Kubeflow Contributor Guide

Welcome to the Kubeflow/MPI-Operator project! We'd love to accept your patches and contributions to this project. Please read the contributor's guide in our docs.

The contributor's guide

shows you where to find the Contributor License Agreement (CLA) that you need to sign,
helps you get started with your first contribution to Kubeflow,
and describes the pull request and review workflow in detail, including the OWNERS files and automated workflow tool.

Set Up Development Environment

If you haven't done so, please follow the instructions here to fork and clone the repository, and then configure the remote repository for the repository you just cloned locally. Note that you'd probably want to clone your forked repository to be under your GOPATH, for example:

mkdir -p ${GOPATH}/src/github.com/kubeflow
cd ${GOPATH}/src/github.com/kubeflow
git clone https://github.com/${GITHUB_USER}/mpi-operator.git

Install Dependencies

We use Go v1.20+ for development and use Go Modules to download and install the dependencies.

Run tests

Unit and integration tests

You can execute all the unit and integration tests via make test.

You can find the unit tests in the same folders as the functional code.

You can find the integration tests in a separate directory, test/integration. Integration tests make use of a real kube-apiserver to test the interaction of the controller with a real Kubernetes API. In these tests, other components are not running, including kubelet or kube-controller-manager.

Consider adding an integration test if your feature makes new API calls.

E2E tests

E2E tests run against a real cluster. In our tests, we create a cluster using kind.

You can run the tests with make test_e2e.

If desired, you can run the tests against any existing cluster. Just make sure that credentials for the cluster are present in ${HOME}/.kube/config and run:

USE_EXISTING_CLUSTER=true make test_e2e

Check Code Style

We use golangci-lint to check issues on code style. Please also check out this wiki for some additional instructions on code review.

You can run formatter and linter with:

make fmt lint

Run

You have to build the image and deploy the standalone YAMLs in a cluster.

make images dev_manifest
kubectl apply -k manifests/overlays/dev

The controller handles MPIJobs with API version v2beta1 or newer.

If you need to use a different registry, or a different tag, you can do:

make IMAGE_NAME=example.com/mpi-operator RELEASE_VERSION=dev make images dev_manifest

To look at the controller's logs, you can do:

kubectl logs -n mpi-operator -f deployment/mpi-operator

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!