Software engineering and middleware: a roadmap

作者: Wolfgang Emmerich

DOI: 10.1145/336512.336542

关键词:

摘要: The construction of a large class distributed systems can be simplified by leveraging middleware, which is layered between network operating and application components. Middleware resolves heterogeneity, facilitates communication coordination Existing middleware products enable software engineers to build that are across local-area network. State-of-the-art research aims push this boundary towards Internet-scale distribution, adaptive reconfigurable for dependable wireless systems. challenge engineering devise notations, techniques, methods tools system systematically exploit the capabilities deliver. 1 I N T R O D U C Various commercial trends have lead an increasing demand Firstly, number mergers companies continuing increase. different divisions newly merged company deliver unified services their customers usually demands integration IT time available delivery such often so short building new not option therefore existing components integrated into appears as integrating computing facility. Secondly, providing decreasing. Often only achieved if procured off-the-shelf then rather than built from scratch. Components may incompatible requirements hardware platforms; they deployed on hosts, forcing resulting distributed. Finally, Internet provides opportunities offer vast potential customers. In setting, it difficult estimate scalability requirements. Permission make digital or hard copies all part work lbr personal classroom use granted without fee provided made tbr profit advantage bear notice lull citation first page. To copy otherwise, republish, post servers redistribute lists, requires prior specific permission and/or tee. Future Sofware Engineering Limerick Ireland Copyright ACM 2000 1-58113-253-0/00/6...$5.00 An e-commerce site was designed cope with given transactions per day suddenly find itself exposed orders magnitude larger. required cannot centralized client-server architectures but system. Distributed integrate legacy components, thus preserving investment, decrease market, scalable tolerant against failures. caveat, however, truly considerably more client/server This because there multiple points failure in system, need communicate each other through network, complicates opens door security attacks. has been devised order conceal these difficulties much possible; As solve real problem simplify construction, rapidly being adopted industry [6]. meet requirements, know what available, one best suited problems at hand, how used architecture, design implementation principal contribution paper assessment both, state-of-the-practice current state-of-the-art research. Software increasingly Any ignores trend will limited impact. We, therefore, analyze influence should agenda. We argue techniques needed focus non-functional selection middleware. identify architecture produce guide selecting right employing meets set highlight transparent address issue.

参考文章(29)
Gregor Kiczales, Jim des Rivières, Daniel G. Bobrow, The art of metaobject protocol MIT Press. ,(1991)
Grady Booch, James Rumbaugh, Ivar Jacobson, The unified software development process ,(1999)
Richard Schreiber, Len Gilman, Distributed Computing with IBM? MQSeries ,(1996)
Gregor Kiczales, Jim Des Rivieres, Daniel G Bobrow, The Art of the Metaobject Protocol ,(1991)
Nathan Goodman, Philip A. Bernstein, Vassco Hadzilacos, Concurrency Control and Recovery in Database Systems ,(1987)
Gian Pietro Picco, Amy L. Murphy, Gruia-Catalin Roman, A Software Engineering Perspective on Mobility ,(1999)
W. Emmerich, W. Schwarz, A. Finkelstein, Markup meets middleware ieee computer society workshop on future trends of distributed computing systems. pp. 261- 266 ,(1999) , 10.1109/FTDCS.1999.818814
Eugene S. Hudders, CICS: A Guide to Internal Structure ,(1994)