Analysis of integration models for service composition

David Liu, Kincho H. Law, Gio Wiederhold
2002 Proceedings of the third international workshop on Software and performance - WOSP '02  
This paper studies service integration infrastructures that support the execution of megaservices -large-scale applications that are composed of autonomous service modules. Integration infrastructures are classified according to their control-flow and data-flow structures. We analyze the effects of data-flows on the performances of the centralized and distributed data-flow models. A mathematical model is built to compare the performances of megaservices. Particularly, aggregated cost and
more » ... ted cost and response time metrics are defined and evaluated. We arrive at the conclusion that the distributed data-flow model is in general superior in performance. We also identify the key system parameters as well as system bottlenecks. The analysis provides recommendations for a few techniques to build high-performance and scalable service integration infrastructures based on the distribution of data-flows. 1. When additional functionalities are needed from other autonomous services, data and requests are again sent out from the megaservice. Due to its simple conceptual model and easy implementation, the 1C1D model is the most common model for current service integration infrastructures. However, megaservices become communication bottlenecks in the 1C1D model. The centralized communication topology makes the 1C1D model unscalable. It is especially problematic in an Internet environment, where the communication links between the megaservice and autonomous services are likely to be of limited bandwidth. On the other hand, the high-speed networks deployed between autonomous services will not be utilized under the 1C1D model. Centralized Data-flow Model Control-flows Data-flows Megaservice M Autonomous Services S1, S2, S3, S4 (b) Centralized Control-flow Distributed Data-flow Model (c) Distributed Control-flow Centralized Data-flow Model (d) Distributed Control-flow Distributed Data-flow Model
doi:10.1145/584369.584393 dblp:conf/wosp/LiuLW02 fatcat:5ygf3slpwjeyflkdt3mpjqrluu