Location finding algorithms for distributed systems

One of the problems encountered in distributed systems is how to find the location of the resources need by a computation. In many situations the location may have to be found at run time, when the resource is accessed, thus the efficiency of the location algorithm will affect the performance of the system. In general, the larger the distributed system, the more the number of processors at which a resource may reside at the time it is accessed. The general problem of resource location in distributed systems has not been addressed adequately, and most of the systems have adopted ad hoc solutions without a careful study of the performance of the algorithms used. In this thesis is studied the problem of finding the location of resources in order to get a better understanding of the factors affecting the cost of a location algorithm. This study will make it possible to judge proposed algorithms as well as to come up with new ones, optimized for particular systems. Most distributed systems are based on bus networks that have broadcast and multicast capabilities. The thesis first describes an efficient location method that takes advantage of the multicast capabilities of these networks to reduce the computation cost of resource location finding. Performance results based on a simulation of the scheme are presented, showing that the method is a simple and efficient one. In another multicast scheme for broadcast networks, the thesis considers a system in which no references to resources are stored in the network except where the resource resides. Besides the CPU-cost, response time costs are also considered, and cost formula is found for the scheme. Based on this cost formula, an algorithm is presented to find an optimal sequence of multicast groups to be used in locating a resource. The thesis then considers the communication costs incurred by location finding algorithms in store-and-forward networks. The thesis presents a model of the usage of hint tables and shows how it affects the performance of finding the location resources. (Abstract shortened with permission of author.)