IP Video Conferencing: No guts, no glory

A couple of months back, I blogged on the interesting challenges that deploying IP video conferencing technologies poses to service quality and service assurance – especially when it comes to network capacity and performance. Since then, the picture is starting to clear regarding how these technologies work, and some of the cost savings they can deliver. I spent some time with an internal Cisco IT representative who is responsible for their internal deployment of telepresence, which they believe is the largest in the world at the present time, with 440 end points as of mid-April. At that time, those 440 represented almost 20% of globally deployed Cisco telepresence units. Cisco has taken it upon themselves to make the investments and see just what kind of difference telepresence can make. And the results are impressive. They recently claimed that across their three-year run in deploying and utilizing telepresence, they have confirmed nearly $150 million in cost savings, mostly due to reduced travel. But what they are equally excited about are the soft values – they feel that their globally-distributed workforce is achieving new levels of collaborative innovation, both internally and with customers, that they did not even conceive of when planning this move to virtual communications. And their internal telepresence suites are booked non-stop - I had to go to a sales office to get time on one. These kind of results get attention from mahogany row - the kind that means you had better be ready when you get the call to deploy. Another interesting metric – a Cisco representative informed me at Interop (two weeks ago) that video now makes up an astounding 80% of traffic on the Cisco internal network. This isn’t all video conferencing – Cisco makes heavy use of streaming video communications and video blogging – but that number is staggering. Anyone else out there giving up 80% of traffic to video?? I also went another round with Avistar and their CTO, Dr. Chris Lauwers. Avistar focuses on desktop video conferencing, which represents another interesting twist to the challenge. Instead of having to change your network to handle a small number of high bandwidth telepresence suites (which can demand as much as 15 Mbps each), what if you decide to give everyone desktop video conferencing and now have to handle hundreds of simultaneous video sessions that might be lower bandwidth demand but collectively could be an even bigger load on your network? Chris had some practical advice to give when planning deployments of their solutions – make sure you are reserving 15-20% of your network bandwidth (via traffic prioritization) for real-time video conference, as a starting point. As your deployment proceeds, you can adjust this up or down to fit the adoption and usage patterns that emerge. The biggest challenge here is that even if desktop video conferencing only demands 3-5 times the bandwidth of a VoIP call, just having a few concurrent sessions could easily overload a typical T1 remote branch connection. Avistar’s answer is to prevent the system from allowing that to happen by intelligently recognizing how many sessions are active versus available bandwidth – a good hedge, but one that has to be configured manually at the present time. I’m still watching this space closely. In fact, EMA is now gathering inputs from practitioners on how they are handling this. If you have VoIP or IP Video Conferencing deployed in your organization, we’d like to hear how it’s going and what changes you’ve made to accommodate this new killer app. Please take our survey and you’ll get a free copy of the results.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Must read: 10 new UI features coming to Windows 10