This research proposes that the use of general purpose parallel architectures combined with parallel relational database technology provides a solution for rapid retrieval and analysis of geographic data in a corporate environment. Changes in corporate IT strategies have made the inclusion of parallel architecture viable, vendor support for parallel database software is evident, and research into parallel geographic information system (GIS) functions make commercial versions likely. To that end a substantial parallel GIS/database environment was developed to test the suitability of parallel processing for corporate GIS. The parallel GIS environment brought together parallel relational database management systems and GIS applications on a parallel machine (Meiko Computing Surface) to investigate the ability of the system to support three corporate requirements: the ability to manage and analyse large volumes of data with reasonable response times, the provision of efficient access for GIS to integrated corporate databases, and the successful support of both transaction processing and longer, more complex transactions in the same environment. A series of performance tests were devised and split into two phases. Databases were constructed for each of the testing phases, including a substantial corporate database of some 15 million rows of real data. Experimental results from the performance tests undertaken showed that the parallel system not only successfully supported both the GIS and parallel relational database management system software, it also provided insights into the use and configuration of the environment, the construction of complex queries using GIS/database interfaces, and the ability to support both transaction processing and complex queries in the a single environment. Finally, the implications and applications of the results are discussed.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:663024 |
| Date | January 2000 |
| Creators | Tranter, Mette |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/16936 |
Page generated in 0.0023 seconds