Bundled References: An Abstraction for Highly-Concurrent Linearizable Range Queries

TL;DR

This paper introduces bundled references, a novel technique for highly concurrent linearizable range queries, improving performance of linked data structures and in-memory databases significantly over existing methods.

Contribution

It proposes bundled references as a new abstraction enabling efficient, linearizable range queries in concurrent linked data structures, with implementations and performance evaluations.

Findings

Achieves up to 3.9x speedup in skip list range queries

Improves binary search tree range query performance by 2.1x

Yields up to 20% performance gain in in-memory database integration

Abstract

We present bundled references, a new building block to provide linearizable range query operations for highly concurrent linked data structures. Bundled references allow range queries to traverse a path through the data structure that is consistent with the target atomic snapshot and is made of the minimal amount of nodes that should be accessed to preserve linearizability. We implement our technique into a skip list, a binary search tree, and a linked list data structure. Our evaluation reveals that in mixed workloads, our design improves upon the state-of-the-art techniques by 3.9x for a skip list and 2.1x for a binary search tree. We also integrate our bundled data structure into the DBx1000 in-memory database, yielding up to 20% gain over the same competitors.