Overview

Pattern objectives

The objective of this pattern is to provide a resilient solution design for Red Hat OpenShift Kubernetes Service on a Virtual Private Cloud (VPC) architecture deployment, on IBM Cloud, that meets high availability requirements for enterprise workloads that require persistent storage. This pattern is intended to:

• Accelerate and simplify solution design by providing a standard IBM Cloud deployment architecture reference following the IBM Architecture Design Framework.
• Provide a prescriptive, end-to-end enterprise-class solution design, with diagrams, component architecture decisions along with rationale for cloud component selection to meet enterprise requirements.
• Help ensure that requirements can be met from a performance, system availability, and security perspective.

Pattern details

Red Hat OpenShift service on IBM Cloud multizone resiliency pattern deploys containerized enterprise workloads that require persistent storage on VPC, using compute, storage, and network cloud resources as well as other cloud services provisioned across multiple availability zones within a single region.

The multizone Red Hat OpenShift cluster pattern provides 99.99% infrastructure availability. For high availability of stateful application, close coordination between application teams, infrastructure teams (including availability of storage), and database teams is essential to design highly available stateful applications on a multizone Red Hat OpenShift cluster.

This pattern uses a multizone workload cluster that is configured with Portworx-Store to provide persistent storage for databases and other stateful application components and a Portworx-Backup cluster for application data backup and recovery.

Deploying Red Hat OpenShift cluster across three availability zones is the recommended option for a highly available cluster to be used for highly available stateful applications. IBM Cloud-managed control plane nodes are automatically distributed across availability zones on IBM Cloud. The network communication across IBM Cloud availability zones has low enough latency to satisfy Red Hat OpenShift etcd and Portworx requirements. Portworx provides a robust data service platform for persistent storage with replication and high availability features across multiple availability zones to run stateful containerized applications. This pattern is not to be used across IBM Cloud regions, which have higher latency for region-to-region network communication.

Red Hat OpenShift service on IBM Cloud multizone resiliency pattern can be used to support business continuity policies or regulatory requirements with country boundaries or geo data residence constraints. It does not support out-of-region disaster recovery.