/nvidia-container-runtime

NVIDIA container runtime

Primary LanguageGoBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

nvidia-container-runtime

GitHub license Package repository

Warning: This project is based on an alpha release (libnvidia-container), it is not intended to be used in production systems.

A modified version of runc adding a custom pre-start hook to all containers.
If environment variable NVIDIA_VISIBLE_DEVICES is set in the OCI spec, the hook will configure GPU access for the container by leveraging nvidia-container-cli from project libnvidia-container.

Usage example

# Setup a rootfs based on Ubuntu 16.04
cd $(mktemp -d) && mkdir rootfs
curl -sS http://cdimage.ubuntu.com/ubuntu-base/releases/16.04/release/ubuntu-base-16.04-core-amd64.tar.gz | tar --exclude 'dev/*' -C rootfs -xz

# Create an OCI runtime spec
nvidia-container-runtime spec
sed -i 's;"sh";"nvidia-smi";' config.json
sed -i 's;\("TERM=xterm"\);\1, "NVIDIA_VISIBLE_DEVICES=0";' config.json

# Run the container
sudo nvidia-container-runtime run nvidia_smi

Installation

Ubuntu distributions

  1. Install the repository for your distribution by following the instructions here.
  2. Install the nvidia-container-runtime package:
sudo apt-get install nvidia-container-runtime

CentOS distributions

  1. Install the repository for your distribution by following the instructions here.
  2. Install the nvidia-container-runtime package:
sudo yum install nvidia-container-runtime

Docker Engine setup

To register the nvidia runtime, use the method below that is best suited to your environment.
You might need to merge the new argument with your existing configuration.

Systemd drop-in file

sudo mkdir -p /etc/systemd/system/docker.service.d
sudo tee /etc/systemd/system/docker.service.d/override.conf <<EOF
[Service]
ExecStart=
ExecStart=/usr/bin/dockerd --host=fd:// --add-runtime=nvidia=/usr/bin/nvidia-container-runtime
EOF
sudo systemctl daemon-reload
sudo systemctl restart docker

Daemon configuration file

sudo tee /etc/docker/daemon.json <<EOF
{
    "runtimes": {
        "nvidia": {
            "path": "/usr/bin/nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}
EOF
sudo pkill -SIGHUP dockerd

Command line

sudo dockerd --add-runtime=nvidia=/usr/bin/nvidia-container-runtime [...]

Environment variables (OCI spec)

Each environment variable maps to an command-line argument for nvidia-container-cli from libnvidia-container.
These variables are already set in our official CUDA images.

NVIDIA_VISIBLE_DEVICES

This variable controls which GPUs will be made accessible inside the container.

Possible values

  • 0,1,2, GPU-fef8089b …: a comma-separated list of GPU UUID(s) or index(es),
  • all: all GPUs will be accessible, this is the default value in our container images,
  • none: no GPU will be accessible, but driver capabilities will be enabled.
  • empty: nvidia-container-runtime will have the same behavior as runc.

NVIDIA_DRIVER_CAPABILITIES

This option controls which driver libraries/binaries will be mounted inside the container.

Possible values

  • compute,video, graphics,utility …: a comma-separated list of driver features the container needs,
  • all: enable all available driver capabilities.
  • empty: use default driver capabilities, determined by nvidia-container-cli.

Supported driver capabilities

  • compute: required for CUDA and OpenCL applications,
  • compat32: required for running 32-bit applications,
  • graphics: required for running OpenGL and Vulkan applications,
  • utility: required for using nvidia-smi and NVML,
  • video: required for using the Video Codec SDK.

NVIDIA_REQUIRE_*

A logical expression to define constraints on the configurations supported by the container.

Supported constraints

  • cuda: constraint on the CUDA driver version,
  • driver: constraint on the driver version,
  • arch: constraint on the compute architectures of the selected GPUs.

Expressions

Multiple constraints can be expressed in a single environment variable: space-separated constraints are ORed, comma-separated constraints are ANDed.
Multiple environment variables of the form NVIDIA_REQUIRE_* are ANDed together.

NVIDIA_DISABLE_REQUIRE

Single switch to disable all the constraints of the form NVIDIA_REQUIRE_*.

NVIDIA_REQUIRE_CUDA

The version of the CUDA toolkit used by the container. It is an instance of the generic NVIDIA_REQUIRE_* case and it is set by official CUDA images. If the version of the NVIDIA driver is insufficient to run this version of CUDA, the container will not be started.

Possible values

  • cuda>=7.5, cuda>=8.0, cuda>=9.0 …: any valid CUDA version in the form major.minor.

CUDA_VERSION

Similar to NVIDIA_REQUIRE_CUDA, for legacy CUDA images.
In addition, if NVIDIA_REQUIRE_CUDA is not set, NVIDIA_VISIBLE_DEVICES and NVIDIA_DRIVER_CAPABILITIES will default to all.

Copyright and License

This project is released under the BSD 3-clause license.