custom-kube-scheduler/kubernetes/cluster/images/hyperkube

hyperkube

hyperkube is an all-in-one binary for the Kubernetes server components.

hyperkube is built for multiple architectures and the image is pushed automatically on every release.

Images for development scenarios

During the development process, it's effective to just build only the required binaries for specific architecture and generate single hyperkube image from it. The following shows how to generate linux/amd64 hyperkube image:

# Run the following from the top level kubernetes directory, to build the binaries necessary for creating hyperkube image.
$ KUBE_BUILD_PLATFORMS=linux/amd64 make kube-apiserver kube-controller-manager kube-proxy kube-scheduler kubectl kubelet

# Create and push the hyperkube image
$ REGISTRY=<registry> VERSION=<image version> ARCH=amd64 make -C cluster/images/hyperkube push

How to release by hand

# First, build the binaries
$ build/run.sh make cross

# Build for linux/amd64 (default)
# export REGISTRY=$HOST/$ORG to switch from staging-k8s.gcr.io

$ make push VERSION={target_version} ARCH=amd64
# ---> staging-k8s.gcr.io/hyperkube-amd64:VERSION
# ---> staging-k8s.gcr.io/hyperkube:VERSION (image with backwards-compatible naming)

$ make push VERSION={target_version} ARCH=arm
# ---> staging-k8s.gcr.io/hyperkube-arm:VERSION

$ make push VERSION={target_version} ARCH=arm64
# ---> staging-k8s.gcr.io/hyperkube-arm64:VERSION

$ make push VERSION={target_version} ARCH=ppc64le
# ---> staging-k8s.gcr.io/hyperkube-ppc64le:VERSION

$ make push VERSION={target_version} ARCH=s390x
# ---> staging-k8s.gcr.io/hyperkube-s390x:VERSION

If you don't want to push the images, run make or make build instead

[]()