Abstracts of Published Articles

PERFORMANCE AND PARALLEL PROCESSING TEAM
[Abstracts of published articles]

[1]
D. Andresen, T. Yang, O. Egecioglu, O. Ibarra, and T. Smith, "Scalability Issues for High Performance Digital Libraries on the World Wide Web,'' Proceedings of ADL '96, Forum on Research and Technology Advances in Digital Libraries, IEEE, Washington D.C., May 1996.
We investigate scalability issues involved in developing high performance digital library systems. Our observations and solutions are based on our experience with the Alexandria Digital Library (ADL) testbed under development at UCSB. The current ADL system provides on-line browsing and processing of digitized maps and other geo-spatially mapped data via the World Wide Web (WWW). A primary activity of the ADL system involves computation and disk I/O for accessing compressed multi-resolution images with hierarchical data structures, as well as other duties such as supporting database queries and on-the-fly HTML page generation. Providing multi-resolution image browsing services can reduce network traffic but impose some additional cost at the server. We discuss the necessity of having a multi-processor DL server to match potentially huge demands in simultaneous access requests from the Internet. We have developed a distributed scheduling system for processing DL requests, which actively monitors the usages of CPU, I/O channels and the interconnection network to effectively distribute work across processing units to exploit task and I/O parallelism. We present an experimental study on the performance of our scheme in addressing the scalability issues arising in ADL wavelet processing and file retrieval. Our results indicate that the system delivers good performance on these types of tasks.

[2]
D. Andresen, T. Yang, V. Holmedahl and O. Ibarra, "SWEB: Towards a Scalable WWW Server on MultiComputers,'' Proceedings of the 10th International Parallel Processing Symposium (IPPS'96), Hawaii, April, 1996.
We investigate the issues involved in developing a scalable World Wide Web (WWW) server on a cluster of workstations and parallel machines. The objective is to strengthen the processing capabilities of such a server by utilizing the power of multicomputers to match huge demands in simultaneous access requests from the Internet. We have implemented a system called SWEB on a distributed memory machine, the Meiko CS-2, and networked workstations. The scheduling component of the system actively monitors the usages of CPU, I/O channels and the interconnection network to effectively distribute HTTP requests across processing units to exploit task and I/O parallelism. We present the experimental results on the performance of this system.

[3]
D. Andresen, T. Yang, O. Ibarra "Towards a Scalable Distributed WWW Server on Networked Workstations'' The Journal of Parallel and Distributed Computing (JP DC), 1996.
We investigate the issues involved in developing a scalable World Wide Web (WWW) server on a cluster of workstations and parallel machines. The objective is to strengthen the processing capabilities of such a server by utilizing the power of multicomputers to match huge demands in simultaneous access requests from the Internet. We have implemented a system called SWEB on a distributed memory machine, the Meiko CS-2, and networked workstations. The scheduling component of the system actively monitors the usages of CPU, I/O channels and the interconnection network to effectively distribute HTTP requests across processing units to exploit task and I/O parallelism. We present the experimental results on the performance of this system.

[4]
D. Andresen, T. Yang, D. Watson, A. Poulakidas, "Dynamic Processor Scheduling with Client Resources for Fast Multi-resolution WWW Image Browsing,'' Proceedings of the 11th International Parallel Processing Symposium (IPPS'97), Geneva, April, 1997.
WWW-based Internet information service has grown enormously during the last few years, and major performance bottlenecks have been caused by WWW server and Internet bandwidth inadequacies. Utilizing client-site computing power and also multi-processor support at the server site can substantially improve the system response time. In this paper, we examine the use of scheduling techniques in monitoring and adapting to workload variation at client and server sites for supporting fast WWW image browsing. We provide both analytic and experimental results to identify the impact of system loads and network bandwidth on response times and demonstrate the effectiveness of our scheduling strategy.

[5]
O. H. Ibarra, "Efficient Access and Delivery of Images for the Digital Library,'' International Conference on High Performance Computing, Keynote Address, Trivandrum, India, December 1996.
We give a brief overview of the Alexandria Digital Library (ADL) project at the University of California at Santa Barbara. ADL's emphasis is on spatial data and images. In an image browsing environment such as the ADL's, there is need for a design that allows viewing image subregions in progressively higher resolutions. We describe a storage scheme that accomplishes good image compression, while providing fast image subregion retrieval at various resolutions. This storage scheme is applied on the quantized coefficient matrices obtained from the application of wavelet transforms to images. Our encoding algorithm uses quadtrees to provide good compression and efficient region indexing, coupled with Huffman coding to further improve the compression ratios. We also briefly describe our work on a multiprocessor server that provides scalable WWW access to ADL.

[6]
A. Poulakidas, A. Srinivasan, O.Egecioglu, O. Ibarra, T. Yang, "Experimental Studies on a Compact Storage Scheme for Wavelet-Based Multiresolution Subregion Retrieval,'' Combined Industry, Space and Earth Science Data Compression Workshop, Snowbird, Utah, April 4, 1996.
Wavelet transforms, when combined with quantization and a suitable encoding, can be used to compress images effectively. In order to use them for image library systems, a compact storage scheme for quantized coefficient wavelet data must be developed with a support for fast subregion retrieval. We have designed such a scheme and in this paper we provide experimental studies to demonstrate that it achieves good image compression ratios, while providing a natural indexing mechanism that facilitates fast retrieval of portions of the image at various resolutions.