A few weeks ago, Level 3 had announced that it had snagged MySpace as a customer. Today the company announced that it has signed up YouTube as a customer. YouTube is going to buy 10 gigabit/second connections from Level 3 to build out a nationwide backbone. What is this? A Web 2.0 makeover?
I guess they need all that bandwidth, up until the point they can find someone else to pay the bills. The company claims that it serves 100 million videos per day, and more than 65,000 videos uploaded everyday. According to Nielsen NetRatings, YouTube has nearly 20M unique users per month.
Still, these kind of Web-meets-telecom press releases make me queasy, and remind me of another time and place when Qwest and its ilk would tout applications on demand. (Funny, AT&T bought a former Q partner USinternetworking for about $300 million in cash and assumed debt.)
Anyway back to Level 3 and Web 2.0. The trick would be to get Jim Crowe to blog, where he can mouth off about people and things he doesn’t much care about. Like George Gilder or SEC? Or author of Broadbandits ;-).
Very interesting development indeed! And if Level 3 then employs p2p technologies to handle their traffic, that would be even more faciniating.
Even level 3 will struggle to handle all that traffic…?
Om, any thoughts regarding Limelight and this deal?
Basic telco colocation making in-roads against complex CDNs?
Inter-site connectivity is something Level(3) has always been good about, and has gotten even better about with all their recent fiber and metro acquisitions.
In regards to ingjerd, all the connections are P2P already (point to point). A peer-to-peer solution would be subobtimal at best. YouTube has an extensive and growing network and expanding out to datacenters requires backhaul for them (for now at least), this is what Level(3) is providing over the 10ge links — inter-site connectivity. If they are buying transit from Level(3) that’s something else.
As for what it has to do with Limelight, well the limelight carrying youtube’s traffic story has been propagated so far but isn’t true. They carry only a fraction of Youtube’s traffic.
Basic colocation will never make in-roads against complex CDNS but two things are happening: First, CDNs will become simple to use like CacheFly and second, complex networks are becoming less complex. Don’t consider what youtube is doing to be “basic telco colocation” because it isn’t. They have extensive peering and datacenters being deployed. It’s far from basic telco stuff.
-david