What is a fault-tolerant cluster

A fault-tolerant cluster is two or more servers combined so that, if one of them fails, the protected virtual machines and applications continue running on the standby server without data loss and with minimal interruption. The Russian HyperSphere software based on KVM virtualisation is installed on two servers and moves virtual machines to the standby node in real time, replacing VMware vSphere and Stratus everRun.

Key facts

  • Two or more servers that stand in for each other without data loss.
  • Memory and disk state of the virtual machine is synchronised between nodes.
  • Practically no interruption on failure, unlike a high-availability cluster.
  • Needed for SCADA, control system and CII servers.
  • Russian HyperSphere software on KVM replaces VMware vSphere and Stratus everRun.

How it works

Applications — SCADA servers, historians, databases, engineering stations — run inside virtual machines. The cluster synchronises the memory and disk state of the protected machine between two servers; on a hardware, power or network failure the machine continues on the second node automatically, without administrator intervention. Management is through a web interface, with monitoring of node status and replication.

Fault tolerance and high availability

A high-availability cluster restarts the virtual machine on another node after a failure — the interruption lasts minutes. A continuously available fault-tolerant cluster keeps the machine state synchronous on both nodes, so there is practically no interruption. The operator level of control systems and data acquisition systems need the second option.

Where it is used

Control system and SCADA servers, telemetry systems, database and corporate application servers that must run continuously, critical infrastructure objects where downtime means disruption of the process. Russian software from the national register replaces imported virtualisation platforms that are no longer supported.

Where it appears on this site

A fault-tolerant cluster on HyperSphere software (developed by STR LLC, partnership agreement since 5 August 2025): installation on two servers, real-time transfer of virtual machines, KVM virtualisation; the company’s specialists are certified in cluster installation and administration; demonstration on request.

Questions and answers

How does a fault-tolerant cluster differ from a high-availability cluster?

A high-availability cluster restarts the virtual machine on another node after a failure, with an interruption of minutes. A fault-tolerant cluster keeps the machine state synchronised on two nodes all the time, so work continues with practically no interruption. SCADA and control system servers need the second option.

Can business applications and databases run in the cluster?

Yes: any server application runs in the protected virtual machines — 1C:Enterprise servers, database servers, application and terminal servers. The cluster protects them against hardware, power and network failures without changes to the applications themselves.

What can replace VMware vSphere and Stratus everRun?

The Russian HyperSphere software based on KVM virtualisation: it is installed on two servers and transfers virtual machines to the standby node in real time. The company’s specialists are certified in its installation and administration — see fault-tolerant cluster on HyperSphere software.

Related terms

Updated: 2026-09-23

CallSend a request