microsoft/kubegpu

A GPU / device extension framework for Kubernetes

Kubernetes GPU Project (KubeGPU)

The KubeGPU consists of two parts, one are the core extensions to Kubernetes (a CRI shim and a custom scheduler), and the other are the device-specific implementations implemented using Golang plugins for further extensibility. The project has been started and being worked on by Microsoft Research Lab in Redmond, USA.

The main project has now been split into the core and the plugins. For the core, please see: [https://github.com/Microsoft/KubeDevice]. The NVIDIA GPU plugins (which can be used in place of the standard NVIDIA device plugin) is located here.

Building the plugin

To build the plugin, make sure you have a Go installation. Then, run

go get github.com/NVIDIA/gpu-monitoring-tools/bindings/go/nvml
go get github.com/Microsoft/KubeDevice-API
go get github.com/Microsoft/KubeGPU
cd $GOPATH/src/github.com/Microsoft/KubeGPU
make

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.

kubegpu's Issues

How to support infiniband NIC to boost GPU communication?

In Readme, you metioned that

it does not have any support for other constraints on GPUs such as minimum GPU memory or multi-GPU connectivity

Can KubeGPU do that ?
Thanks

More instruction for installation?

The nvidia device plugin has a pool support. And the support of GPU topology understanding is a great feature. I'd like to have I try.

But for my pool understanding of kubernetes, could you please give more information for the installation and usage?

Thanks.

Installation

Could you be more specific on how to install the project? I guess I need go, are there any other dependencies?