You are here


Multi-objective Traffic Engineering for Data Center Networks

Data centers with huge logical storage capacity are now base of many Internet and cloud computing services. Deploying by switched Ethernet networks, data centers are mostly dominated by different variants of spanning tree protocol. The need of a better use of limited spanning tree links is required to find solutions for the load balancing problem in data centers. In this paper, our different Local Search approaches for generating a good set of spanning trees in data centers with many VLANs are evaluated by different multi-criteria assessment methods.

Traffic Engineering for Multiple Spanning Tree Protocol in Large Data Centers

The size of the capacity of data centers have been growing significantly during the last years. Most data centers rely on switched Ethernet networks. A drawback of the Ethernet technology is that it relies on the spanning tree protocol (or variants of it) to select the links that are used to forward packets inside the data center. In this paper we propose a Constrained-Based Local Search optimization scheme that is able to efficiently compute the optimum spanning tree in large data center networks.

The LS(Graph) library

The LS(Graph) is a library developed on top of the Comet programming language for modeling and solving Constrained Optimum Trees and Constrained Optimum Paths problems on graphs by local search.

The Aeon Library

Aeon is a library for the Comet Programming language aimed at simplifying the expression and resolution of scheduling problems.

Dynadec Training

Dynadec Training (June 4, 2009) given at JFPC-09.

Data and results for the Steel Mill Slab Problem

The Steel Mill Slab problem is the problem 38 of CSPLib. The steel mill slab problem is to assign colored sized orders to slabs of different capacities such that the total loss is minimized and such that at most two differentz colors are present in each slabs.

Aeon Benchmark Results

Annexe to the article “Aeon: Synthesizing Scheduling Algorithms from High-Level Models” with detailed results on several benchmarks.