Apache YuniKorn (Incubating)

Apache YuniKorn (Incubating) is a light-weighted, universal resource scheduler for container orchestrator systems. It was created to achieve fine-grained resource sharing for various workloads efficiently on a large scale, multi-tenant, and cloud-native environment. YuniKorn brings a unified, cross-platform scheduling experience for mixed workloads consists of stateless batch workloads and stateful services.

Build

Run the script build-docker-images.sh to build docker images.

# specify the docker repo
./build-docker-images.sh -r <REPO_NAME> -v <VERSION>

# for example, the following command will
# build 3 docker images like below:
#  foo/yunikorn-scheduler-k8s:0.8.0
#  foo/yunikorn-scheduler-admission-controller:0.8.0
#  foo/yunikorn-web:0.8.0
./build-docker-images.sh -r foo -v 0.8.0

Run YuniKorn on an existing K8s cluster

The simplest way to run YuniKorn is to use our helm charts, you can find the templates in the release package helm-charts. There are a few prerequisites:

1. A existing K8s cluster is up and running.
2. Helm chart client is installed.

Once you've built your own docker images, you will need to replace the docker image name in the helm chart templates, open helm-charts/yunikorn/values.yaml and replace the docker image addresses with ones you built. Then simply run command:

helm install ./yunikorn

For more instructions, please refer to User Guide.