From WiscKey to Bourbon: A Learned Index for Log-Structured Merge Trees

Abstract

We introduce BOURBON, a log-structured merge (LSM) tree that utilizes machine learning to provide fast lookups. We base the design and implementation of BOURBON on empirically-grounded principles that we derive through careful analysis of LSM design. BOURBON employs greedy piecewise linear regression to learn key distributions, enabling fast lookup with minimal computation, and applies a cost-benefit strategy to decide when learning will be worthwhile. Through a series of experiments on both synthetic and real-world datasets, we show that BOURBON improves lookup performance by 1.23x-1.78x as compared to state-of-the-art production LSMs.

References in corpus (7)

• End to End Learning for Self-Driving Cars
• A Comprehensive Analysis of Deep Regression
• On the Computational Efficiency of Training Neural Networks
• RadixSpline: A Single-Pass Learned Index
• Fast Algorithms for Segmented Regression
• A Scalable Learned Index Scheme in Storage Systems
• Piecewise Linear Approximation in Data Streaming: Algorithmic Implementations and Experimental Analysis

Cited by in corpus (4)

• Are Updatable Learned Indexes Ready?
• Constructing and Analyzing the LSM Compaction Design Space (Updated Version)
• LeaFTL: A Learning-Based Flash Translation Layer for Solid-State Drives
• Updatable Learned Indexes Meet Disk-Resident DBMS -- From Evaluations to Design Choices