Shared cluster components

Shared cluster components provide underlying functionality for the IBM Cloud Pak® for Data control plane and services. Depending on the services that you plan to install or the features that you want to use in IBM Cloud Pak for Data, you might need to install one or more required components on your cluster.

Shared cluster components are different from end-user services, such as the Watson™ Studio or Watson Query services. Unlike typical services, end users do not interact directly with the shared cluster components.

Use the following information to determine whether you need to install any of these components on your cluster.

Component Required if... Details
IBM Cloud Pak foundational services Required for all Cloud Pak for Data installations. The IBM Cloud Pak foundational services are installed once on the cluster and can be used by all instances of Cloud Pak for Data and any other IBM® Cloud Paks on the cluster.

You can install the IBM Cloud Pak foundational services when you install Cloud Pak for Data.

Cloud Pak for Data Scheduling service
The scheduling service offers enhancements over the default Kubernetes scheduler, such as:
  • Advanced pod placement logic
  • Resource sharing
  • Quota logic
You must install the scheduling service if you plan to use:
  • The quota enforcement feature in Cloud Pak for Data
  • The Watson Machine Learning Accelerator service
The scheduling service is installed once on the cluster and can be used by all instances of Cloud Pak for Data on the cluster.

You can install the scheduling service when you install Cloud Pak for Data.

If you have an existing Cloud Pak for Data installation, you must delete all the pods in the deployment to allow the scheduling service to take over.

For details, see:

Cloud Pak for Data common core services
The common core services provide features that are used by multiple Cloud Pak for Data services, including:
  • Connections
  • Deployment management
  • Job management
  • Notifications
  • Projects
  • Search
  • Metadata repositories
You are installing a service that relies on the common core services. The common core services services are installed once in a given Red Hat® OpenShift® project. Any services in the project that rely on the common core services can use the same installation.
The common core services are automatically installed by the services that require them:
  • If you install or upgrade a service that requires the common core services and the common core services are not installed or are not at the required level, the service will install or upgrade them.
  • If you install or upgrade a service that requires the common core services and the common core services are already installed at the required level, the service will use the existing installation.

For details, see the System requirements.