Prof. Lorenzo Alvisi(University of Texas at Austin)
"Reasoning with MAD distributed systems"
Decentralized approaches spanning multiple administrative domains (MAD) are an increasingly effective way to deploy services. Popular examples include peer-to-peer (p2p) applications, content distribution networks, and mesh routing protocols. Cooperation lies at the heart of these services. Yet, when working together is crucial, a natural question is: "What if users stop cooperating?" After all, users in a MAD system are vulnerable not only to the failure of some of the equipment in the system, or to the possibility that some users may behave maliciously, but also to the possibility that users may selfishly refrain from sharing their resources as the protocol would require.
In this setting, it is hard to put a bound on the number of components in the systems that deviate from their correct specification. is it still possible under such circumstances to build systems that not only provide provable guarantees in terms of their safety and liveness properties but also yield practical performance?Bio: Lorenzo Alvisi is a Professor in the Department of Computer Sciences at the University of Texas at Austin and a Visiting Chair Professor at Shanghai Jiao Tong University. Lorenzo holds a Ph.D. (1996) and M.S. (1994) in Computer Science from Cornell University, and a Laurea summa cum laude in Physics from the University of Bologna, Italy. His research interests are in dependable distributed computing. He is a Fellow of the ACM and the recipient of an Alfred P. Sloan Fellowship and an NSF CAREER Award, as well as of several teaching awards. He serves on the editorial boards of the ACM Transactions on Computer Systems (TOCS), ACM Computing Surveys, and Springer's Distributed Computing. In addition to distributed systems, Lorenzo is passionate about western classical music and red Italian motorcycles.
|Zeit:||Mittwoch, 31.10.2012, 10.30 Uhr|
|Ort:||MPI-SWS Gebäude Saarbrücken, Wartburg, 5. Etage|
|Hinweis:||Der Vortrag wird live zum MPI-SWS Gebäude nach Kaiserslautern, Raum 206 übertragen.|