DMCA
Diplomarbeit Distributed Processing of Data Streams in P2P Networks (2005)
BibTeX
@MISC{Scholl05diplomarbeitdistributed,
author = {Tobias Scholl and Erstgutachter Prof and Alfons Kemper and Ph. D and Zweitgutachter Prof and Dr. Winfried Hahn and Betreuer Dipl. -inf and Richard Kuntschke and Prof A. Kemper and Ph. D},
title = {Diplomarbeit Distributed Processing of Data Streams in P2P Networks},
year = {2005}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
In many industrial and scientific cooperations and communities, globally distributed institutions share their data and computational resources to compose socalled virtual organizations. Efficient parallelization, network-aware load-balancing, and distributed processing of heterogeneous data sources are exemplary for arising challenges. Virtual observatories play an integer role to face data explosion and to support distributed research in the fields of astronomy and astrophysics. The appearance of many general issues in these communities is a great incentive to probe further. As amounts of data are rising very fast due to technical advances, traditional publish-subscribe approaches such as data shipping reach their border of feasibility. Disseminating data as streams to fellow researchers or realizing several key applications such as matching processes over a Grid-based Peer-To-Peer infrastructure could gain significant benefits. This thesis introduces an approach for distributed data processing which applies concepts from distributed query processing of persistent data on data streams. It describes further an interface to integrate user-defined operations into a data stream management system and demonstrates how such a system can be monitored and evaluated with a prototype. Contents
Keyphrases
data stream p2p network diplomarbeit distributed processing virtual observatory computational resource data explosion socalled virtual organization technical advance many general issue heterogeneous data source grid-based peer-to-peer infrastructure significant benefit integer role persistent data great incentive network-aware load-balancing data stream management system data processing user-defined operation traditional publish-subscribe approach efficient parallelization several key application institution share query processing scientific cooperation
