Online selection of quorum systems for RAMBO reconfiguration

Authors:
Laurent Michel;Martijn Moraal;Alexander Shvartsman;Elaine Sonderegger;Pascal Van Hentenryck
Affiliations:
University of Connecticut, Storrs, CT;University of Connecticut, Storrs, CT;University of Connecticut, Storrs, CT;University of Connecticut, Storrs, CT;Brown University, Providence, RI
Venue:
CP'09 Proceedings of the 15th international conference on Principles and practice of constraint programming
Year:
2009

Citing 5
Cited 1

Optimization by simulated annealing: an experimental evaluation. Part I, graph partitioning

Operations Research
A Majority consensus approach to concurrency control for multiple copy databases

ACM Transactions on Database Systems (TODS)
RAMBO: A Reconfigurable Atomic Memory Service for Dynamic Networks

DISC '02 Proceedings of the 16th International Conference on Distributed Computing
FAB: enterprise storage systems on a shoestring

HOTOS'03 Proceedings of the 9th conference on Hot Topics in Operating Systems - Volume 9
Long-lived Rambo: Trading knowledge for communication

Theoretical Computer Science

Load balancing and almost symmetries for RAMBO quorum hosting

CP'10 Proceedings of the 16th international conference on Principles and practice of constraint programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

RAMBO is the Reconfigurable Atomic Memory for Basic Objects, a formally specified algorithm that implements atomic read/write shared memory in dynamic, rapidly changing networking environments. RAMBO is particularly apt at dealing with volatile environments such as mobile networks. To maintain availability and consistency, even as hosts join, leave, and fail, RAMBO replicates objects and uses reconfigurable quorum systems. As the system dynamically changes, RAMBO installs new quorum configurations. This paper addresses the reconfiguration problem with three approaches based on a finite-domain model, an hybrid master-slave decomposition and a parallel composite to find optimal or near-optimal configurations. Current behaviors of RAMBO participants are observed, gossiped, and used as predictors for future behaviors, with the goal of finding quorum configurations that minimize read and write operation delays without affecting correctness and fault-tolerance properties of the system.