Introduction

A key-value store for modern storage devices.

Summary

Modern NVMe solid state drives offer significantly higher bandwidth and lower latency than prior storage devices. Current key-value stores struggle to fully utilize the bandwidth of such devices. This paper presents SplinterDB, a new key-value store explicitly designed for NVMe solid-state-drives.

SplinterDB is designed around a novel data structure (the STBε-tree) that exposes I/O and CPU concurrency and reduces write amplification without sacrificing query performance. STBε-tree combines ideas from log-structured merge trees and Bε-trees to reduce write amplification and CPU costs of compaction. The SplinterDB memtable and cache are designed to be highly concurrent and to reduce cache misses.

On a number of micro- and macro-benchmarks, SplinterDB outperforms RocksDB, a state-of-the-art key-value store, by a factor of 6–10x on insertions and 2–2.6x on point queries, while matching RocksDB on small range queries. Furthermore, SplinterDB reduces write amplification by 2x.

Details

We are hiring engineers to incubate SplinterDB. Apply

Researchers

External Researchers

  • Martin Farach-Colton

Related Publications

Category

  • Active VMW Software Systems Projects

Research Areas

  • Data Structures
  • Databases
  • External-Memory Data Structures
  • Key-Value Stores
  • Write-Optimized Data Structures