The increasing use of high-volume scientific data and new performance characteristics of modern hardware platforms need new methods for an efficient distributed data management. Replication and synchronization, concurrency, consistency, and fault-tolerance are the core research fields of this working group. We implement our theoretical results in practical data storage systems such as distributed NoSQL databases, object stores, distributed blockchains, and file systems. Often, modern hardware capabilities such as byte-addressable non-volatile memory or one-sided communication of modern networks offer new challenges and eventually more effective solutions.
In cooperative, interdisciplinary research and development projects with scientific and industrial partners, we find new, efficient solutions or improve, revise, and adapt previous approaches (Scalaris, PacioFS, etc.) in the context of real-world applications. Our research topics include: transactional key/value stores (Scalaris), replicated state machines, conflict-free replicated data types, and distributed file systems and object stores (PacioFS, XtreemFS).