Red Hat has unveiled a strategy to minimize server requirements for high-availability edge deployments of its OpenShift platform, addressing customer concerns over escalating hardware expenses. In a recent post, Red Hat engineers Daniel Froehlich and Paul Lancaster highlighted rising interest in alternative architectures for large-scale edge environments, driven by the financial burden of maintaining a third node across numerous locations. This challenge has intensified amid recent spikes in hardware pricing.
The company already provides a two-node OpenShift configuration using an “arbiter” device—equipped with minimal specifications like two virtual CPUs, 8GB RAM, and 50GB storage—to maintain quorum and prevent cluster split-brain scenarios. However, even the modest cost and logistical complexity of deploying arbiter devices are now prompting further innovation.
Red Hat now employs fencing techniques adapted from its Red Hat Enterprise Linux High Availability Add-On, leveraging Corosync and Pacemaker technologies. Fencing ensures cluster integrity by mandating that Pacemaker forcibly powers down or reboots disconnected nodes. When a node loses connectivity, Corosync marks it as failed, while Pacemaker assumes control without risking concurrent active nodes—a critical improvement for simplified edge environments.
The approach requires servers with baseboard management controllers compatible with the Redfish API. However, limitations exist: if a previously disconnected node reactivates, consistency cannot be guaranteed during reboot. In cases where both nodes fail simultaneously, manual recovery procedures may still be necessary.
This development is pivotal for edge computing use cases, such as real-time video analytics in retail locations, where cost-effective resilience is essential. The two-node OpenShift configuration also supports OpenShift Virtualization, positioning Red Hat as a stronger competitor against VMware in edge infrastructure solutions.
