Chapter 1. Architecture of OpenShift AI
Red Hat OpenShift AI is a fully Red Hat managed cloud service that is available as an Add-on to Red Hat OpenShift Dedicated and to Red Hat OpenShift Service on Amazon Web Services (ROSA).
OpenShift AI integrates the following components and services:
At the service layer:
- OpenShift AI dashboard
- A customer-facing dashboard that shows available and installed applications for the OpenShift AI environment as well as learning resources such as tutorials, quick start examples, and documentation. You can also access administrative functionality from the dashboard, such as user management, cluster settings, accelerator profiles, and notebook image settings. In addition, data scientists can create their own projects from the dashboard. This enables them to organize their data science work into a single project.
- Model serving
- Data scientists can deploy trained machine-learning models to serve intelligent applications in production. After deployment, applications can send requests to the model using its deployed API endpoint.
- Data science pipelines
- Data scientists can build portable machine learning (ML) workflows with data science pipelines, using Docker containers. This enables your data scientists to automate workflows as they develop their data science models.
- Jupyter (Red Hat managed)
- A Red Hat managed application that allows data scientists to configure their own notebook server environment and develop machine learning models in JupyterLab.
- TrustyAI
- Data scientists can review local, global, and time-series explainers for predictive models in both enterprise and data science applications. They can also use tools to detect bias and drift. These capabilities help organizations to deliver higher quality, unbiased AI-based applications.
- Distributed workloads
- Data scientists can use multiple nodes in parallel to train machine-learning models or process data more quickly. This approach significantly reduces the task completion time, and enables the use of larger datasets and more complex models.
The distributed workloads feature is currently available in Red Hat OpenShift AI as a Technology Preview feature only. Technology Preview features are not supported with Red Hat production service level agreements (SLAs) and might not be functionally complete. Red Hat does not recommend using them in production. These features provide early access to upcoming product features, enabling customers to test functionality and provide feedback during the development process.
For more information about the support scope of Red Hat Technology Preview features, see Technology Preview Features Support Scope.
At the management layer:
- The Red Hat OpenShift AI Operator
- A meta-operator that deploys and maintains all components and sub-operators that are part of OpenShift AI.
- Monitoring services
- Alertmanager, OpenShift Telemetry, and Prometheus work together to gather metrics from OpenShift AI and organize and display those metrics in useful ways for monitoring and billing purposes. Alerts from Alertmanager are sent to PagerDuty, responsible for notifying Red Hat of any issues with your managed cloud service.
When you install the OpenShift Data Science Add-on in the Cluster Manager, the following new projects are created:
-
The
redhat-ods-operatorproject contains the Red Hat OpenShift AI Operator. -
The
redhat-ods-applicationsproject installs the dashboard and other required components of OpenShift AI. -
The
redhat-ods-monitoringproject contains services for monitoring and billing. -
The
rhods-notebooksproject is where notebook environments are deployed by default.
You or your data scientists must create additional projects for the applications that will use your machine learning models.
Do not install independent software vendor (ISV) applications in namespaces associated with OpenShift AI add-ons unless you are specifically directed to do so on the application tile on the dashboard.
