December 2007 Archives

In my last blog, I talked about disconnecting my desktop phone. Here’s my 3 reasons why this is the exception rather than the rule in today’s world.

#1 Ease of use: familiar look and feel of a phone, including use of a handset, hands-free operation, and numeric key pad. IP phones with multi-line (and even color bit-mapped displays) and USB ports have taken this to new levels. IP phones can support corporate dashboards; corporate and departmental directories with click-to-call; conference managers that simplify chairman controls and enhance security; push-to-talk capabilities; zone paging that speed up communications; and visual voicemail that accelerates voicemail handling. They can support alerts, such as security alerts, weather alerts, IT alerts, travel advisories and company announcements. Finally, the phone, whether wired, cordless or wireless, can be converged with the PC to provide a unified experience in a unified communications context.

#2 Reliability/quality: voice quality impacts during PC background activities, booting in the morning, restarting during the day (e.g. due to security patches) and lack of operation during power failures, all contribute to the perception that ‘I want/need a phone to do my job”. While the PC platform is getting better every year, we’re not there for many environments! That said, the PC is a critical business tool- answering the phone while your PC is not available does not necessarily allow you to conduct business in any meaningful way. This is a reality of business in the 21st century!

#3 Cost: The TCO for a traditional phone is in the range of a few dollars/day, orders of magnitude less than a PC. Basic SIP phones have brought the cost points close to those of analog sets. In addition, MACs and wiring costs come down with IP telephony. If you have a PC, isn’t the incremental cost for IP telephony close to zero? Yes but what about the cost of headsets which have a relatively limited life expectancy- my experience is a couple of years max!

So the question is generally less about ‘should I invest in phones?’ and more about what’s the right functionality/price, given unified communications, PC functionality and most importantly job roles and the people that do these jobs.

I have always had a desk phone over my 36 year career. Until last Friday!

Firstly, a VT100 dumb terminal (sort of a very thin client running at 1200 bps) and a phone; then an IBM 3270 running off of a mainframe and a phone; then a Mac and a phone; then a PC and a phone; and, over the last number of years, a laptop docking station, monitor and a phone.

My office phone has been a constant companion, and kept me in touch some 300,000 times over those years.

Last Friday, I unplugged my desk phone!

On the road and at home I use my laptop equipped with a UC client and headset. The primary use of my desk phone was as a speakerphone (I have a closed office). I got an LG-Nortel USB phone 8501 (which was designed to work with the Microsoft Office Communicator) and plugged it into my docking station. The 8501 includes a handset and a speakerphone but no keypad; it’s not really a phone, since I dial and accept calls on my UC client, leveraging my personal and corporate directory and click to call capabilities. It does the job nicely.

If I get a call before I start up my laptop (or if I’m in the process of rebooting it), then I have set up my UC client to ring my cell phone, so no missed calls. Otherwise, it’s business as usual: my office number is unchanged and my unified messaging works as before. And I’m ready for the future: 100% SIP-based.

This isn't for everyone, but sure works for me.

One drawback. Ask me to dial 1-800-4Nortel and I have to flip open my cell phone to help me convert this to a number I can dial;)

Let’s look a little further out from the Gartner Enterprise Networking Summit. I took back three messages:

1) Dr. Lippman of MIT said during his keynote: “Technology doesn't change things; society does. It's out of our hands". "It's not about products but about architectures, tools and enablers". This certainly aligns with our communications enablement multi-vendor, multi-domain SOA-based framework that was announced in November with IBM.

2) Gartner spent a lot of time talking to WYNIWYG (What You Need Is What You Get). WYNIWYG unleashes applications and communications through a context driven architecture. To deliver on this promise, solutions have to provide orchestration in real-time, and be multi-vendor and multi-domain. I might not have come up with this acronym, but wholeheartedly agree with the importance of context and relationships.

3) Gartner predicted that ‘Voice as an Application’ will be a reality in a relatively short three years, and will effect 20% of IT budgets. That’s our bet too in partnering deeply with MSFT and IBM.

These may not impact your budget in 2008 but certainly gives one lots to think about. Your thoughts?

The recurring conference theme: the job of IT and how we approach this challenge/opportunity is fundamentally changing. These challenges and opportunities arise from a megatrend, we call Hyperconnectivity.

My personal highpoint: describing our joint UC solutions to a room filled to capacity with MSFT customers and being able to address the needs of both Nortel voice and non-Nortel voice enterprises with the broadest UC solution in the industry. And then having Mike Lucas of Indiana U describe his deployment of the MSFT-Nortel UC solution on a very large scale.

Most direct recommendation from Gartner: "Now is the time to invest in UC". No single vendor can do it all, so partnership are key. In fact, you have probably written your last VoIP RFP!

Most evident truth: There is a direct correlation between how much enterprises pay and how dependent they are on Cisco. Mark Fabbi of Gartner gave an example of two brokerage banks, who each spent $20-30M on L2-3 solutions. One had a Cisco-only network and had 40% discounts; the other had a multi-vendor approach and got 60%. We see this everyday.

Biggest red flag: Gartner flagged Cisco’s network centric architecture as a potential impediment to communications enablement of business processes. This is in contrast to our software-centric approach.

Weirdest travel moment: my flight out of Nashville was cancelled due to fog in DC, half the people left and then it was announced “if the conditions don’t change, we are leaving in 30 minutes” and we did…. Absurd!

Feel free to comment on any of these.

One of the greatest impacts of Hyperconnectivity is in the security area.

Multiple approaches to security enforcement should be used in different parts of the network, operating under enterprise-wide policies. This ‘layered defense’ is further bolstered by adopting an open-security philosophy that embraces a security ecosystem leveraging security leaders such as Symantec, Check Point and Sourcefire.

I recently spoke to an Israeli financial institution that was experiencing 25,000 attacks per day (not a typo!). It has adopted a layered multi-vendor approach to security, including using firewall technologies from different vendors in different parts of the network. In this way, it has eliminated single points of perimeter security failure in the network.

Layered defense in a hyperconnected networking environment must provide highly scalable capabilities, including pervasive end-point security with integration with Microsoft NAP, media security to protect critical multimedia data in transit, virtualization and accelerated perimeter security to logically separate different forms of traffic over a converged IP network, and core network security incorporating dynamic threat protection and real-time device vulnerability fingerprinting.

Security 101 says it all starts with the need to continually revisit enterprise security policies and people processes associated with them. Remember, your security decisions today are setting the foundation for security in the hyperconnected world.

When was the last time you took a critical look at your security policy?

Hyperconnectivity will make the enterprise more dependent on its network than ever before. Real-time UC, security, building management, telemetry, asset tracking and hazmat management will drive the need for real-time reliable networks with extremely fast restoration times to avoid impacting these applications when failures occur.

In November Cisco introduced its Virtual Switching System (VSS), which they claim to be 'innovative'. The good news is that Cisco finally acknowledges that their legacy approaches (e.g. a combination of multiple per VLAN fast spanning trees and IP routing techniques) just can't meet the needs of the hyperconnected enterprise: too complex, too ineffective and too slow for real-time apps.

So is this a leap-frog strategy or a me-too strategy? I would suggest it's an also-ran strategy. Let me explain.

Did you know that we have had switch clustering for 6 years (you may know this as Split Multi Link Trunking- SMLT), meeting the needs of demanding customers such as the NY Times, Mumbai International Airport and the Vancouver Olympic Committee? Cisco innovation claim is hollow!

Did you know Cisco's VSS requires high-end I/O modules and 10GigE inter-switch links, and comes with a price tag in the $200K range? Real-time reliability is not just a core network requirement of enterprises with deep pockets! Resiliency should be end to end and available to all. Our approach spans our entire portfolio of Ethernet modular and stackable switches and operates over a range of links speeds.

Did you know that VSS provides double density with half the brains? In technical terms, two switches look like one, but there is only a single active Supervisor/CPU Module. Our approach uses an active-active architecture across up to 4 switches with no single points of failure.

Do you want some added perspective? Take a look at Phil Edholm's blog.

In my Oct 22 posting "Beware the Single Vendor as Trusted Advisor: Gartner", I discussed with you the pitfalls of single vendor networking. Well opportunity knocks and I had a chance to debate Cisco on this hot topic.

Take a look and tell me what you think about dropping Cisco in favor of best in breed vendors.

Let's talk about Hyperconnectivity and network implications of unified communications and of an explosion in network-connected devices - for example, in the realms of energy and property management, asset and location tracking, telemetry and enhanced security systems. This is enabled by low-cost sensors and actuators that can detect over 100 different physical parameters, including temperature, radiation levels, door closures, visual and audio signals and location - and that can cost-effectively transmit this information.

Scaling the network by a factor of 10 to 100, the most obvious of a number of new requirements, can’t be achieved without fundamentally streamlining current network environments. Hyperconnectivity demands simplification on a grand scale, transforming the network into a business optimized infrastructure. This will be as big as the transformation from departmental LANs to enterprise IP networking.

So how will you simplify your network, while delivering performance across an ever-broader range of applications?