Putting IPTV to the Test
This was an interesting, risky, and bold exercise by all the parties involved. And it worked.
EANTC spent gobs of time, resources, and energy on the project. Cisco did the same, in addition to risking its reputation. And for us, of course, we risked both time and money.
Well, now we know. We now know Cisco's gear can handle all the things a carrier IPTV network would ask of it. And we also know that it requires a lot of complex network design and careful software implementations in order to deliver true carrier-quality IPTV.
The largest service providers are of course very secretive about the IPTV reality. Few global IPTV networks are scaling to hundreds of thousands of users, let alone a million. So this was an important test of how far these networks can be pushed.
The full test report is available here: Testing Cisco's IPTV Infrastructure.
Credit Cisco for stepping up to this test: Few vendors would risk it. Cisco achieved the requisite scale through lots of customization, design, and proprietary software features.
Of course, there were glitches: Cisco had to rewrite some code, as well as replace some defective linecards. This is the sort of thing with which I'm sure telecom engineers are familiar – for us, however, it's a unique insight into what happens when you are building a massively scaleable carrier network.
Carsten Rossenhövel, Managing Director of EANTC, points out the scale of the test alone was daunting: "Here we had emulated 120 DSLAMs, more than 200 TV channels."
The request for proposal (RFP) asked for a network designed to serve 1 million IPTV customers with voice, video, and Internet access. The requirements included 20 high-definition and 200 standard-definition TV channels – plus 22-Gbit/s worth of video on demand (VOD) service, representing 6,400 simultaneous users per point of presence (POP). Cisco's first test was to answer the RFP.
At first, EANTC standardized the software for all the Cisco 7600 routers in the network, but later found that it needed to install a different version in order to run multicast failover tests. The CRS-1 software had an MPLS problem that had been discovered by a real-life customer at roughly the same time, requiring a software patch.
To a certain extent, the test validated Cisco's use of proprietary features including Call Admission Control (CAC) and Video Quality Experience (VQE). This led EANTC to the conclusion that, in fact, carrier-quality IPTV networks may require vendor customization, and that the real world probably isn't yet ready for fully interoperable, standards-based IPTV networks.
The bottom line? EANTC concludes that carrier IPTV will continue to require a high degree of vendor customization and design. This is why the largest vendors have moved to large-scale, cross-group integration approaches in their IPTV deployments.
"In order to successfully implement and make money from these services, [carriers] must partner with a vendor that will do extra and sophisticated network design work, and that can supply additional technology to prevent network over-subscription and other problems," the report states.
We already know that IPTV in carrier networks in 2007 is certainly not plug-and-play, but now we know how and why.
I'd like to personally think EANTC and Cisco for cooperating in this exercise, helping us open up the book on carrier IPTV networks. Without this test, we wouldn't have had a detailed look at what exactly is involved in building a network designed to handle a million-user IPTV network in the telecom environment.
— R. Scott Raynovich, Editor in Chief, Light Reading