Add support for nvidia MIG

[piyush-jena]

Issue number:

Closes #

Description of changes:

• Added migrations for the following settings and configuration-files

settings.kubelet-device-plugins.nvidia.device-partitioning-strategy
settings.kubelet-device-plugins.nvidia.mig
configuration-files.nvidia-k8s-device-plugin-mig-conf

• Rebased with core-kit version bump

Testing done:

Migration test between v1.32.0 and v1.33.0

v1.32.0

[ssm-user@control]$ apiclient get settings.kubelet-device-plugins.nvidia
{
  "settings": {
    "kubelet-device-plugins": {
      "nvidia": {
        "device-id-strategy": "index",
        "device-list-strategy": "volume-mounts",
        "device-sharing-strategy": "none",
        "pass-device-specs": true
      }
    }
  }
}
bash-5.1# cat /etc/os-release
NAME=Bottlerocket
ID=bottlerocket
VERSION="1.32.0 (aws-k8s-1.29-nvidia)"
PRETTY_NAME="Bottlerocket OS 1.32.0 (aws-k8s-1.29-nvidia)"
VARIANT_ID=aws-k8s-1.29-nvidia
VERSION_ID=1.32.0
BUILD_ID=cacc4ce9
HOME_URL="https://github.com/bottlerocket-os/bottlerocket"
SUPPORT_URL="https://github.com/bottlerocket-os/bottlerocket/discussions"
BUG_REPORT_URL="https://github.com/bottlerocket-os/bottlerocket/issues"
DOCUMENTATION_URL="https://bottlerocket.dev"
bash-5.1# updog check-update -a --json
[
  {
    "variant": "aws-k8s-1.29-nvidia",
    "arch": "x86_64",
    "version": "1.33.0",
    "max_version": "1.33.0",
    "waves": {
      "0": "2025-02-07T12:26:28.495108532Z",
      "20": "2025-02-07T15:26:28.495108532Z",
      "102": "2025-02-08T11:26:28.495108532Z",
      "307": "2025-02-09T11:26:28.495108532Z",
      "819": "2025-02-11T11:26:28.495108532Z",
      "1228": "2025-02-12T11:26:28.495108532Z",
      "1843": "2025-02-13T11:26:28.495108532Z"
    },
    "images": {
      "boot": "bottlerocket-aws-k8s-1.29-nvidia-x86_64-1.33.0-1fb8b819-dirty-boot.ext4.lz4",
      "root": "bottlerocket-aws-k8s-1.29-nvidia-x86_64-1.33.0-1fb8b819-dirty-root.ext4.lz4",
      "hash": "bottlerocket-aws-k8s-1.29-nvidia-x86_64-1.33.0-1fb8b819-dirty-root.verity.lz4"
    }
  }
]
bash-5.1# updog update -i 1.33.0 -r -n
Starting update to 1.33.0
Reboot scheduled for Fri 2025-02-07 11:37:00 UTC, use 'shutdown -c' to cancel.
Update applied: aws-k8s-1.29-nvidia 1.33.0

Post upgrade to v1.33.0 - Unrelated lines in apiclient get configuration-files have been omitted for brevity.

[ssm-user@control]$ apiclient get settings.kubelet-device-plugins.nvidia
{
  "settings": {
    "kubelet-device-plugins": {
      "nvidia": {
        "device-id-strategy": "index",
        "device-list-strategy": "volume-mounts",
        "device-partitioning-strategy": "none",
        "device-sharing-strategy": "none",
        "pass-device-specs": true
      }
    }
  }
}
[ssm-user@control]$ apiclient get configuration-files
    ****
    "nvidia-k8s-device-plugin-mig-conf": {
      "path": "/etc/nvidia-migmanager/nvidia-migmanager.toml",
      "template-path": "/usr/share/templates/nvidia-k8s-device-plugin-mig-conf"
    },
    ****
bash-5.1# journalctl -u nvidia-migmanager
Feb 07 11:36:39 ip-172-31-0-165.us-west-2.compute.internal systemd[1]: Starting NVIDIA MIG manager service...
Feb 07 11:36:39 ip-172-31-0-165.us-west-2.compute.internal nvidia-migmanager[1394]: 11:36:39 [INFO] nvidia-migmanager started
Feb 07 11:36:39 ip-172-31-0-165.us-west-2.compute.internal nvidia-migmanager[1394]: 11:36:39 [INFO] Fetching GPU devices data ...
Feb 07 11:36:39 ip-172-31-0-165.us-west-2.compute.internal nvidia-migmanager[1394]: 11:36:39 [WARN] Found NVIDIA Device but couldn't confirm variant.
Feb 07 11:36:39 ip-172-31-0-165.us-west-2.compute.internal systemd[1]: Finished NVIDIA MIG manager service.
bash-5.1# signpost rollback-to-inactive

Post downgrade back to v1.32.0 - Unrelated lines in apiclient get configuration-files have been omitted for brevity.

[ssm-user@control]$ apiclient get settings.kubelet-device-plugins.nvidia
{
  "settings": {
    "kubelet-device-plugins": {
      "nvidia": {
        "device-id-strategy": "index",
        "device-list-strategy": "volume-mounts",
        "device-sharing-strategy": "none",
        "pass-device-specs": true
      }
    }
  }
}
[ssm-user@control]$ apiclient get configuration-files
    ****  
    ****

Terms of contribution:

By submitting this pull request, I agree that this contribution is dual-licensed under the terms of both the Apache License, version 2.0, and the MIT license.

[bcressey]

New defaults and migrations look good to me.