Consolidating flows with implicit deadlines for energy-proportional data center networks

To reduce the energy consumption of a large number of network devices in a data center, energy-efficient schemes use various heuristics to consolidate traffic to fewer switches. Most of these works, however, ignore the flow-level performance, which is one of the most critical requirements in production data centers. Hence, flow rate allocation should be considered together with flow path selection to guarantee flow-level performance and in the meantime save energy of network devices. For this reason, we present a framework to ensure that the energy consumption for data center network (DCN) is proportional to the traffic and to guarantee the flow-level performance. Our solution consists of two components: (i) flow rate allocation to meet flows' deadlines and (ii) flow path selection to use fewer switches. We compare our framework with existing techniques under synthetic traffic patterns. Results show that our framework could save, on average, 20% of network energy than the always-on baseline, while maintaining the better flow-level performance, and achieving good running time and fault tolerance simultaneously.