Skype goes down worldwide for first time since 2007

Updated--It wasn't exactly a red-letter day for Skype: Users trying to access the service to make calls Wednesday morning had trouble connecting, a problem confirmed by Skype in a Twitter message. The VoIP service continued to update users through the day, saying by 4:00 p.m. EST that service is "now gradually returning to normal--we expect it may take several hours for everyone to be able to sign in again, however."

The problem, Skype explained on its website, originated among its supernodes, many of which (Skype wouldn't say how many) were "taken offline by a problem affecting some versions of Skype." The site went on to say "Our engineers are creating new ‘mega-supernodes' as fast as they can" and referred visitors back to Twitter for status updates.

To add to the ignominy, in a classic cause-equals-effect motion, the number of Skype users complaining about the outage on Twitter likely caused the annoying "fail whale" to raise its head above the waves for a few hours Wednesday as well, says tech blog TheNextWeb.com.

Failure of the supernodes is a pretty critical issue, notes Charles Studt, Vice President of Product Marketing at IntelePeer, a cloud communications provider.

"In any modern IP communications network, that centralized directory function is pretty important obviously for being able to manage essentially, where are the end points located or what routes should I take to find them for a given process flow," he told FierceTelecom.

"The takeaway probably is, obviously those pieces are really critical to how the network operates, and ... whether those supernodes were just unavailable because someone pulled a plug in a server room or whether they were doing a software upgrade globally to all the supernodes and something didn't go so well in the upgrade ...but regardless of the reason, when that intelligence leaves the network, then you have a big problem."

The last time Skype experienced a major outage was in 2007, caused by what the company called an "algorithm deficiency."

For more:
- L.A. Times has this blog post
- see Skype's statement
- and this article

Related articles:
How do you and your company deal with VoIP phone service outages?
Skype, Fring and the real cost of VoIP and video calling