Implementing Dynamic Code Assembly for Client-Based Query Processing

As determined by the applications’ requirements, nonstandard DBMS are usually conceived for client/server architectures [5, 12]. While the server is responsible for general data-management tasks and for precomputing data, application-oriented processing is done in the mainmemory buffer at the client. Since the expressive power of the query language should be available not only for loading/unloading the buffer, but also for more sophisticated processing tasks on the buffer contents, mainmemory based query processing must be supported [6, 14]. A simple navigational buffer interface is no longer satisfactory. Instead, a declarative interface is required to allow querying the buffer contents. Additionally, adequate main-memory indices should be supported to cope with large application buffers.