Theoretical properties of two problems of distribution of interrelated data

Efficient techniques are needed to allocate resources across distributed sites, in order to minimize processing costs and to improve response times for users. This research investigates algorithms for the efficient partitioning of multiple interrelated data objects across the sites within a computer system. Interrelated data objects commonly arise in database systems, and their interrelation presents problems in placement. It is desirable to allocate these data objects so that objects which are related to each other, are in close proximity. Two methods of achieving this is to, perform the allocation so that interrelated data are placed at adjacent nodes, or placed so that the maximum distance between two interrelated data members (the interrelated class diameter), is minimal.