The Convergence Challenge
A few years back the buzz was all about voice over IP. Now the action is centered on adding video to the networking mix. And when that begins in earnest, observers say that a majority of today's networks will be hard pressed to keep up the pace.
May 1, 2008
Print this page
A growing number of enterprises are looking at a variety of ways to deploy video over IP networks for everything from collaboration between working sites to surveillance and education.
For those few with the infrastructure in place to handle all this extra traffic, integrating access controls, alarm systems and equipment monitoring is also becoming part of the monthly planning sessions.
Besides the “gee whiz” factor of getting real-time access to video over IP, there are a number of practical reasons why companies are taking a serious look at building network capacity to accommodate the collaborative capabilities it can bring.
The initial driver for collapsing all voice, video and data onto one backbone is the fact that it can cut costs, while simplifying overall maintenance and cabling requirements.
“VoIP was the catalyst for the genesis of this migration,” says Boris Kluck, director of collaboration solutions for Telus Communications Inc. in Toronto. “Now organizations are asking what they can layer on [to their networks] to drive out costs and enhance the return they’re making on their investment.”
Cost savings may have triggered the initial interest, but the more compelling drivers behind convergence are much bigger than dollars and cents metrics.
For one, globalization is placing increasing demand on enterprises to be able to communicate and collaborate with branch offices, trading partners or customers at any location, regardless of the time shifts.
“Companies in geographically disparate locations are looking to move from basic communication and conferencing technology to visual collaboration tools to shrink that distance,” says Kluck.
Corporate responsibility, and more specifically the growing focus on a green agenda, is also becoming a strong motivator for enterprises to push convergence to minimize travel time for workers.
Workforce attitudes are also playing into the equation Kluck adds. “In this age of YouTube, people have enjoyed using these technologies at home and expect the same set of tools in the workplace.”
At this point in time, video streaming is quickly gaining ground in internal messaging applications. However, other applications such as IP surveillance are just now coming of age and will soon be part of the network planning mix in the months to come.
Tony Gasson, vice president of Global Unified Communications for AT&T Corp. in London, England, reports that collaboration technologies are experiencing a “strong, strong growth rate” of about 40% per annum.
“It’s clear there has been a huge swing even in the last six months in terms of what collaboration technologies have to offer a business,” he says. “A few years ago it was videoconferencing, but it was expensive and cumbersome to set up. Then audio conferencing democratized the whole concept of virtual business meetings. Now people are seeing the value of putting video, audio and data onto one platform for enterprise deployment. Beyond the cost savings and the green issues, there is a deeper underlying awareness that there can realize tremendous productivity gains.”
Taxing the system
While there is no question that convergence holds significant appeal, the question is, can enterprises leverage the network infrastructure they have in place to support it? For many that answer is probably not.
Ron Groulx, product manager for Fluke Networks Canada in Mississauga, Ont., notes that most organizations are simply not prepared to handle latency sensitive applications like VoIP and IP TV.
Brian Donovan, manager, technology solutions for Graybar Canada Ltd. in Halifax, confirms there are huge pressures to drive more bandwidth and many of today’s infrastructures simply cannot handle it. “DSL and cable modems simply won’t cut it,” he says. “To have all services running over the Internet, you’re pushing the envelope to 100 Mbps, which is what the average user is hooking into. You need speeds up to 1 gig at least. In conjunction with that higher bandwidth, you have to manage it all correctly.”
Jason Riel, data centre product service specialist for Cisco Systems Canada Co. in Toronto, says that From a technology perspective, as you start putting more and more on a network, 10 gig should be your starting point. With larger throughput, you might even have to go to 40 gig.”
One could liken the demand to a tried and true analogy, says Luc Adriaenssens, senior vice president of R&D and technology, CommScope Enterprise Solutions in Richardson, Tex. “The old saying a picture is worth a thousands words applies to voice communications over data. One might say that a video is worth a thousand pictures and a thousand times the bandwidth requirements.”
The implications of this bandwidth demand has a decided impact in multiple facets of network planning and design, including critical areas such as storage and backup functions says Riel. “A lot of technology is now aligned over IP to make networks and is smart enough to route IP traffic according to quality and prioritization requirements.”
Another area of stress on the network manager that goes hand in hand with convergence is security, notes Scott Morrison, vice president, engineering and chief architect at Layer 7 Technologies Inc. in Vancouver. “As more and more applications are added to the network, managers need to focus on a more layered security model both inside and outside the corporate firewall in order to secure the streams of protocols,” he says.
“They have to recognize authorization and access requirements, in order to secure the various streams while maintaining the integrity of the transmission. (Convergence) makes it all that more difficult to stay ahead of the game.”
What you don’t know can hurt you: Industry spokespeople across the board say that before thinking about redesigning, reinstalling or upgrading your network, a critical first step is figuring out what your current infrastructure can do when it comes to adding the video factor.
“You might see the benefits that video will bring, but you also need to understand that the infrastructure needs to be validated and upgraded to make sure it can handle the extra stress put on the networks,” says Adriaenssens. “Whether you’re talking active switches or WAN bandwidth or cabling, you have to ensure that everything is capable of handling the higher requirements.” Amir Hameed, director applications sales, Avaya Canada Corp. in Markham, Ont. confirms that testing is the number one priority before you start layering on those bandwidth-hungry video applications.
“Clearly a network/readiness assessment perspective is absolutely critical. It’s incumbent on managers to do begin with a proper bandwidth analysis based on utilization patterns,” he says.
Establishing if you have the application and hardware infrastructure in place to support the added bandwidth requirements must also take into account specific user requirements such as whether it is to enable virtual agents, mobile workers or on-site staff. A key point to remember when checking on video support capabilities is not forgetting to factor in data and voice requirements and utilization patterns.
The proof of the value of proper assessment techniques says Hameed, lies in the fact that any installations that have experienced major problems usually skipped the first stages of a proper network assessment. “In our experience, where an assessment was done, the results have been brilliant. Where organizations tried to rush or cut corners, they had to go back and troubleshoot because it wasn’t provisioned properly.”
Groulx states the approach to the problem quite simply. “To test is to know.” Testing in the voice/data/video world is not so simple because it assumes a number of added dimensions many network managers are not used to addressing.
“In standard network testing you check for capacity and utilization,” he explains. “Yet many managers are not taking into account timing variables as
sociated with new applications like VoIP and IP TV. These are very latency sensitive applications, and enterprises have no idea what impact they will have on the network and how it will all come together once they are rolled out over IP.”
Stab in the dark
In the brave new converged world, capacity or utilization problems are not always about bandwidth he adds. “The natural reaction to handling slowdowns in the data world is to add bandwidth. But to blindly throw bandwidth or new technology systems is a stab in the dark. Today you need to be more intelligent in understanding network design and traffic flow.”
Latency in a converged world could be a result of timing or a network design that has failed to optimize equipment performance. Therefore, a key concept that every network manager should be aware of today is performance routing, says Groulx. “Technology from Cisco, for example, has the intelligence to understand QoS requirements and can look at specific types of traffic to determine the optimal route based on the potential for jitter or delay.”
New IP and Ethernet test tools are available for testing the effects of latencies on links or circuits over any given time. One of the latest additions to the Fluke portfolio for example is the MetroScope, a device that is specifically designed to provide SLA testing and IP visibility for installing and troubleshooting Ethernet services.
Experts agree that while a relatively low percentage of organizations are actually deploying IP-based video today, it is not a matter of whether they will or not. It is merely a matter of when.
To prepare for the inevitable Adriaenssens says, “Enterprises should be getting prepared for this coming evolution on their networks. They need to make sure they have the right cabling, the right standards and the right monitoring tools in place to handle the reliability demands of the future.”
The importance of reliability cannot be underestimated when it comes to layering applications onto IP. “If a PC crashes today you’re out of commission and productivity drops,” says Adriaenssens. “As more and more applications converge, a system crash has an even greater impact on productivity. To continually improve the reliability of the network you have to ensure that the quality of all components is top notch as well as the operational strategy for administering the network.”
When selecting components, enterprises need to ensure that the products they choose support open standards and interoperability. “Standards are a big consideration,” says Hameed. “Long gone are the days of a single vendor, end-to-end approach. Enterprises are evaluating everything so should expect vendors to adhere to standards to allow equipment to interoperate.”
Interoperability is especially critical for enterprises in the process of transitioning from legacy systems to IP-based infrastructures.
Donovan says that a good part of Graybar’s time is spent working with customers to build strategies to build on connecting legacy equipment to leverage the benefits of IP.
“There are gateway strategies that allow legacy equipment to talk over IP,” he says. “The idea is to enhance applications today while keeping their current infrastructure investment for a few more years until the time comes to replace it.”
Gasson claims that AT&T’s solutions have been built with future enterprise needs in mind. Part of that is providing the ability to manage network load balancing in real time within its AT&T Connect service.
“If you have 50 users, they can connect to an optimizer at a branch office,” he explains. “Rather than having 50 loads going through, you get the equivalent load of one user, which dramatically reduced bandwidth demands.
At this juncture in the network evolution, intelligent infrastructure management is rapidly becoming the de facto standard in best practices.
This means putting added sophistication into such functions as change management for video applications to ensure that transitions can be made smoothly, and problems quickly diagnosed and corrected. “When you add intelligence, performance improves dramatically,” says Adriaenssens.
“The penetration of folks deploying IP video throughout the enterprise is low right now, but it’s clear that it’s the future,” he adds. “Whether it’s this year or the next, everyone knows is coming. Companies now have it in mind when planning their infrastructure, even if they’re not implementing it yet.”
Denise Deveau is a Toronto-based freelance writer. She can be reached at firstname.lastname@example.org.
A key point to remember when checking on video support capabilities is not forgetting to factor in data and voice requirements and utilization patterns.