SC//HyperCore virtualization suite now supports 2-node clusters with physical witness support. Leveraging a lightweight device dedicated to quorum, organizations get the resiliency and high-availability behavior of a three-node cluster, without the cost, power, or footprint of a third compute node.
For years, edge infrastructure has been shaped by excess. Every new application, especially AI workloads, seemed to demand another hardware system, more power, more space, and a bigger budget. That approach was tolerable when hardware was cheap and inefficiency was easy to absorb. But rising hardware costs, tighter budgets, and shrinking edge footprints have pushed many organizations toward leaner 2-node infrastructure. The tradeoff? Native 2-node clusters have typically had drawbacks in the form of complex configuration or weaker availability guarantees. Our approach has previously been to to move to a full three-node cluster—which solved the reliability gap but reintroduced the exact cost, power, and space problems that drove the move to 2-node in the first place.
For many edge environments, a three-node cluster simply isn't realistic. There's no budget, no rack space, no power headroom to justify it. Until now, that meant living with compromised availability.
Closing the Gap With a Lightweight Witness
Organizations that need dependable uptime for business-critical workloads but require a highly available 2-node virtualization platform due to cost, power, or footprint reasons now have a way to close the availability gap and achieve 3-node-level resiliency, at a fraction of the cost. SC//HyperCore virtualization suite with physical witness support is a hyperconverged infrastructure (HCI) solution that supports a lightweight, low-cost cluster configuration by adding a witness device dedicated solely to quorum.
The witness is not a compute node—it does not run storage or host virtual machines. Its only job is helping the cluster maintain quorum, giving 2-node clusters 3-node-level resiliency without the expense, cloud dependency, or added licensing and workload overhead.
For edge deployments, SC//HyperCore virtualization suite with witness support enables IT and infrastructure teams to:
- Improve availability during a node-loss event, with the same I/O delay as a full three node cluster Recover automatically, with no manual intervention required
- Eliminate cloud dependency by keeping quorum entirely on-premises
- Preserve rack space and reduce hardware costs
- Scale seamlessly to a larger cluster whenever you're ready
Enterprise-Grade Resilience, Edge-Ready Simplicity
SC//HyperCore virtualization suite with physical witness support is designed to deliver 3-node reliability with 2-node simplicity.
Simple to deploy.
Configure the witness through cluster initialization—no separate console, no complex setup.
Full high availability, near-instant failover.
During a node-loss event, failover is completely automatic ensuring application availability. The witness handles quorum only—it doesn't run SCRIBE storage or act as a VM host—while a relaxed SCRIBE data placement spec keeps data protected even in degraded states. Since the witness never sits in the data or workload path, its failure never interrupts applications.
No internet dependency.
Quorum resolves locally, so clusters keep full HA behavior in secure or isolated environments, with no cloud-based witness required.
Zero-touch recovery.
The cluster auto-recovers from node failures without human intervention.
Seamless scalability.
Add nodes later to expand into a 3+-node cluster, without re-architecting your deployment.
Optimized for Edge.
With a minimal footprint, the witness needs just 4GB of RAM, an x86 CPU, and enough storage for HCOS—negligible added power draw or rack space.
If you've previously ruled out a three-node cluster because of cost or footprint, we'd love to show you how a 2-node cluster with a physical witness can deliver the resilience your workloads require. Contact your Scale Computing, Inc. representative to learn more or schedule a demonstration.