• 2 min read
Red Hat cuts edge OpenShift to two servers
Red Hat is pushing a two-server OpenShift edge setup to avoid the cost of a third node or arbiter as hardware prices climb.
Image: The Register
Red Hat is pitching a two-server design for high-availability OpenShift at the edge, arguing that customers increasingly see a third node as too expensive to power, maintain, and deploy across hundreds or thousands of sites.
In a post published last week, Daniel Froehlich and Paul Lancaster said customers are asking for “alternative topologies for large-scale edge deployments” because of “the prohibitive cost of powering, maintaining, and deploying a third node across hundreds or thousands of sites — a motivation that has only grown stronger in light of the recent steep increase in hardware prices.”
Red Hat already had a two-node option using an arbiter: a small device with at least two virtual CPUs, 8GB of RAM, and 50GB of solid-state storage. Its job is to maintain quorum and prevent the split-brain problem, where two disconnected nodes both assume control and risk overwriting each other’s changes.
Now Red Hat is moving to remove even that extra system. The company has adapted technology from the Red Hat Enterprise Linux High Availability Add-On — specifically Corosync and Pacemaker — to use fencing instead.
“Fencing imposes consistency by requiring that Pacemaker on the surviving node forcefully powers down or reboots the failed or disconnected node.”
The idea is straightforward: if one node loses contact with the other, Corosync flags a failure and Pacemaker treats the missing node as dead, taking over its workload rather than risking a split-brain state. But there is a hardware requirement: servers need a baseboard management controller that supports the Redfish API.
Recommended reading
Why AI agent rules break without OS context
Red Hat also acknowledges limits. In some cases, such as when a node comes back after its independent power supply recovers, the company cannot guarantee consistency on restart. If both nodes go down and only one comes back, manual intervention may be needed.
The setup matters because edge sites such as retail locations are expected to run inferencing hardware for tasks like near-real-time video analysis. Red Hat’s two-node OpenShift design also supports OpenShift Virtualization, giving it another angle in the fight for customers reconsidering VMware.
Enterprise Editor
Marcus follows the money. He covers enterprise software, cloud architecture, and the tectonic shifts in Big Tech strategy. He translates dense earnings calls and complex M&A activity into actionable insights about where the industry is actually heading. If a tech giant makes a silent pivot, Marcus is usually the first to notice.
via The Register
