Network Management and Search in Loosely Coupled Distributed Systems

Loosely coupled distributed systems are those in which nodes can join, leave, or fail at high rates without centralized control and network connections or topology can be highly unstable. Examples include self-organizing Internet overlay networks, wireless networks, and local-area machine clusters without centrally coordinated management. This project studies several problems on network management and search in loosely coupled distributed systems. On network management, we consider scalable network structure management (NSDI'04), random membership subset management (INFOCOM'05, Trans. on Computers), and non-uniform information dissemination (ICNP'03). Exploiting the data popularity skewness in real-world applications, we investigate popularity-driven adaptive distributed data management, specifically for object replication degree customization (EuroSys'08), correlated object placement (ICDCS'08), and Bloom filter hash number customization (PODC'08). In addition to network management, we also study the support for search, an extremely effective utility for many higher-level distributed services. Our study includes peer-to-peer full-text keyword search (WebDB'05) and object search (IPTPS'06).

Project Members:

Ross Camara, undergraduate researcher at University of Rochester (graduated)
John Heidkamp, undergraduate researcher at University of Rochester (graduated)
Pin Lu, Microsoft Inc.
Justin Moore, undergraduate researcher at University of Rochester (graduated)
Amy L. Murphy, Center for Scientific and Technological Research (IRST), Bruno Kessler Foundation (FBK), Italy
Kai Shen, University of Rochester
Ming Zhong, Google Inc.

Contact:

Kai Shen (kshen@cs.rochester.edu)

Support:

This project is supported by the National Science Foundation under Grant No. 0312925. Link to the award abstract.

Note:

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Publications:

Demo and Research Artifact: