Mission Statement

The Dependable and Scalable Software Systems Group at LUH conducts research on all levels of the system software stack, ranging from operating systems to middleware to big data processing platforms and distributed machine learning systems.

Our approach is integrated (looking at multiple layers at every time) and holistic (following the problem through all layers of the system stack). We are particularly interested in exploring and rethinking the boundaries and interfaces of the system software stack, collapsing layers or co-designing new interfaces where necessary to fulfill the common non-functional requirements like low latency, high sustained throughput, frictionless scalability in terms of number of concurrent users or number of instances, and increased energy and cost efficiency.

Our approach is empirical, grounded in reality, by relying real-world or close to real-world workloads for experimentation and benchmarking in realistic settings at scale, and thereby establishing meaningful baselines for assessing different dimensions of improvement.

Our analysis is thorough, we want to identify shortcomings in the state-of-the-art and identify areas that need to be re-envisioned based on technological pulls by emerging applications, or that are likely to get disrupted by upcoming technological pushes from the hardware side.

Our thinking is bold and ambitious, we dare to challenge the established patterns to design, implement, and research novel and disruptive systems that push the boundaries of what system software is able to achieve.

We target scientific publications at the highest international peer-reviewed conference and journal level.

We aim at making systems available to the public as open source project and creating viable ecosystems around our innovation to achieve far-reaching and long-term sustainable impact.

Our goal is to make research data available to the scientific community and make experiments reproducible in the spirit of transparency.