Experimental Study about Join Algorithms of Distributed Database System

Along with the widespread use of distributed database system,execution and optimization of query have become a critical factor that affects the performance of distributed database system.The direct-join algorithm and semi-join algorithm are the most widely applied two query optimization algorithms.Semi-join is generally deemed to have better performance than direct-join.This paper will make an analysis about these two algorithms and make experiments to test them with joining by equivalence in the local area network with big data loaded.At last,based on the result of experiments,after quantizing the superiority of semi-join to direct-join,get a conclusion that semi-join is better than direct-join in actual situations.