Some ISPs are turning away paid high-speed transit traffic from video companies, citing the cost of the equipment needed to upgrade to 10 gigabit Ethernet connections. These ISPs are sticking with slower connections, even though the growing volume of video traffic could create periodic capacity challenges. This trend was noted earlier this month by Bill Norton, a founder of Equinix and leading researcher on Internet peering, who predicts that the surging popularity of online video will bring "a new wave of disruption that potentially dwarfs currently peered Internet traffic."
"This has become a significant issue NOW because a few of the largest US ISPs are turning away these n*10G Internet video transit customers," Norton said in a post in the North American Network Operators Group (NANOG), which prompted a lengthy discussion of online video's impact on network capacity. The topic has become even more relevant in light of the potentially disruptive impact of the peer-to-peer IPTV app Joost (The Venice Project) and NetFlix' announcement that it will stream full-length feature films.
Several providers, including Equinix, say video growth is driving strong demand for 10 gigabit Ethernet connections, high-capacity pipes that enable providers to move enormous volumes of traffic. But the volume of traffic generated by online video is altering the economics of these connections. In some cases, the financial benefits of big-pipe peering don't offset the short-term expense of network upgrades needed to support 10 gigabit Ethernet equipment. While the parties can continue to peer over existing connections, the growing volume of video traffic could stress networks and affect performance.
Peering allows two providers exchanging large volumes of traffic to save money by connecting directly, rather than routing traffic through their paid Internet connections (known as "transit"). Peering doesn't provide access to the entire Internet, only the other provider's customers. Peering arrangements are sometimes "settlement free," but in some instances fees will be paid by one provider, typically the one sending the greater amount of traffic onto the other's network. Even with these settlement fees, peering can be cheaper than reaching the same users using paid transit. In recent years cable companies, content providers and regional ISPs have used this strategy to save money by peering directly, thus keeping traffic off Tier 1 backbones.
Why are some ISPs reluctant to make 10 gigabit peering connections? Norton gave an example from an unidentified Tier 1 ISP. "First, the network equipment currently deployed hasn't been paid for and they would have to go back and argue for more (money) for a forklift upgrade," Norton wrote, using industry jargon for a major equipment upgrade. "Which leads to the second reason - the colos are out of space, power or both. Usually both. So a forklift upgrade may be needed to replace the current gear to handle the emerging n*10G Video traffic demand."
A recent report by Deloitte analyst Igal Brightman warned that such cost/benefit imbalances are a significant challenge for the industry. "There needs to be a business case for investing in new capacity," Brightman wrote. "However if the price paid for bandwidth continues falling, there may be no economic rationale for either upgrading or laying new fiber. Thus the cost of wholesale transmission may well have to rise.
"This may be a far larger issue than net neutrality," Brightman added. "It is about the long-term commercial viability of broadband provision. The whole industry needs to find a satisfactory solution to the issue of monetizing rapidly increasing usage to fund future growth, without disenfranchising customers."
Some in the NANOG discussion said the largest markets - particularly, New York, Los Angeles and San Francisco - should be able to accommodate demand for wide-pipe connections. One NANOG veteran saw this as a cyclical issue. "Don't we go through this every other network generation?" wrote Sean Donelan. "We'll have wailing and gnashing of teeth for many months but somehow things will balance out again as technology, equipment and revenue catch up with each other. And then it will start over again with someone wanting 100GE connections, and then 1,000GE connections, etc, etc."
A key issue is the cost of 10 gigabit Ethernet equipment. "10-Gig gear has been slower to commoditize than its predecessors," according to an analysis last year by Business Communications Review. "Unlike Gigabit Ethernet, which offered a significant discount per gigabit of bandwidth over Fast Ethernet, 10-Gigabit Ethernet continues to cost a bandwidth price premium that discourages its broad adoption."
But the pricing trend has improved as component prices drops. "We forecast that by late 2007, average selling prices of 10-Gigabit Ethernet will follow the pattern set by Gigabit Ethernet and Fast Ethernet, offering a bandwidth discount over lower-speed technologies," the report concluded. "In turn, we anticipate stronger adoption of 10-Gigabit Ethernet."