The performance of query execution in distributed data- base systems depends mainly on the efficiency of the distributed algorithms used for executing query operations and the efficiency of the queryprocessing strate...
详细信息
The performance of query execution in distributed data- base systems depends mainly on the efficiency of the distributed algorithms used for executing query operations and the efficiency of the queryprocessing strategies. In this paper, we present and evaluate a queryprocessing strategy which is based on pipelining and data-flow techniques, We develop timing equations for calculating the performance of four join algorithms: 1) rusted block, 2) hash, 3) sort-merge, and 4) pipelined sort-merge. They are used to execute the join operation in a query in distributed fashion and in pipelined fashion. Based on these equations and similar sets of equations developed for other relational algebraic operations, we evaluated the performance of query execution using the different join algorithms. The effects of varying the values of 1) processing time, 2) I/O time, 3) communication time, 4) buffer size, and 5) join selectivity on the performance of the pipelined join algorithms are investigated. The results are compared to the results obtained by employing the same algorithms for executing queries using the distributedprocessing approach which does not exploit the vertical concurrency of the pipelining approach. These results clearly establish the benefits of pipelining. [ABSTRACT FROM AUTHOR]
暂无评论