2009년 1월 18일 일요일

Partitioned hash-join 참고자료

1 Partitioned Hash-Join

Figure 3.9: Partitioned hash-join algorithm
\includegraphics[height=5cm]{operator/gfx/radixjoin}

The process of performing partitioned hash-join can be divided into two steps. During the first step, both source relations are partitioned into $H$ separate clusters according to the value of a hash function calculated over the join attribute. In the second, corresponding clusters are joined using any traditional join method - Figure 3.9. This solution proves effective when number of clusters is selected in such a way, that they are small enough to fit into cache memories.


Hash Join의 특징 및 사용기준

more..






- 참고자료 -



계속 추가 ..


댓글 없음:

댓글 쓰기