qdina-bench

This repository provides a way to benchmark qDINA, or indeed any other divergent index recommendation algorithm.

Note

This benchmark is merely a comparison of query execution times and not an actual implementation of the TPC-H benchmark (which reports query-per-hour performance, power, and throughput metrics). For an implementation of TPC-H for divergent indexing, see this repository.

This benchmark implements three phases, any or all of which may be run:

• Generation of TPC-H/DS table data and queries
• Loading of generated table data into PostgreSQL databases
• Execution of the benchmark, which creates recommended secondary indexes and then measures query execution times

Installation

Prerequisites

This benchmark is built on Python 3.12, though other versions may be compatible. Download the desired runkit(s) from the TPC website. TPC-H requires gcc, while TPC-DS requires more packages:

sudo apt-get install gcc make flex bison byacc git gcc-9

To run the benchmark, a cluster of PostgreSQL databases must be instantiated and connection permissions configured, etc. This benchmark was written for Postgres 17.