Chapter 4. Support removals

This section describes major changes in support for user-facing features in Red Hat OpenShift AI. For information about OpenShift AI supported software platforms, components, and dependencies, see Supported configurations.

4.1. Data science pipelines v1 upgraded to v2

Previously, data science pipelines in OpenShift AI were based on KubeFlow Pipelines v1. Data science pipelines are now based on KubeFlow Pipelines v2, which uses a different workflow engine. Data Science Pipelines (DSP) 2.0 is enabled and deployed by default in OpenShift AI. It is no longer possible to deploy, view, or edit the details of pipelines that are based on DSP 1.0 from the dashboard. For more information, see Enabling Data Science Pipelines 2.0.


DSP 2.0 contains an installation of Argo Workflows. OpenShift AI does not support direct customer usage of this installation of Argo Workflows. To install or upgrade to OpenShift AI with DSP 2.0, ensure that there is no existing installation of Argo Workflows on your cluster.

If you want to use existing pipelines and workbenches with DSP 2.0 after upgrading OpenShift AI, you must update your workbenches to use the 2024.1 notebook image version and then manually migrate your pipelines from DSP 1.0 to 2.0. For more information, see Upgrading to DSP 2.0.

4.2. Removal of bias detection (TrustyAI)

Starting with OpenShift AI 2.7, the bias detection (TrustyAI) functionality has been removed. If you previously had this functionality enabled, upgrading to OpenShift AI 2.7 or later will remove the feature. The default TrustyAI notebook image remains supported.

4.3. Version 1.2 notebook container images for workbenches are no longer supported

When you create a workbench, you specify a notebook container image to use with the workbench. Starting with OpenShift AI 2.5, when you create a new workbench, version 1.2 notebook container images are not available to select. Workbenches that are already running with a version 1.2 notebook image continue to work normally. However, Red Hat recommends that you update your workbench to use the latest notebook container image.

4.4. NVIDIA GPU Operator replaces NVIDIA GPU add-on

Previously, to enable graphics processing units (GPUs) to help with compute-heavy workloads, you installed the NVIDIA GPU add-on. OpenShift AI no longer supports this add-on.

Now, to enable GPU support, you must install the NVIDIA GPU Operator. To learn how to install the GPU Operator, see NVIDIA GPU Operator on Red Hat OpenShift Container Platform (external).

4.5. Kubeflow Notebook Controller replaces JupyterHub

In OpenShift AI 1.15 and earlier, JupyterHub was used to create and launch notebook server environments. In OpenShift AI 1.16 and later, JupyterHub is no longer included, and its functionality is replaced by Kubeflow Notebook Controller.

This change provides the following benefits:

  • Users can now immediately cancel a request, make changes, and retry the request, instead of waiting 5+ minutes for the initial request to time out. This means that users do not wait as long when requests fail, for example, when a notebook server does not start correctly.
  • The architecture no longer prevents a single user from having more than one notebook server session, expanding future feature possibilities.
  • The removal of the PostgreSQL database requirement allows for future expanded environment support in OpenShift AI.

However, this update also creates the following behavior changes:

  • For IT Operations administrators, the notebook server administration interface does not currently allow login access to data scientist users' notebook servers. This is planned to be added in future releases.
  • For data scientists, the JupyterHub interface URL is no longer valid. Update your bookmarks to point to the OpenShift AI Dashboard.

The JupyterLab interface is unchanged and data scientists can continue to use JupyterLab to work with their notebook files as usual.