Section: Research Program

Research rationale

The research of Regal addresses both theoretical and practical issues of Computer Systems, i.e., its goal is a dual expertise in theoretical and experimental research. Our approach is a “virtuous cycle” of algorithm design triggered by issues with real systems, which we prove correct and evaluate theoretically, and then eventually implement and test experimentally.

Regal's major challenges comprise communication, sharing of information, and correct execution in large-scale and/or highly dynamic computer systems. While Regal's historically focused in static distributed systems, since some years ago we have covered a larger spectrum of distributed computer systems: multicore computers, clusters, mobile networks, peer-to-peer systems, cloud computing systems, and other communicating entities such as swarms of robots. This holistic approach allows the handling of related problems at different levels. Among such problems we can highlight communication between cores, consensus, fault detection, scalability, search and diffusion of information, allocation resource, replication and consistency of shared data, dynamic content distribution, and multi-core concurrent algorithms.

Computer Systems is a rapidly evolving domain, with strong interactions with industry and modern computer systems, which are increasingly distributed. Ensuring persistence, availability, and consistency of data in a distributed setting is a major requirement: the system must remain correct despite slow networks, disconnection, crashes, failures, churn, and attacks. Easiness of use, performance, and efficiency are equally fundamental. However, these requirements are somewhat conflicting, and there are many algorithmic and engineering trade-offs, which often depend on specific workloads or usage scenarios. At the same time, years of research in distributed systems are now coming to fruition, and are being used by millions of users of web systems, peer-to-peer systems, gaming and social applications, or cloud computing. These new usages bring new challenges of extreme scalability and adaptation to dynamically-changing conditions, where knowledge of the system state might only be partial and incomplete. Therefore, the scientific challenges of the distributed computing systems listed above are subject to additional trade-offs which include scalability, fault tolerance, dynamics, and virtualization of physical infrastructure. Algorithms designed for traditional distributed systems, such as resource allocation, data storage and placement, and concurrent access to shared data, need to be redefined or revisited in order to work properly under the constraints of these new environments.

In in particular, Regal focuses on three key challenges:

• the adaptation of algorithms to the new dynamics of distributed systems;

• data management on extreme large configurations;

• the adaptation of execution support to new multi-core architectures.

We should emphasize that these challenges are complementary: the two first challenges aim at building new distributed algorithms and strategies for large and dynamic distributed configurations whereas the last one focusses on the scalability of internal OS mechanisms.