Stewart A. Schuster

RAP: an associative processor for data base management

Recent concepts in data base management systems (DBMS) necessitate making the logical view and the physical representation of data distinct from each other. Currently, this requirement has to be realized in the environment of conventional Von Neumann architecture. This creates the need for several levels of indirection for mapping one structure into the other. Also, efficient search mechanisms are needed to handle large data bases within concurrent processing and on-line response limits. The implementation of these requirements results in software complexities and inefficiencies in the following way. Pointer mechanisms for mapping structures and providing fast access paths have to be implemented by software and data. These pointers are extra data requiring extensive overhead in storage, access time, and maintenance.

Performance evaluation of a relational associative processor

A relational associative processor called RAP has been designed to provide hardware support for the use and manipulation of a relational data base. In this paper, we describe the relational operations provided by the RAP hardware, and devise a representative approach to providing the same relational operations with conventional software and hardware. Analytic models are constructed for both RAP and the conventional system. All simplifying assumptions are made to favor the conventional system. The execution times of several of the operations are shown to be improved with RAP by several orders of magnitude.

RAP. 2 - an Associative Processor for data bases

RAP - a Relational Associative Processor - is a back-end or peripheral device to augment a general purpose computer for implementing a data base management system (DBMS). Its architecture is based on the fact that data base operations are inherently set-oriented and that data base addressing is best accomplished through associative reference to achieve high data independence. RAP utilizes these characteristics by combining the features of associative and array processors. Previous publications on RAP have dealt separately with the details of the first version of its architecture [1,2,3,4] language interface [5,6] and performance evaluation [7,8,9,]. This paper provides details on a recently evolved, faster, and more flexible architecture for RAP.

A multi-level relational system

A relation can be conceptually viewed as a table of data. The tables heading defines the relations name, the column headings are the attribute names, and each row corresponds to an n-tuple of data values describing a single entity. The set of values which can be used in a column is called a domain. A relational data base is composed of a set of time varying relations inter-related through common domains.

Query execution and index selection for relational data bases

An algorithm to evaluate primitive Boolean selections over single relations is presented. It will be argued that the algorithm is efficient with respect to the number of relational accesses and with respect to the merging of inverted lists. The algorithms unique quality is its efficiency in evaluating selections over partially inverted relations. A simple cost function is used to drive the algorithm along the most efficient access paths. The cost function can also be used to predict its response time which then forms the basis of a procedure to suboptimize the selection of the domains to be inverted. The domains to be inverted are selected by analyzing, with respect to their costs, a sample of queries. Such a method does away with usual methods of updating usage counters for every domain and relation in the system. In this approach, the selection of a good set of inverted lists is based on the algorithm which uses those lists.

Exploiting parallelism in a Relational Associative Processor

The Relational Associative Processor (RAP) is a special purpose non-numeric back-end processor used in supporting general Data Base Management Systems. In particular, it is ideally suited for supporting a relational data base. The architecture and instruction set of RAP are discussed in this context. It is the purpose of this paper to show that RAP performance can be enhanced considerably by more fully exploiting its parallel nature. It is shown that a greater degree of concurrent activity will result in better overall performance. An operating system executive to support this high level of concurrency is proposed and modelled using simulation techniques. Both the analytic and simulation results support the value of the proposals.

Optimal contiguous area location in geographical parcel configurations

Abstract A definition of optimal area, path, and point sites is proposed in terms of contiguous geographic parcels. This definition is formalized using a linear graph model of a parcel configuration. The graph model is used to develop integer programming formulations of site location problems. The paper deals primarily with the area problem showing that the problem of locating an area can subsume the problems of path and point location.