SIGMOD2024

DPconv: Super-Polynomially Faster Join Ordering

Mihail Stoian, Andreas Kipf

5 citations

Abstract

We revisit the join ordering problem in query optimization. The standard exact algorithm, DPccp, has a worst-case running time of ๐‘‚ (3 ๐‘› ). This is prohibitively expensive for large queries, which are not that uncommon anymore. We develop a new algorithmic framework based on subset convolution. DPconv achieves a superpolynomial speedup over DPccp, breaking the ๐‘‚ (3 ๐‘› ) time-barrier for the first time. We show that the instantiation of our framework for the ๐ถ max cost function is up to 30x faster than DPccp for large clique queries. CCS CONCEPTS โ€ข Information systems โ†’ Query optimization.