The security of the SC//HyperCore™ HyperVisor begins long before software reaches your clusters. Scale Computing builds its engineering team the same way it builds the product. The team uses a hand-picked set of open-source packages. But full-time Scale Computing employees write all software that ships as part of the SC//HyperCore software. This page explains how that model, its CI/CD pipelines, and the testing program give every release a trusted supply chain.
Trusted software
The SC//HyperCore software is the foundation of SC//Platform™. It is the hypervisor for SC//HyperCore clusters. The software bundles a chosen mix of adapted open-source parts and in-house smart software into a simple, appliance-style operating system. Built on KVM, it works directly with the Scale Computing Reliable Independent Block Engine (SCRIBE) storage layer. So the same software stack manages compute, storage, and traffic within the cluster.
The system monitors and heals itself in almost all cases. As a result, it shows far less surface to admins, and to would-be attackers, than a traditional virtualization stack does.
Scale Computing uses its own development and testing process. It combines Scrum and Kanban practices to keep several branches of code for current and upcoming releases. With this branching model, engineering and product teams can release SC//HyperCore software updates without rushing. It also makes sure that each release weighs features and security together.
Scale Computing cryptographically signs all customer-facing software updates and patches to prove they are authentic and intact. The CI/CD pipelines that build and publish those updates have strict security controls to prevent unauthorized access and tampering. This supports a trusted supply chain for everything delivered to your clusters.
Automated testing
A dedicated quality assurance team and a large automated testing program are key to keeping the software stable and secure. They also let the development team stay focused on the product. Hundreds of thousands of system and unit tests run each week against builds in progress. The test suite itself is always being watched, reviewed, and improved.
Every release of every Scale Computing product goes through functional and performance tests. It also goes through a set list of automated and manual security tests, including:
- Port scans to confirm that only intended services are reachable on cluster nodes.
- Targeted security scans against the management interfaces and APIs.
- Penetration tests against the platform and its supporting services.
External penetration testing
Beyond internal testing, Scale Computing hires expert third parties to run regular external penetration tests. These cover SC//Fleet™ Manager and the public servers that support SC//HyperCore remote support and software updates. These independent tests check the in-house defenses and bring new risks to light. They also give ongoing proof that the platform stays resilient as threats change. Findings from outside tests feed back into the same development, testing, and incident response processes used for any other security work.
In-house development, signed updates, thorough automated testing, and independent penetration testing work together. They let Scale Computing ship new features and security patches with confidence. They also let you apply them to your clusters with confidence.