AllConcur: Leaderless Concurrent Atomic Broadcast (Extended Version)
Abstract
Many distributed systems require coordination between the components involved. With the steady growth of such systems, the probability of failures increases, which necessitates scalable fault-tolerant agreement protocols. The most common practical agreement protocol, for such scenarios, is leader-based atomic broadcast. In this work, we propose AllConcur, a distributed system that provides agreement through a leaderless concurrent atomic broadcast algorithm, thus, not suffering from the bottleneck of a central coordinator. In AllConcur, all components exchange messages concurrently through a logical overlay network that employs early termination to minimize the agreement latency. Our implementation of AllConcur supports standard sockets-based TCP as well as high-performance InfiniBand Verbs communications. AllConcur can handle up to 135 million requests per second and achieves 17x higher throughput than today's standard leader-based protocols, such as Libpaxos. Thus, AllConcur is highly competitive with regard to existing solutions and, due to its decentralized approach, enables hitherto unattainable system designs in a variety of fields.
Cite
@article{arxiv.1608.05866,
title = {AllConcur: Leaderless Concurrent Atomic Broadcast (Extended Version)},
author = {Marius Poke and Torsten Hoefler and Colin W. Glass},
journal= {arXiv preprint arXiv:1608.05866},
year = {2017}
}
Comments
Overview: 18 pages, 7 sections, 10 figures, 3 tables. Modifications from previous version: added Figure 4; added, in Section 4.4, a paragraph describing the construction of Gs digraphs; added Section 4.5, a theoretical comparison between AllConcur and leader-based agreement; added, in Section 5, a comparison to unreliable agreement; rephrased several paragraphs to improve clarity