We’re thrilled to announce the 1.0 release of S2, the Scalable Scheduler designed to bring true cluster queues, modern configuration management, and cloud-native capabilities to HPC clusters.
One of the most requested features from customers and large scale users on the Grid Engine mailing lists is true cluster-wide queues, especially for sites migrating from LSF. While Grid Engine 6.0 introduced cluster queues, it lacks critical functionality, particularly queue-level fairshare. S2 addresses this gap by offering semantics similar to LSF queues, making migrations smoother and more intuitive.
Another common pain point with previous-generation job schedulers is the reliance on key-value configuration files. While simple, this format is prone to parsing errors and lacks robust validation. S2 solves this by adopting TOML, a modern, structured configuration language with broad third-party tooling support. This ensures more reliable parsing, easier validation, and seamless templating for complex deployments.
1.0 is just the first step in our multi-year roadmap. With queue-level fairshare already in development and additional performance optimizations on the horizon, we’re committed to delivering groundbreaking enhancements in the coming year. Stay tuned for updates as we continue to evolve S2 to meet the demands of modern HPC environments.
 |
| Hello World |
