Blame it on older Windows clients. That's the verdict, forensically
, from Skype's investigation into what caused the recent massive Skype outage.
Skype is a P2P service, and relies on supernodes, which exist outside firewalls, for its directory service. The current Windows version of Skype is 18.104.22.168. However, a bug in 22.214.171.124 caused those clients to crash. Here's what Skype said:
On Wednesday, December 22, a cluster of support servers responsible for offline instant messaging became overloaded. As a result of this overload, some Skype clients received delayed responses from the overloaded servers. In a version of the Skype for Windows client (version 5.0.0152), the delayed responses from the overloaded servers were not properly processed, causing Windows clients running the affected version to crash.
The worst news is that around 50 percent of all Skype users globally were running the 126.96.36.199 version of Skype for Windows. The crashes caused approximately 40 percent of those clients to fail. In effect, 20 percent of clients overall failed, as the newer client, older (4.0) Windows clients, and clients on other platforms were unaffected by the initial problem.
However, once those 188.8.131.52 Windows clients crashed, that meant a number of supernodes, approximately 25 – 30 percent, failed. That's when things really started to fail on the Skype network.
What can Skype do to keep this from happening again? They will be reviewing their procedures for "automatic" updates to end users, but in addition to that the company outlined plans, including strengthening their core systems, in their forensic blog post.
In addition, the company sent an email to affected users, issuing a $1 credit voucher for the inconvenience. While that seems ludicrously small, it should be noted that landline companies don't give users a credit for an outage, nor do broadband companies. Thus, it's not as bad a deal as it might sound.