The Hyflow project is developing concurrency control abstractions, protocols, and mechanisms for current and emerging multicore architectures, cluster systems, and geographically distributed systems. A particular focus is to understand what concurrency control abstractions promote high programmability for these architectures, and how to support those programming abstractions with high performance, scalability, and dependability. A closely related focus is to build open-source experimental systems that embody the techniques, as well as incorporate them into existing open-source infrastructures (e.g., Infinispan).
Ongoing efforts are exploring high performance fault-tolerant transactional memory for cluster systems, high performance software transactional memory for multicore architectures, scalable transactional data structures, and scalable transactional protocols for geographically distributed systems.
- "Remote Invalidation: Optimizing the Critical Path of Memory Transactions", A. Hassan, R. Palmieri, and B. Ravindran, 28th IEEE International Symposium on Parallel and Distributed Processing (IPDPS 2014), May 19-23, 2014, Phoenix, Arizona, USA
- "Automated Data Partitioning for Highly Scalable and Strongly Consistent Transactions", A. Turcu, R. Palmieri and B. Ravindran, The 7th ACM International Systems and Storage Conference, SYSTOR 2014, June 10-12, 2012, Haifa, Israel