Using Openshift

You can use containerized execution using Openshift 4 as the underlying Kubernetes engine.

Setup

Create your Openshift cluster

Dataiku leverages Kubernetes as the base cluster API for workload runtime. Dataiku does not manage OpenShift clusters. Only “unmanaged” operation is possible. Attach your Openshift cluster to DSS as if it was a regular kubernetes cluster.

You’ll be able to configure the memory allocation for each container and per-namespace in Dataiku DSS using multiple containerized execution configurations.

Prepare your local oc commands

Follow the Openshift documentation to ensure the following on your local machine

  • The oc command is installed and can interact with the cluster. This will require a dedicated service account with authentication credentials. Follow the Openshfit documentation on service accounts

Create new image build and containerized execution configurations

  • Go to Administration > Settings > Containerized Execution.

  • Add a new image build configuration.

  • Select Openshift as the builder type.

  • Select the Openshift namespace in which you want the build to be performed. By default, images are published into the registry in the same namespace where the build action is performed.

  • Add a new container execution configuration of type Kubernetes.

Warning

Changing the Openshift namespace will most likely require rebuilding all container images built by Dataiku. Depending on your setup, images are only published in this namespace.

You can optionally configure additional namespaces you want images to be made available using Target Namespaces. Once the build is finished, the images will be copied into the target namespaces.