Red Hat OpenShift on OCI for Bare Metal and GPU shapes

We’re excited to announce the general availability for customers to launch their Red Hat OpenShift workloads on bare metal instances and support for GPU instances accelerated by NVIDIA.

Red Hat OpenShift provides a complete application environment that enables a consistent experience for applications, regardless of the underlying platform. Our mutual customers have invested time and resources in configuring and running Red Hat OpenShift. They want to bring workloads to Oracle Cloud Infrastructure (OCI) to take advantage of flexible compute, block storage autotuning, and powerful services such as Oracle’s Autonomous Database.

Red Hat OpenShift is validated and certified on bare metal instances. We’re also launching new features on the storage drivers Container Storage Interface (CSI) and Cloud Controller Manager (CCM) to enable customers to run high-performance Kubernetes workloads efficiently.

Red Hat OpenShift on OCI

With this release, you can bring your Red Hat OpenShift license to launch your workloads on bare metal instances that provide customers direct access to hardware resources without the overhead of virtualization, giving full control over configuration to enable running high-performance large-scale workloads. With the addition of the Raw Block Volume access launch, you can now access the volume directly, instead of mounting storage as a file system, which provides more performance benefits.

As part of this release, we removed a previous limitation of Cloud Controller Manager (CCM) on using only one virtual network interface card (VNIC) to route all traffic. We enhanced the driver to support a secondary VNIC to route traffic for OpenShift to handle all incoming customer traffic to enhance performance via a dedicated network.

Select Availability GPU Instances Accelerated by NVIDIA

We’re excited to announce select availability of support for GPU instances accelerated by NVIDIA on Red Hat OpenShift. RedHat OpenShift is validated on the GPU shapes, BM.GPU.A100 and BM.GPU.H100. This early release provides customers with the opportunity to run AI workloads on OCI. You can now run OpenShift on a standalone NVIDIA A100 or NVIDIA H100 GPU and set up a remote direct memory access (RMDA) manually to run their workloads in a cluster mode. 

This release is the result of a collaborative effort between Red Hat and Oracle, who have worked hard to bring this offering into the hands of our customers. We’re excited to see the AI applications that you build on this high-performance platform.

And we’re not done! We’re soon introducing the support of Red Hat OpenShift AI on OCI. Stay tuned!

Join us at NVIDIA GTC March 17-21 booth #1515 to learn more about our partnerships!

 

Try for yourself

Get started with Red Hat OpenShift on Oracle Cloud Infrastructure. We’re excited for you to bring your extended workloads, including Red Hat OpenShift environments, onto the cloud built for enterprise applications. For more information, see the documentation and the following resources: