A query propagation approach to improve CORBA Trading Service scalability

Abstract

Existing CORBA traders, at least most of them, support the core functions of the OMG specification of the Trading Service. We believe that this is useful and can help potential users to use such a service in heterogeneous environments. However we also believe that this is not sufficient because such environments deal with dynamic information and often require scalability (e.g. stock market applications). The CORBA Trading Service uses static information recorded in different components of a trading graph, which reduces its ability to deal with dynamic environments. This paper proposes solutions to the issue of type management and query routing in the context of CORBA Trading Service to improve the scalability and the quality of the results returned by the core trader functions. We propose a query routing mechanism that uses dynamic information recorded within different traders. Some of this information, such as hit factor, is calculated based on the number of offers a remote trader can address and their relative hops away. Finally, we demonstrate that the proposed approach has led to better performance for the core CORBA trader functions.