scispace - formally typeset
Proceedings ArticleDOI

Merging traffic to save energy in the enterprise

Reads0
Chats0
TLDR
Algorithms to enable network designers to re-architect their networks using the merge network are presented and a feasibility study is presented using the authors' College of Engineering network as a working example to illustrate how this approach would work and the resultant energy savings of almost 47%.
Abstract
In typical enterprise networks, a large fraction of ports see utilization of less than 5% at peak times and close to zero utilization otherwise. Therefore, the normal architecture of one switch port per end-host is very wasteful because of the need for high port density switches to support numerous end users. In this paper we propose merging traffic from multiple end-hosts and feeding that to small port density switches that can replace the high port density switches. The energy savings from such a redesign are significant. The innovative part of this paper is the design of a low-power Merge network that is used to merge traffic from N incoming links to be fed to K switch ports and for sending traffic from the K-port switch to N links. Further, we present algorithms to enable network designers to re-architect their networks using the merge network, and a feasibility study using our College of Engineering network as a working example to illustrate how this approach would work and the resultant energy savings of almost 47%.

read more

Content maybe subject to copyright    Report

Citations
More filters
Proceedings Article

Traffic merging for energy-efficient datacenter networks

TL;DR: The idea of traffic aggregation, in which low traffic from N links is combined together to create H <; N streams of high traffic, is proposed, which shows that energy-proportional datacenter networks are indeed possible.
Proceedings ArticleDOI

Applying traffic merging to datacenter networks

TL;DR: The results show that energy-proportional datacenter networks are indeed possible and can be accomplished with minimal latency and energy costs while simultaneously allowing us a deterministic way of switching link rates between maximum and minimum.
Journal ArticleDOI

Putting the cart before the horse: merging traffic for energy conservation

TL;DR: This article proposes a different approach to the problem, which involves aggregating traffic from multiple input links prior to feeding them to the switch interfaces, and results are that energy consumption, measured as fraction of interfaces that are awake, scales linearly with load for all loads and the algorithms are fully deterministic yielding zero packet loss.
Journal ArticleDOI

Minimizing Energy Consumption of FatTree Data Center Networks

TL;DR: In this paper, the authors derived formulas for the minimum number of active switches needed in a fat-tree data center network for arbitrary types of loading and derived expressions for the expected traffic loss when these networks are overloaded with external (Internet) traffic.
Journal ArticleDOI

A semantic enhanced Power Budget Calculator for distributed computing using IEEE 802.3az

TL;DR: An energy budget calculator that includes the energy model of 802.3az compliant Ethernet devices and supports the resource management service is devised and a solution for enhancing the calculator by using a semantic energy information system is presented.
References
More filters
Book

Queueing Systems. Volume 1: Theory

TL;DR: The purpose of this document is to summarize the main points of the book written by Leonard Kleinrock, titled, ‘Queueing Systems’, which is about queueing systems.
Proceedings ArticleDOI

Improving the Scalability of Data Center Networks with Traffic-aware Virtual Machine Placement

TL;DR: This paper designs a two-tier approximate algorithm that efficiently solves the VM placement problem for very large problem sizes and shows a significant performance improvement compared to existing general methods that do not take advantage of traffic patterns and data center network characteristics.
Proceedings ArticleDOI

ElasticTree: saving energy in data center networks

TL;DR: This work presents ElasticTree, a network-wide power1 manager, which dynamically adjusts the set of active network elements -- links and switches--to satisfy changing data center traffic loads, and demonstrates that for data center workloads, ElasticTree can save up to 50% of network energy, while maintaining the ability to handle traffic surges.
Proceedings ArticleDOI

Power Awareness in Network Design and Routing

TL;DR: This paper describes the power and associated heat management challenges in today's routers and advocates a broad approach to addressing this problem that includes making power-awareness a primary objective in the design and configuration of networks, and in theDesign and implementation of network protocols.
Related Papers (5)
Trending Questions (1)
Why we don't find port fastethernet0/0 on Catalyst switches?

Therefore, the normal architecture of one switch port per end-host is very wasteful because of the need for high port density switches to support numerous end users.