Generic Design Patterns for Tunable and High-Performance SSD-based Indexes

Abstract

A number of data-intensive systems require using random hash-based indexes of various forms, e.g., hashtables, Bloom filters, and locality sensitive hash tables. In this paper, we present general SSD optimization techniques that can be used to design a variety of such indexes while ensuring higher performance and easier tunability than specialized state-of-the-art approaches. We leverage two key SSD innovations: a) rearranging the data layout on the SSD to combine multiple read requests into one page read, and b) intelligent request reordering to exploit inherent parallelism in the architecture of SSDs. We build three different indexes using these techniques and conduct extensive studies showing their superior performance and flexibility.