IEEE Communications Magazine · 2013 · 16 citations · 12 references
ComputingComputer ScienceComputer SystemsCloud Performance OptimizationComputer NetworksInterconnection NetworksFat-tree NetworksLoad BalancingResource SchedulingRoutingRouting AlgorithmsData Center NetworkData Center Network ArchitectureData Center Network TopologyData Center NetworkingLoad Balancing AlgorithmsThroughputTraffic Load BalancingTraffic Pattern Analysis
In this article, we analyze and compare different data center environments in terms of the maximal throughput they offer to end nodes. In particular, we examine to which extent balancing can increase the throughput guaranteed for an arbitrary traffic pattern. We consider and compare two-phase balancing algorithms which are optimized assuming different underlying routing algorithms. It will be shown that optimized load balancing significantly improves the maximal loss-free throughput. A suite of routing algorithms is examined on three cost-effective data center topologies: fat-tree, dragonfly, and flattened butterfly.
12
Albert Greenberg, James R. Hamilton, Navendu Jain et al. · 2009 · 2.1K citations
A scalable, commodity data center network architecture
Mohammad Al-Fares, Alexander Loukissas, Amin Vahdat · 2008 · 1.7K citations · Full text
Hedera: dynamic flow scheduling for data center networks
Mohammad Al-Fares, Sivasankar Radhakrishnan, Barath Raghavan et al. · 2010 · 1.5K citations
Chuanxiong Guo, Guohan Lu, Dan Li et al. · 2009 · 1.4K citations · Full text
Improving datacenter performance and robustness with multipath TCP
Costin Raiciu, Sébastien Barré, Christopher Pluntke et al. · 2011 · 679 citations