
sudo apt-get install -y nvidia-docker2
sudo pkill -SIGHUP dockerd
Start a docker container with following command
- Ubuntu 16.04
sudo docker run --runtime=nvidia \
--privileged --network host --name cuda-docker2 \
-e NVIDIA_VISIBLE_DEVICES=all -e NVIDIA_DRIVER_CAPABILITIES=compute,utility,video \
-it nvidia/cuda:10.0-runtime-ubuntu16.04
- Ubuntu 18.04
sudo docker run --runtime=nvidia \
--privileged --network host --name cuda-docker2 \
-e NVIDIA_VISIBLE_DEVICES=all -e NVIDIA_DRIVER_CAPABILITIES=compute,utility,video \
-it nvidia/cuda:10.0-runtime-ubuntu18.04
- Ubuntu 20.04
sudo docker run --runtime=nvidia \
--privileged --network host --name cuda-docker2 \
-e NVIDIA_VISIBLE_DEVICES=all -e NVIDIA_DRIVER_CAPABILITIES=compute,utility,video \
-it nvidia/cuda:11.0-runtime-ubuntu20.04
In this docker container, you can install Ant-Media-Server Enterprise edition. It automatically uses hardware encoder.
Alternatively, you can use Ant Media Server Docker file and just change the the line FROM ubuntu:20.04
to `FROM nvidia/cuda:11.0-runtime-ubuntu20.04' and run the container with the parameters above
If you have any questions or comments, just send an email to contact at antmedia dot io or fill the contact form.
- Introduction
- Quick Start
- Installation
- Publishing Live Streams
- Playing Live Streams
- Conference Call
- Peer to Peer Call
- Adaptive Bitrate(Multi-Bitrate) Streaming
- Data Channel
- Video on Demand Streaming
- Simulcasting to Social Media Channels
- Clustering & Scaling
- Monitor Ant Media Servers with Apache Kafka and Grafana
- WebRTC SDKs
- Security
- Integration with your Project
- Advanced
- WebRTC Load Testing
- TURN Servers
- AWS Wavelength Deployment
- Multi-Tenancy Support
- Monitor Ant Media Server with Datadog
- Clustering in Alibaba
- Playlist
- Kubernetes
- Time based One Time Password
- Kubernetes Autoscaling
- Kubernetes Ingress
- How to Install Ant Media Server on EKS
- Release Tests
- Spaceport Volumetric Video
- WebRTC Viewers Info
- Webhook Authentication for Publishing Streams
- Recording Streams
- How to Update Ant Media Server with Cloudformation
- How to Install Ant Media Server on GKE
- Ant Media Server on Docker Swarm
- Developer Quick Start
- Recording HLS, MP4 and how to recover
- Re-streaming update
- Git Branching
- UML Diagrams