UniqAI/nvidia-container-toolkit
1
0
Fork 0
mirror of https://github.com/NVIDIA/nvidia-container-toolkit synced 2024-11-22 16:29:18 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
b0f7a3809f
nvidia-container-toolkit/cmd/nvidia-container-runtime
History
Evan Lezar c76946cbcc Add nvidia-container-runtime.runtimes config option 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2022-05-09 09:38:12 +02:00
..
modifier Add compute capability of first device as arch property 2022-05-05 14:11:30 +02:00
logger.go Copy code from nvidia-container-runtime 2021-09-07 13:13:03 +02:00
main_test.go Add nvidia-container-runtime.runtimes config option 2022-05-09 09:38:12 +02:00
main.go Add debug log for command line arguments 2022-05-05 13:47:39 +02:00
README.md Add csv discovery mode to experimental runtime 2022-04-07 10:45:19 +02:00
runtime_factory_test.go Add nvidia-container-runtime.runtimes config option 2022-05-09 09:38:12 +02:00
runtime_factory.go Add nvidia-container-runtime.runtimes config option 2022-05-09 09:38:12 +02:00

README.md

The NVIDIA Container Runtime

The NVIDIA Container Runtime is a shim for OCI-compliant low-level runtimes such as runc. When a create command is detected, the incoming OCI runtime specification is modified in place and the command is forwarded to the low-level runtime.

Standard Mode

In the standard mode configuration, the NVIDIA Container Runtime adds a prestart hook to the incomming OCI specification that invokes the NVIDIA Container Runtime Hook for all containers created. This hook checks whether NVIDIA devices are requested and ensures GPU access is configured using the nvidia-container-cli from project libnvidia-container.

Experimental Mode

The NVIDIA Container Runtime can be configured in an experimental mode by setting the following options in the runtime's config.toml file:

[nvidia-container-runtime]
experimental = true

When this setting is enabled, the modifications made to the OCI specification are controlled by the nvidia-container-runtime.discover-mode option, with the following mode supported:

  • "legacy": This mode mirrors the behaviour of the standard mode, inserting the NVIDIA Container Runtime Hook as a prestart hook into the container's OCI specification.
  • "csv": This mode uses CSV files at /etc/nvidia-container-runtime/host-files-for-container.d to define the devices and mounts that are to be injected into a container when it is created.

Notes on using the docker CLI

The docker CLI supports the --gpus flag to select GPUs for inclusion in a container. Since specifying this flag inserts the same NVIDIA Container Runtime Hook into the OCI runtime specification. When experimental mode is activated, the NVIDIA Container Runtime detects the presence of the hook and raises an error. This requirement will be relaxed in the near future.