Bulletin board: a scalable and robust eventually consistent shared memory over a peer-to-peer overlay

Authors:
Vita Bortnikov;Gregory Chockler;Alexey Roytman;Mike Spreitzer
Affiliations:
IBM Research;IBM Research;IBM Research;IBM Research
Venue:
ACM SIGOPS Operating Systems Review
Year:
2010

Citing 9
Cited 0

Exploiting virtual synchrony in distributed systems

SOSP '87 Proceedings of the eleventh ACM Symposium on Operating systems principles
The Transis approach to high availability cluster communication

Communications of the ACM
Horus: a flexible group communication system

Communications of the ACM
Bimodal multicast

ACM Transactions on Computer Systems (TOCS)
Astrolabe: A robust and scalable technology for distributed system monitoring, management, and data mining

ACM Transactions on Computer Systems (TOCS)
GoCast: Gossip-Enhanced Overlay Multicast for Fast and Dependable Group Communication

DSN '05 Proceedings of the 2005 International Conference on Dependable Systems and Networks
A Distributed Service Management Infrastructure for Enterprise Data Centers Based on Peer-to-Peer Technology

SCC '06 Proceedings of the IEEE International Conference on Services Computing
Live Distributed Objects: Enabling the Active Web

IEEE Internet Computing
Quicksilver Scalable Multicast (QSM)

NCA '08 Proceedings of the 2008 Seventh IEEE International Symposium on Network Computing and Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present the design and early experience with a completely new implementation of the Bulletin Board, a topicbased distributed shared memory service employed by commercial-grade application middleware, to achieve robustness and administrative simplicity with adequate latency and costs at the required throughput and scale. To facilitate scalability, only weak consistency is provided. For robustness and ease of use, the implementation is designed in a fully peer-to-peer fashion leveraging the weakly consistent group communication services provided by a semi-structured overlay network. We discuss issues in providing good (while not perfect) stability and reliability at tolerable cost. We address scalability issues, such as supporting large numbers of processes, large subscription spaces, and complex interest patterns. We also consider comprehensive API instrumentation.