Wednesday, March 23, 2011

The Advent of OpenFlow Protocols

The Open Networking Foundation (ONF) has just been organized to create protocols that would make it possible for firms to control the processing of Internet transactions on switches and routers. ONF includes leading firms involved in Internet networking, such as Broadcom, Brocade, Ciena, Cisco, Citrix, Dell, Deutsche Telekom, Ericsson, Facebook, Force10, Google, Hewlett-Packard, I.B.M., Juniper, Marvell, Microsoft, NEC, Netgear, NTT, Riverbed Technology, Verizon, VMWare and Yahoo. *

The new ONF protocols, named “Open Flow” will radically change how communication networks will operate in the future. This protocol will become an industry standard as an add-on feature to the existing IPv4 and IPv6 protocols. Open Flow code will be also embedded in network controllers, switches and routers. The first router that uses OpenFlow is already in place. Prototype installations have been in place at over dozen universities since 2008.

The objective of ONF is to make networks programmable in much the same way that individual computers can be programmed to perform specific tasks.  This represent a major departure form the current approach in which the Internet switches and routers are pre-defined so that they cannot be modified to accommodate dynamic fluctuations as the traffic on networks keeps changing. OpenFlow focuses on controlling how packets are forwarded through network switches and routers.

In the past one of the key components of any system could not be programmed. That was the network that connected computing nodes. Under OpenFlow it will now be possible to customize networks to the applications that are actually being run.

OpenFlow protocols and associated software should open up hardware and software systems that control the flow of Internet data packets, systems that have until now been closed and vendor proprietary. This will cause a new round of innovation that will be focused principally upon the emerging computing systems, known as cloud computers, that require a variety of network services that currently are not available.

For instance, OpenFlow will permit setting up on-demand “express lanes” for voice and data traffic that is mission critical. Software will allow combining several fiber optic backbones temporarily for particularly heavy information loads and then have circuits automatically separate when a data rush hour is over. Another use of OpenFlow will be load balancing across an entire network, so that diverse data centers will be able to shuttle the workload so that performance does not deteriorate.

OpenFlow will be an open interface for remotely controlling the forwarding tables in network switches, routers, and access points. Based on such capabilities user firms will be able to build networks with a much wider scope, especially involving wireless communications. For example, OpenFlow will enable more secure default fail-overs, wireless networks with smooth handoffs, scalable data centers, host mobility, more energy-efficient allocation of resources and ready deployment of improvised new networks.

SUMMARY
OpenFlow warrants immediate attention even though large-scale implementation may be three to five years ahead. However, it will alter the architecture of networks, such as GIG, to a significant extent. The programmability of networks will change the role of the GIG from just a transmission medium to a component that becomes an active part of the design for DoD. Such planning should be starting soon. OpenFlow equipment will have to be provided for any acquisitions that will have a life of well over ten years.


* http://www.nytimes.com/2011/03/22/technology/internet/22internet.html?_r=1&partner=rss&emc=rss

No comments:

Post a Comment

For comments please e-mail pstrassm@gmu.edu