SQLite: Past, Present, and Future

• We optimize SQLite for analytical data processing. We identify key bottlenecks in SQLite and discuss the advantages and disadvantages of potential solutions. We integrate our optimizations into SQLite, resulting in overall 4.2X speedup on SSB. • We identify several performance measures specific to embeddable database engines, including library footprint … tinues to be the most widely used database engine in the world, the drastic changes in both hardware capabilities and software demands have exposed SQLite to a unique set of challenges. The expansion of hardware capabilities calls for a deeper evalua- tion into the underlying implementation of SQLite. Notably, SQLite generally does not use multiple threads, which limits its ability … it is likely that certain workloads, particularly those that include complex OLAP, would benefit from multithreading. Furthermore, SQLite’s row-oriented storage format and execution engine are suboptimal for many OLAP operations. In general, SQLite is con- sidered not to be competitive with state-of-the-art OLAP-focused … use cases of embeddable database engines, they are representative of the most common workloads. We identify key bottlenecks in SQLite’s OLAP performance, discuss the tradeoffs of potential so- lutions, and present the performance impact of our optimizations. Finally, we discuss the “footprint” of SQLite: the amount of memory