GPU Manager is used for managing the nvidia GPU devices in Kubernetes cluster. It implements the DevicePlugin interface of Kubernetes. So it's compatible with 1.9 of Kubernetes release version.

To compare with the combination solution of nvidia-docker and nvidia-k8s-plugin, GPU manager will use native runc without modification but nvidia solution does. Besides we also support metrics report without deploying new components.

To schedule a GPU payload correctly, GPU manager should work with gpu-admission which is a kubernetes scheduler plugin.

GPU manager also supports the payload with fraction resource of GPU device such as 0.1 card or 100MiB gpu device memory. If you want this kind feature, please refer to vcuda-controller project.

1. Build binary

• Prerequisite
  • CUDA toolkit

make

2. Build image

• Prerequisite
  • Docker

make img

Prebuilt image can be found at thomassong/gpu-manager

GPU Manager is running as daemonset, and because of the RABC restriction and hydrid cluster, you need to do the following steps to make this daemonset run correctly.

• service account and clusterrole

kubectl create sa gpu-manager -n kube-system
kubectl create clusterrolebinding gpu-manager-role --clusterrole=cluster-admin --serviceaccount=kube-system:gpu-manager

• label node with nvidia-device-enable=enable

kubectl label node <node> nvidia-device-enable=enable

• submit daemonset yaml

kubectl create -f gpu-manager.yaml

There is nothing special to submit a Pod except the description of GPU resource is no longer 1 . The GPU resources are described as that 100 tencent.com/vcuda-core for 1 GPU and N tencent.com/vcuda-memory for GPU memory (1 tencent.com/vcuda-memory means 256Mi GPU memory). And because of the limitation of extend resource validation of Kubernetes, to support GPU utilization limitation, you should add tencent.com/vcuda-core-limit: XX in the annotation field of a Pod.

Notice: the value of tencent.com/vcuda-core is either the multiple of 100 or any value smaller than 100.For example, 100, 200 or 20 is valid value but 150 or 250 is invalid

• Submit a Pod with 0.3 GPU utilization and 7680MiB GPU memory with 0.5 GPU utilization limit

apiVersion: v1
kind: Pod
metadata:
  name: vcuda
  annotations:
    tencent.com/vcuda-core-limit: 50
spec:
  restartPolicy: Never
  containers:
  - image: <test-image>
    name: nvidia
    command:
    - /usr/local/nvidia/bin/nvidia-smi
    - pmon
    - -d
    - 10
    resources:
      requests:
        tencent.com/vcuda-core: 50
        tencent.com/vcuda-memory: 30
      limits:
        tencent.com/vcuda-core: 50
        tencent.com/vcuda-memory: 30

• Submit a Pod with 2 GPU card

apiVersion: v1
kind: Pod
metadata:
  name: vcuda
spec:
  restartPolicy: Never
  containers:
  - image: <test-image>
    name: nvidia
    command:
    - /usr/local/nvidia/bin/nvidia-smi
    - pmon
    - -d
    - 10
    resources:
      requests:
        tencent.com/vcuda-core: 200
        tencent.com/vcuda-memory: 60
      limits:
        tencent.com/vcuda-core: 200
        tencent.com/vcuda-memory: 60

If you have some questions about this project, you can first refer to FAQ to find a solution.