Международный журнал передовых инноваций, мыслей и идей

Абстрактный

MINER: An Improved Adaptive Join Algorithm

C. Naga Pradeep Kumar, A. Ananda Rao

Adaptive join algorithms were created to overcome the drawbacks of traditional join algorithms in emerging data integration or online aggregation environments. The input relations to adaptive joins are continuously retrieved from remote sources. The main objective for designing these algorithms is to i) start producing the first output tuples as soon as possible ii) produce the remaining results at a fast rate. One of the early adaptive join algorithm Multiple Index Nested-loop Reactive join (MINER) is a multi-way join operator used for joining an arbitrary number of input sources. Here MINER was limited to chain joins. In this paper, MINER is extended to support snowflake joins, where each relation may participate in joins with more than two join attributes. It will improve producing result tuples at a significantly higher rate, while making better use of the available memory.