Fall 2021 Electrical Engineering and Computer Science (EECS) Seminar Series

"Join Ordering Without Statistics"

Asoke Datta,

University of California, Merced

Faculty Host: Professor Florin Rusu

Abstract:

The goal of a database query optimizer is to find an optimal join order, which is crucial for efficient query execution. In order to compute the optimal join order, existing solutions employ extensive data synopses and correlations between table attributes. In this presentation, we introduce Simplicity Simplified (Simpli-Squared), a simple join ordering algorithm that does not use any statistics or cardinality estimates, yet achieves unexpectedly good results. We compare Simpli-Squared with four state-of-the-art query optimization methods on three database systems and eleven different configurations. Experiments on the JOB benchmark in PostgreSQL and MonetDB (CPU accelerated database systems) show that Simpli-Squared achieves runtimes having an increase of up to 16% – and sometimes even a reduction. In MapD (GPU accelerated database systems), Simpli-Squared achieves >18x improvement in runtime compare with MapD default optimizer.

Speaker Bio:

Asoke Datta is a fourth-year Ph.D. student at the University of California Merced working with Dr. Florin Rusu. He received his BSc. degree in computer science from Leading University, Bangladesh in 2014. He worked for Accenture in Dhaka, Bangladesh between 2014-2017 as a system engineer. His research focus is on query optimization in databases and data management systems. The link to his personal webpage is: https://asoke26.github.io/adatta2/