1.3. Installing the Red Hat OpenShift AI RPM
To use AI models in MicroShift deployments, install the Red Hat OpenShift AI (Red Hat OpenShift AI Self-Managed) RPM with a new MicroShift installation. You can also install the RPM on an existing MicroShift instance if you restart the system.
The microshift-ai-model-serving RPM contains manifests that deploy kserve, with the raw deployment mode enabled, and ServingRuntimes objects in the redhat-ods-applications namespace.
Red Hat OpenShift AI is 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 the following link:
Prerequisites
- The system requirements for installing MicroShift have been met.
- You have root user access to your machine.
-
The OpenShift CLI (
oc) is installed. - You configured your LVM VG with the capacity needed for the PVs of your workload.
- You have the RAM and disk space required for your AI model.
- You configured the required accelerators, hardware, operating system, and MicroShift to provide the resources your model needs.
- Your AI model is ready to use.
Procedure
Install the MicroShift AI-model-serving RPM package by running the following command:
$ sudo dnf install microshift-ai-model-servingAs a root user, restart the MicroShift service by entering the following command:
$ sudo systemctl restart microshiftOptional: Install the release information package by running the following command:
$ sudo dnf install microshift-ai-model-serving-release-info注記The
microshift-ai-model-serving-release-infoRPM contains a JSON file with image references useful for offline procedures or deploying a copy of aServingRuntimeCustom Resource to your namespace during a bootc image build.
Verification
Verify that the
kservepod is running in theredhat-ods-applicationsnamespace by entering the following command:$ oc get pods -n redhat-ods-applicationsExample output
NAME READY STATUS RESTARTS AGE kserve-controller-manager-7fc9fc688-kttmm 1/1 Running 0 1h
Next steps
- Create a namespace for your AI model.
- Package your model into an OCI image.
- Configure a model-serving runtime.
- Verify that your model is ready for inferencing.
- Make requests against the model server.