Big(ger) Sets: decomposed delta CRDT Sets in Riak

TL;DR

This paper introduces a decomposed delta CRDT set implementation in Riak that significantly improves write performance for high-cardinality sets by optimizing delta-replication and disk decomposition, enabling larger sets with maintained consistency.

Contribution

The paper presents a novel approach to CRDT set implementation in Riak that enhances write efficiency and supports larger sets through disk decomposition and optimized delta-replication.

Findings

Supports sets hundreds of times larger than previous Riak sets

Achieves better than O(n) performance for insert and remove operations

Maintains consistency while improving performance for high-cardinality sets

Abstract

CRDT[24] Sets as implemented in Riak[6] perform poorly for writes, both as cardinality grows, and for sets larger than 500KB[25]. Riak users wish to create high cardinality CRDT sets, and expect better than O(n) performance for individual insert and remove operations. By decomposing a CRDT set on disk, and employing delta-replication[2], we can achieve far better performance than just delta replication alone: relative to the size of causal metadata, not the cardinality of the set, and we can support sets that are 100s times the size of Riak sets, while still providing the same level of consistency. There is a trade-off in read performance but we expect it is mitigated by enabling queries on sets.