IMADA - Department of Mathematics and Computer Science |
Data stream processing has a wide applicability, ranging from environmental monitoring to financial monitoring through computer network management. Such applications often require the processing of continuous queries on active data streams, which arrive in a continuous and massive manner. Traditional DBMSs, which assume data are stored passively and user queries only have to be executed once, are unsuitable to perform such tasks. Hence, many recent research efforts have been devoted to building DSMSs (Data Stream Management Systems) to facilitate the above applications. To provide scalable and reliable stream processing services, building a distributed DSMS is inevitable. In this talk, our recent efforts to enhance the scalability and adaptability of a distributed DSMS will be presented, which address several critical performance issues, such as data stream dissemination, query distribution, query operator placement and ordering. Open problems in this area will also be discussed.
SDU HOME | IMADA HOME | Previous Page Last modified: Mon Jan 14 09:39:03 CET 2008 Joan Boyar (joan@imada.sdu.dk) |