Thursday, December 23, 2010
Skype Scrambling To Fix Global 'Supernode' Outage
A failure of "supernodes" that connect users took Skype down on Wednesday, and as of Thursday morning, millions of users around the world were still disconnected. Skype's Peter Parkes said about 30 percent of Skype users were back online. An analyst said the anger among Skype users shows the Internet phone service is now an enterprise player.
Skype went down suddenly Wednesday -- and still hasn't come back online for millions of users around the world. As of Thursday morning, about five million people are back online.
Peter Parkes, Skype's blogger-in-chief, said the ability of one Skype user to find another relies on what the company calls "supernodes." On Wednesday, he explained, a number of those supernodes failed due to a software issue. Skype has identified the issue and engineers are working to resolve the problem.
"Millions of you are already reporting that you can now sign in to Skype normally, and we estimate that there are already almost five million people online," Parkes said. "As a guide, this is around 30 percent of what we'd expect at this time of day -- and that number is increasing all the time. Unfortunately, it's not possible for us to predict on an individual level when you'll be able to sign in again, and we thank you for your patience in the meantime."
Brad Shimmin, an analyst at Current Analysis, said it's interesting to watch the reaction -- not so much of Skype but of Skype's consumer and business users -- to the outage . As he sees it, more moderate expectations of availability driven by the consumerization of IT , coupled with the seemingly impervious nature of the Internet itself, have created a strange brew where customers anticipate outages like this yet still feel outraged when they occur.
"Certainly Skype is no longer a tiny, private startup playing the role of market disruptor," Shimmin said. "If anything, the angry shouting that has followed on the heels of this outage point to the fact that Skype is now an enterprise player, and as such it must adhere to a higher standard of performance and, if not performance, then responsibility."
Shimmin pointed to vendors like Microsoft that post financial credits to customer accounts whenever they fail to meet a self-published service level agreement for services like BPOS/Office 365. Another interesting facet of this outage, he said, stems from Skype's description of the problem.
"The company indicated that the problem stemmed not from loss of connectivity, but rather from a loss of communication between users and Skype supernodes, which appear to be directory-service platforms, much like the Internet's domain naming system servers," Shimmin said. "It's too bad Skype's servers don't behave in the same, redundant manner as DNS servers, which are by and large responsible for the resiliency of the Internet itself."
Skype explained that it isn't a network like a conventional phone or IM network. Instead, it relies on millions of individual connections between computers and phones to keep things up and running. That's where the supernodes come in. Under normal circumstances, Skype explained, there are a large number of supernodes available.
"Our engineers are creating new 'mega-supernodes' as fast as they can, which should gradually return things to normal," Skype said in an update. "This may take a few hours, and we sincerely apologize for the disruption to your conversations. Some features, like group video calling, may take longer to return to normal."