This section will guide you through deploying Kubeflow alongside Rok, using installation manifests provided by Arrikto. Kubeflow is a platform and integrated toolkit for Machine Learning built on Kubernetes. The Arrikto Enterprise Kubeflow (EKF) distribution extends the capabilities of OSS Kubeflow with additional automation, reproducibility, portability, and security features.
You will deploy Kubeflow in an edit/commit/apply manner. More specifically, you will
- Edit: Kubeflow comes with some default kustomizations. You will change
them by editing the
deployoverlay of those kustomizations.
- Commit: You will commit all changes to your GitOps repository.
- Apply: You will apply everything to Kubernetes.
What You’ll Need¶
- A configured management environment.
- Your clone of the Arrikto GitOps repository.
- An existing Kubernetes cluster.
- A configured Rok user.