TY - GEN
T1 - Optimal deployment of eventually-serializable data services
AU - Michel, Laurent
AU - Shvartsman, Alexander
AU - Sonderegger, Eldine
AU - Van Hentenryck, Pascal
PY - 2008
Y1 - 2008
N2 - Replication is a fundamental technique for increasing throughput and achieving fault tolerance in distributed data services. However, its implementation may induce significant communication costs to maintain consistency between the replicas. Eventually-Serializable Data Service (ESDS) has been proposed to reduce these costs and enable fast operations on data, while still providing guarantees that the replicated data will eventually be consistent. This paper reconsiders the deployment phase of ESDS, in which a particular implementation of communicating software components must be mapped onto a physical architecture. This deployment aims at minimizing the overall communication costs, while satisfying the constraints imposed by the protocol. Both MIP and CP models are presented and applied to realistic ESDS instances. The experimental results indicate that both models can find optimal solutions and prove optimality. The CP model, however, provides orders of magnitude improvements in efficiency. The limitations of the MIP model and the critical aspects of the CP model are discussed. Symmetry breaking and parallel computing are also shown to bring significant benefits.
AB - Replication is a fundamental technique for increasing throughput and achieving fault tolerance in distributed data services. However, its implementation may induce significant communication costs to maintain consistency between the replicas. Eventually-Serializable Data Service (ESDS) has been proposed to reduce these costs and enable fast operations on data, while still providing guarantees that the replicated data will eventually be consistent. This paper reconsiders the deployment phase of ESDS, in which a particular implementation of communicating software components must be mapped onto a physical architecture. This deployment aims at minimizing the overall communication costs, while satisfying the constraints imposed by the protocol. Both MIP and CP models are presented and applied to realistic ESDS instances. The experimental results indicate that both models can find optimal solutions and prove optimality. The CP model, however, provides orders of magnitude improvements in efficiency. The limitations of the MIP model and the critical aspects of the CP model are discussed. Symmetry breaking and parallel computing are also shown to bring significant benefits.
UR - http://www.scopus.com/inward/record.url?scp=44649129397&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=44649129397&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-68155-7_16
DO - 10.1007/978-3-540-68155-7_16
M3 - Conference contribution
AN - SCOPUS:44649129397
SN - 354068154X
SN - 9783540681540
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 188
EP - 202
BT - Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems - 5th International Conference, CPAIOR 2008, Proceedings
T2 - 5th International Conference on Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, CPAIOR 2008
Y2 - 20 May 2008 through 23 May 2008
ER -