OceanStore
John Kubiatowics,
University of California at Berkeley
Recovery
Oriented Computing
Armando Fox, Stanford
University
zFS
Shlomi Dolev, Ben-Guiron
University
During the previous two years a prototype of a self-stabilizing
distributed file system was designed and implemented. The research
project was in collaboration with Ealan A. Henis, Julian Satran,
and Alan Azagury from IBM Haifa Research Lab and resulted in several
publications including a paper in the International Workshop on
Self-Repairing and Self-Configurable Distributed Systems, Osaka,
Japan 2002. The master (now PhD) student Ronen Kat together with
a number of undergraduate students and a software engineer, Albina
Budker, have developed a prototype under the supervision of Professor
Shlomi Dolev. The prototype of the system will be available to interested
researchers.
Practical Selfish Routing
Srinivasan Seshan,
Carnegie Mellon University
In order
to maximize the reliability and performance between nodes of a distributed
application, a site must carefully consider the choice of ISPs and
the selection of routes among those advertised by the chosen ISPs.
Our work involves i) large-scale network measurements to characterize
the expected performance of typical Internet paths, and ii) measurement
and analysis of the performance and reliability benefits of connecting
to multiple ISPs at an enterprise or data center.
|