mirror of https://github.com/NVIDIA/nvidia-container-toolkit synced 2024-11-22 08:18:32 +00:00

History

Avi Deitcher 179d8655f9 Move nvidia-ctk hook command into own binary This change creates an nvidia-cdi-hook binary for implementing CDI hooks. This allows for these hooks to be separated from the nvidia-ctk command which may, for example, require libnvidia-ml to support other functionality. The nvidia-ctk hook subcommand is maintained as an alias for the time being to allow for existing CDI specifications referring to this path to work as expected. Signed-off-by: Avi Deitcher <avi@deitcher.net>		2024-05-21 12:19:44 +02:00
..
cdi	Move nvidia-ctk hook command into own binary	2024-05-21 12:19:44 +02:00
config	Use : as a config --set list separator	2024-05-08 17:26:59 +02:00
hook	Move nvidia-ctk hook command into own binary	2024-05-21 12:19:44 +02:00
info	run goimports -local against the entire codebase	2023-12-01 11:13:17 +01:00
runtime	Add support for adding additional containerd configs	2024-05-17 12:58:08 +02:00
system	Specify DRIVER_ROOT consistently	2024-02-09 14:28:56 +01:00
main.go	run goimports -local against the entire codebase	2023-12-01 11:13:17 +01:00
README.md

README.md

NVIDIA Container Toolkit CLI

The NVIDIA Container Toolkit CLI nvidia-ctk provides a number of utilities that are useful for working with the NVIDIA Container Toolkit.

Functionality

Configure runtimes

The runtime command of the nvidia-ctk CLI provides a set of utilities to related to the configuration and management of supported container engines.

For example, running the following command:

nvidia-ctk runtime configure --set-as-default

will ensure that the NVIDIA Container Runtime is added as the default runtime to the default container engine.

Configure the NVIDIA Container Toolkit

The config command of the nvidia-ctk CLI allows a user to display and manipulate the NVIDIA Container Toolkit configuration.

For example, running the following command:

nvidia-ctk config default

will display the default config for the detected platform.

Whereas

nvidia-ctk config

will display the effective NVIDIA Container Toolkit config using the configured config file, and running:

Individual config options can be set by specifying these are key-value pairs to the --set argument:

nvidia-ctk config --set nvidia-container-cli.no-cgroups=true

By default, all commands output to STDOUT, but specifying the --output flag writes the config to the specified file.

Generate CDI specifications

The Container Device Interface (CDI) provides a vendor-agnostic mechanism to make arbitrary devices accessible in containerized environments. To allow NVIDIA devices to be used in these environments, the NVIDIA Container Toolkit CLI includes functionality to generate a CDI specification for the available NVIDIA GPUs in a system.

In order to generate the CDI specification for the available devices, run the following command:\

nvidia-ctk cdi generate

The default is to print the specification to STDOUT and a filename can be specified using the --output flag.

The specification will contain a device entries as follows (where applicable):

An nvidia.com/gpu=gpu{INDEX} device for each non-MIG-enabled full GPU in the system
An nvidia.com/gpu=mig{GPU_INDEX}:{MIG_INDEX} device for each MIG-device in the system
A special device called nvidia.com/gpu=all which represents all available devices.

For example, to generate the CDI specification in the default location where CDI-enabled tools such as podman, containerd, cri-o, or the NVIDIA Container Runtime can be configured to load it, the following command can be run:

sudo nvidia-ctk cdi generate --output=/etc/cdi/nvidia.yaml

(Note that sudo is used to ensure the correct permissions to write to the /etc/cdi folder)

With the specification generated, a GPU can be requested by specifying the fully-qualified CDI device name. With podman as an exmaple:

podman run --rm -ti --device=nvidia.com/gpu=gpu0 ubuntu nvidia-smi -L