We present the results of performance tests to compare two protocols for three-tier computing using the Windows NT operating system. Three-tier computing features a data server for stored databases (Tier 1), an application server that runs applications (Tier 2), and a simple client program that runs on desktop machines that presents the user interface (Tier 3). Three protocols are available to communicate between Tier 2 and 3: intelligence computer architecture (ICA) with and without data compression, and X Window. We measured the performance of the three protocols in a multi-user environment in which we simulated the workload imposed by typical users. We found that, for Microsoft Office 97 and Lotus Notes applications, the X Window protocol uses approximately twice the network bandwidth of ICA, without compression. We also found that compressed ICA generates roughly one third less network traffic than uncompressed ICA at a cost of 20% of additional processor utilization.

We integrate structured data and text using the unchanged, standard relational model. We started with the premise that a relational system could be used to implement an information retrieval (IR) system. After implementing a prototype to verify that premise, we then began to investigate the performance of a parallel relational database system for this application. We also tested the effect of query reduction on accuracy and found that queries can be reduced prior to their implementation without incurring a significant loss in precision/recall. This reduction also serves to improve run-time performance. After comparing our results to a special purpose IR system, we conclude that the relational model offers scalable performance and includes the ability to integrate structured data and text in a portable fashion. © 1997 John Wiley & Sons, Inc.

Abstract Files are frequently organized into records, and keys are associated with those records. A directory is a list of key‐address pairs, where each pair consists of a key and an address associated with that key. The addresses may be those of records in the file, or of intermediate data, such as inverted lists. A fundamental problem is searching a directory efficiently for known keys, in order to find all addresses associated with those keys. Several problems of finding keys in directories are considered. Methods of searching both ordered and unordered directories are analyzed; directories are classified as they organize unused space: compacted, distributed, or chained. Searching on ranges of key values, rather than exact key matching, is also considered. Finally, design tradeoffs among the classes of directory and searching strategies are developed.