Prioritized Restreaming Algorithms for Balanced Graph Partitioning

TL;DR

This paper introduces a novel family of prioritized restreaming algorithms for balanced graph partitioning, combining ideas from restreaming and label propagation, leading to improved empirical performance on real-world graphs.

Contribution

It presents a new algorithm family that integrates constraint management and priority ordering, outperforming existing scalable partitioning methods.

Findings

Prioritized restreaming algorithms outperform existing methods on real-world graphs.

Dynamic ordering based on ambivalence yields the best cut quality.

Static degree-based ordering performs nearly as well as dynamic ordering.

Abstract

Balanced graph partitioning is a critical step for many large-scale distributed computations with relational data. As graph datasets have grown in size and density, a range of highly-scalable balanced partitioning algorithms have appeared to meet varied demands across different domains. As the starting point for the present work, we observe that two recently introduced families of iterative partitioners---those based on restreaming and those based on balanced label propagation (including Facebook's Social Hash Partitioner)---can be viewed through a common modular framework of design decisions. With the help of this modular perspective, we find that a key combination of design decisions leads to a novel family of algorithms with notably better empirical performance than any existing highly-scalable algorithm on a broad range of real-world graphs. The resulting prioritized restreaming algorithms employ a constraint management strategy based on multiplicative weights, borrowed from the restreaming literature, while adopting notions of priority from balanced label propagation to optimize the ordering of the streaming process. Our experimental results consider a range of stream orders, where a dynamic ordering based on what we call ambivalence is broadly the most performative in terms of the cut quality of the resulting balanced partitions, with a static ordering based on degree being nearly as good.