A week after the biggest Skype outage in recent memory (it lasted around 24 hours,) the company CIO posted an interesting rundown of how and why the failure occurred, giving a glimpse as to how Skype works behind the scenes. On Wednesday, December 22, a Skype server handling offline messaging became overloaded, resulting in delayed messages. Due to a bug, a version of Skype for Windows did not process those delayed messages correctly, which made them crash. This led to around 25% of the total Skype supernodes, the clients directing connections and logins in the Skype network, to crash. Since Skype clients have protections built in so that they do not overload the systems they run on, the large amount of crashed clients being restarted caused a massive load on the network, causing more supernodes to shut down to protect themselves. Almost the whole Skype network was thus brought to a halt. Check out the post to view more details, and how the Skype team brought everything back online.
Read More | Skype
Skype suffered a major outage yesterday where many users, possibly millions, were left without service. It started early in the morning, with people finding that they were unable to login, see their friends list, or place calls. At first, Skype tweeted that they were looking into the issue, but it soon became clear that the problem was affecting a large number of people, for several hours.
Skype put up a blog post in the afternoon to explain what the situation was, and how it was a problem affecting many users that caused the number of supernodes, those systems that are used by the Skype network to connect calls and process logins, to be taken offline. They added that they are working on creating "mega-supernodes" to take over that role and rectify the problem. According to their estimates, voice calls should be functioning within hours, but video calls and group features may be down longer. This may end up being one of the most serious issue that the company has had in the past several years.
Read More | Skype