Skype, the world's leader in internet telephony services, experienced an outage which affected millions of its customers.  Jonathan Richards, The Times - 20 August 2007 has written a piece explaining the situation...

---

 

Skype, the internet telephony provider, said today that a massive, simultaneous restart of its users' computers was to blame for the two-day outage its service suffered last week.

Millions of users collectively downloading a routine security update from Microsoft and then rebooting their machines had caused what the company called a "critical disruption", prompting its peer-to-peer network to collapse.

"The high number of restarts caused a flood of log-in requests," a post on the Skype blog read, "which, combined with the lack of peer-to-peer network resources, prompted a chain reaction that had a critical impact."

The company's network has "an inbuilt ability to self-heal," the post said, "however this event revealed a previously unseen software bug which prevented the self-healing function from working quickly. Regrettably as a result of this disruption, Skype was unavailable for to the majority of its users for approximately two days."

Microsoft releases a security update on the second Tuesday of every month, and for August, 'Patch Tuesday' - as the day has become known - fell last week, two days before Skype's service went down.

Microsoft said today that it always recommends users reboot after they have downloaded its security patches, and it was unclear why this particular update had had such a dramatic effect on the Skype network.

The Skype blog post said that the disruption had been "unprecedented in terms of its impact and scope," and added - by way of disclaimer - that "very few technologies or communications networks today are guaranteed to operate without interruptions."

As many as 220 million Skype users were unable to make cheap internet phone calls as a a result of the outage, which began on Thursday afternoon.

Analysts were surprised by the explanation, pointing out that software updates were common, and would not normally present a problem even if 6 million people - which is the estimated number of "active" Skype users at any one time, tried to log on immediately afterwards.

Steve Blood, an voice over internet protocol (VoIP) analyst at Gartner, speculated that perhaps in this case, the update required the users' computers to send information to Skype, which would have "invoked extra processing power" and placed the network under greater than normal strain.

"Either way they'll be wanting to improve their procedure to cope with software updates - it's pretty embarrassing; if this kind of thing starts to happen, it's not long before people begin to lose confidence in the service," Mr Blood said.

Skype, which uses peer-to-peer technology to connect calls rather than routing them through a central hub, said that it had "identified and introduced a number of improvements to its software" to ensure that users would not be similarly affected "in the unlikely possibility of this combination of events recurring."

The company, which was bought by by eBay for $2.6 billion two years ago, allows its members to call each other for free using their computer as well as call traditional phones at cheaper than market rates.

A number of businesses have also begun to use it to reduce the cost of personal calls made by employees while abroad.