gpu

Documentation

Overview

Package gpu implements GPU plugin.

Index

• type Config
• type Tester
  • func New(cfg Config) (Tester, error)

Types

type Config

type Config struct {
	Logger    *zap.Logger
	Stopc     chan struct{}
	Sig       chan os.Signal
	EKSConfig *eksconfig.Config
	K8SClient k8sClientSetGetter
	Namespace string
}

Config defines GPU configuration.

type Tester

type Tester interface {
	// InstallNvidiaDriver installs the Nvidia device plugin for Kubernetes.
	// After GPU worker nodes join the cluster, one must apply the Nvidia
	// device plugin for Kubernetes as a DaemonSet.
	// ref. https://docs.aws.amazon.com/eks/latest/userguide/create-managed-node-group.html
	// ref. https://docs.aws.amazon.com/eks/latest/userguide/gpu-ami.html
	// ref. https://github.com/NVIDIA/k8s-device-plugin
	InstallNvidiaDriver() error
	// RunNvidiaSMI launches a pod manifest that launches a Cuda container that
	// runs "nvidia-smi" on a GPU worker node.
	// ref. https://docs.aws.amazon.com/eks/latest/userguide/gpu-ami.html
	RunNvidiaSMI() error
}

Tester defines GPU tester.

func New

func New(cfg Config) (Tester, error)

New creates a new Job tester.