Optimization Models for Configuring Distributed Computer Systems

This paper develops models for designing the architecture of distributed computing systems of the type used to support corporate management and control activities. The data for the design process consist of two major data sets describing the inputs and outputs to the information system and the relationships between them. The model accepts a listing of the relevant data sources in the organization, their physical locations, the number and rate of transactions originating from those locations and the amount of data that have to be transferred before the transactions can be updated, the list of reports that have to be generated by the system, their generation frequency and their physical distribution of those reports to end users. For each report type, the sources of data needed to generate it and the volume of data that have to be transferred for each run.