Energy proportional datacenter networks

Abstract

Numerous studies have shown that datacenter computers rarely operate at full utilization, leading to a number of proposals for creating servers that are energy proportional with respect to the computation that they are performing. In this paper, we show that as servers themselves become more energy proportional, the datacenter network can become a significant fraction (up to 50%) of cluster power. In this paper we propose several ways to design a high-performance datacenter network whose power consumption is more proportional to the amount of traffic it is moving-that is, we propose energy proportional datacenter networks.We first show that a flattened butterfly topology itself is inherently more power efficient than the other commonly proposed topology for high-performance datacenter networks. We then exploit the characteristics of modern plesiochronous links to adjust their power and performance envelopes dynamically. Using a network simulator, driven by both synthetic workloads and production datacenter traces, we characterize and understand design tradeoffs, and demonstrate an 85% reduction in power -which approaches the ideal energy-proportionality of the network.Our results also demonstrate two challenges for the designers of future network switches: 1) We show that there is a significant power advantage to having independent control of each unidirectional channel comprising a network link, since many traffic patterns show very asymmetric use, and 2) system designers should work to optimize the high-speed channel designs to be more energy efficient by choosing optimal data rate and equalization technology. Given these assumptions, we demonstrate that energy proportional datacenter communication is indeed possible.

Honeyguide: A VM Migration-Aware Network Topology for Saving Energy Consumption in Data Center Networks

“…Abts et al [23] argue that a flattened butterfly topology [14], which is a multi-dimensional direct network, is more power efficient than the other commonly proposed topologies. However, to deploy a flattened butterfly topology, we have to change all the cabling of the existing network topologies.…”

Green Data Center Networks: Challenges and Opportunities

“…The fat-tree architecture delivers high end-to-end bandwidth at a much lower cost and energy consumption [6]. Subsequently proposed successors of the fat-tree, such as VL2 and Portland, also use the Clos structure and commodity network equipment [7]. Kim et al proposed the flattened butterfly DCN architecture using high radix routers [8].…”

“…As the servers and cooling infrastructures are becoming more energy efficient, it is expected that the power consumption share of the networking devices within a data center will increase. It has been projected that the energy consumption of the network component will increase by up to 50% within data centers [7]. Moreover, the bandwidth demands of new network applications are doubling every 18 months [4].…”

“…As our reliance on fossil fuels is expected to rise substantially in the coming decades [2], factors, such as: (a) environmental aspects, (b) high energy costs, (c) increased demand, and (d) economic consequences are driving the need to consider energy efficiency as one of the foremost data center design concerns [7]. The heat dissipation from the networking devices makes a significant contribution to the overall thermal signature of the data center.…”

Network congestion avoidance through Speculative Reservation

“…In these systems, the network bottlenecks usually occur on internal network channels due to under-provisioned global bandwidth. More recently, there has been a shift towards building system networks with full bisection bandwidth as new data center and cloud computing technologies increase the demand for global network communication [3,4,[14][15][16]18]. In networks with ample bisection bandwidth, congestion occurs almost entirely at the edge of the network.…”