Processor allocation strategies for multiprocessor database machines

Abstract

In this paper four alternative strategies for assigning processors to queries in multiprocessor database machines are described and evaluated. The results demonstrate that SIMD database machines are indeed a poor design when their performance is compared with that of the three MIMD strategies presented.Also introduced is the application of data-flow machine techniques to the processing of relational algebra queries. A strategy that employs data-flow techniques is shown to be superior to the other strategies decribed by several experiments. Furthermore, if the data-flow query processing strategy is employed, the results indicate that a two-level storage hierarchy (in which relations are paged between a shared data cache and mass storage) does not have a significant impact on performance.

A stream-oriented approach to distributed query processing in a local area network

“…(5) If the manipulated page is the last one of the operand relation, the execution of this relational operation is terminated. (6) If the manipulated page is not the last page, control returns to (2).…”

“…(5) If the compared inner-relation page is the last page and more than one outer-relation page remain to be produced, then the producer node which creates inner-relation pages is initialized and control returns to (2) to compare the subsequent operand page of the outer-relation with all of the inner-relation pages. (This means reproduction of the inner-relation by the producer node.…”

“…To e n h a n c e the p r o c e s s i n g p e r f o r m a n c e of relational operations, many schemes or algorithms have been p r o p o s e d (see e.g., [1,2,6,7,8,13,15]). However, conventional schemes or algorithms are not a l w a y s .…”