EEvA: Fast Expert-Based Algorithms for Buffer Page Replacement

TL;DR

This paper introduces a new family of expert-based page replacement algorithms that outperform existing methods in buffer management, especially on custom data patterns, with low overhead demonstrated on TPC-C benchmarks.

Contribution

The paper proposes a novel family of buffer page replacement algorithms with theoretical foundations and extensive experimental validation, including implementation in an open-source DB kernel.

Findings

Superior performance on custom data access patterns

Low computational overhead on TPC-C benchmark

Theoretical guarantees for the proposed algorithms

Abstract

Optimal page replacement is an important problem in efficient buffer management. The range of replacement strategies known in the literature varies from simple but efficient FIFO-based algorithms to more accurate but potentially costly methods tailored to specific data access patterns. The principal issue in adopting a pattern-specific replacement logic in a DB buffer manager is to guarantee non-degradation in general high-load regimes. In this paper, we propose a new family of page replacement algorithms for DB buffer manager which demonstrate a superior performance wrt competitors on custom data access patterns and imply a low computational overhead on TPC-C. We provide theoretical foundations and an extensive experimental study on the proposed algorithms which covers synthetic benchmarks and an implementation in an open-source DB kernel evaluated on TPC-C.