Scalable Distributed Systems
- FMKe: a Real-World Benchmark for Key-Value Data Stores Proceedings of the 3rd International Workshop on Principles and Practice of Consistency for Distributed Data, April 2017
- Observing the Consistency of Distributed Systems Proceedings of the 15th International Workshop on Erlang, September 2016
- Cure: Strong semantics meets high availability and low latency 36th IEEE International Conference on Distributed Computing Systems, June 2016
- Cure: Strong semantics meets high availability and low latency Research Report, February 2016
- Antidote: the highly-available geo-replicated database with strongest guarantees SyncFree Project, August 2016
- Observing the Consistency of Distributed Systems Erlang 2016, Workshop on Erlang, Sun 18 - Sat 24 September 2016
- Causal consistency in geo-replicated systems Research Training Group Weak Consistency (weacon), July 25th, 2014
Developing highly scalable distributed systems, in particular partitioned and geo-replicated database systems. Due to the limitation described by CAP theorem, geo-replicated systems are nowadays moving towards eventual consistency. With no or less synchronisation such systems are able to provide high availability and scalability. I am interested in enabling stronger semantics in such systems without compromising scalability; using CRDTs, transactions and weak consistency semantics such as eventual consistency and causal consistency.
Related Project : LightKone SyncFree
Related Publications :