microsoft / pai / 11683

34%

push

travis-ci

Support nvidia-container-runtime for gpu isolation (#2352)

* Support nvidia container runtime for gpu isolation

Support nvidia-container-runtime for gpu isolation.
Set NVIDIA_VISIBLE_DEVICES to void to avoid conflict with runc, reference:
https://github.com/NVIDIA/nvidia-container-runtime#nvidia_visible_devices.

Works for w/ and w/o --runtime=nvidia.

Closes #1667.

* Set default runtime to runc

Set runtime to runc explicitly to overwrite default runtime.

* Drop `MKNOD` capability

Drop `MKNOD` capability when starting container.
`nvidia-smi` will make all gpu devices show up under /dev,
drop [mknod](https://linux.die.net/man/2/mknod) capability to avoid this.

Reference: https://github.com/NVIDIA/nvidia-docker/issues/170

438 of 738 branches covered (59.35%)

Branch coverage included in aggregate %.

4146 of 7969 relevant lines covered (52.03%)

3.44 hits per line