Interleaving a Join Sequence with Semijoins in Distributed Query Processing

In distributed query processing the conventional approach to reduce the amount of data transmission is to rst apply a sequence of semijoins as reducers and then ship the resultant relations to the nal site to carry out the join operations Recently it has been shown that the approach of applying a combination of joins and semijoins as reducers can lead to substantially larger reduction on data transmission required In this paper we develop an e cient heuristic approach to determine an e ective sequence of semijoin and join reducers Semijoins whose execution will reduce the amount of data transmission required to perform a join sequence are termed bene cial semijoins for that join sequence Note that bene cial semijoins include the conventional pro table semijoins and the gainful semijoins that are not pro table themselves but become bene cial due to the inclusion of join reducers This type of dependency between semijoin and join reducers complicates the identi cation of bene cial semijoins and the ordering in the reducer sequence In this paper we rst obtain a sequence of join reducers and map it into a join sequence tree In light of the join sequence tree we derive important properties of bene cial semijoins These properties are then applied to develop an e cient algorithm to determine the bene cial semijoins which can be inserted into the join sequence Examples are also given to illustrate this approach Our results show that the approach of interleaving a join sequence with bene cial semijoins are not only e cient but also e ective in reducing the total amount of data transmission required to process distributed queries Index Terms Distributed query processing gainful semijoins bene cial semijoins join sequence tree reducible set