On Improvement of the Volcano Search and Optimization Strategy

The ever-increasing load on databases dictates that queries do not need to be processed one by one. Multi-query optimization seeks to optimize queries grouped in batches instead of one by one. Multi-query optimizers aim at identifying inter and intra query similarities to bring up sharing of common sub-expressions and hence saving computer resources like time, processor cycles and memory. Of course, the searching takes some resources but so long as the saved resources are more than those used, there is a global benefit. Since queries are random and from different sources, similarities are not guaranteed but since they are addressed to the same schema, it is likely. The search strategy need to be intelligent such that it continues only when there is a high probability of a sharing (hence resource saving) opportunity. We present a search strategy that assembles the queries in an order such that the benefits are high, that detects null sharing cases and therefore terminates the similar sub-expressions’ search as well as removing sub-expressions which already exist else where so as to reduce subsequent searching procedures for a global advantage.

AMS Subject Classification:

68M20, 68P20, 68Q85