Log in to track your reading progress.
Overview
Discover why parallel computing offers an ideal foundation for enterprise-scale data warehousing - and how to make the right strategic and tactical decisions about these critical technologies. What are the right questions to ask? What aren't the vendors telling you? How do benchmarks relate to real-world results? Parallel Systems in the Data Warehouse evaluates all three major parallel hardware architectures, comparing scalability, performance and sustainable I/O bandwidth: symmetric multiprocessors (SMP), distributed memory (DM) or "massively parallel" machines, and distributed shared memory (DSM) systems ("non-uniform memory access" machines). Consistently high throughput is critical to the success of data warehousing - and traditional techniques like normalization and indexing aren't enough anymore. Learn smarter table partitioning strategies, star schema and other powerful techniques to fully exploit the inherent scalability and high I/O bandwidth of parallel hardware.Editorials
Booknews
Examines the use of parallel computing in the context of an enterprise-scale data warehouse. The text evaluates the three major parallel hardware architectures: symmetric multiprocessors, distributed memory machines, and distributed shared memory systems. Also looks at software architecture for parallel database systems, data partitioning and utilities, data mining on parallel systems, benchmarking and implementation, and on-line analytical processing. Includes four case studies of hardware/software combinations currently being used in the industry. Annotation c. by Book News, Inc., Portland, Or.Book Details
Published
November 3, 1997
Publisher
Prentice Hall
Pages
416
Format
Hardcover
ISBN
9780136806042