Is Downtime More Frequent, or More Visible?

Is Downtime More Frequent, or More Visible?
Are leading Internet sites reliable enough? The New York Times examines web downtime today in a front-page story, which focuses on users’ growing reliance upon web services. «Now the Web is an irreplaceable part of daily life, and Internet companies have plans to make us even more dependent on it,» writes Brad Stone. «The problem is that this ideal requires Web services to be available around the clock — and even the Internet’s biggest companies sometimes have trouble making that happen.»

Is downtime really more frequent? Or is it just more visible? The answer is both. Blogs, monitoring services and forums service notice of outages within minutes (see Web Outages and Damage Control 2.0 for more on this phenomenon). We’ll all survive outages at free services like Twitter and Blogger. But there have also been a number of lengthy outages at hosting services that have taken a real-world financial toll on users. Here’s a look at some of the high-profile outages of the past year, and some less prominent incidents that nonetheless affected large numbers of users:

• June 1: An explosion and fire at a Houston data center operated by The Planet leaves 9,000 servers offline. At least 3,000 of those servers remained offline for days as the dedicated hosting firm sought to recover.
• May 23: Thousands of web sites hosted by Orlando reseller hosting specialist HostDime were offline for hours because of power issues in one of the company’s data centers.
• May 15: A power outage at the data center for The Intercontinental Exchange (ICE) led to a suspension of all commodities trading on the ICE exchanges.

• March 24: The web site for the popular movie rental service Netflix goes dark for more than nine hours due to a database problem, which also delays DVD deliveries.
• Feb. 24: YouTube is knocked offline for two hours when Pakistan Telecom claims its IP addresses, sparking a debate about whether the outage was a botched effort to block Pakistanis’ access to the site, or a deliberate political IP hijacking.
• Dec. 4, 2007: Torrential rains in the Pacific Northwest flood a T-Mobile data center in Bothell, Wash., affecting customer activations and the performance of the Tmobile.com web site. The outage reportedly took down servers supporting T-Mobile service activation portals, company web sites and myFaves updates.
• Nov. 26: In one of the most disruptive outages, thousands of merchants using Yahoo Stores find their web sites unable to process transactions on Cyber Monday, the most hyped online shopping day of the year.
• Nov. 13: A truck hits a utility transformer, knocking out power to the main Dallas data center for Rackspace. The outage was widely noticed in the blogosphere, as Rackspace hosts many popular Web 2.0 sites and blog services.
• Nov. 6: Many customers of Alabanza say their sites were offline for up to 72 hours as they were migrated to a data center at NaviSite, which acquired Alabanza in August. NaviSite was moving customer accounts from Alabanza’s main data center in Baltimore to a NaviSite facility in Andover, Mass.
• July 31: A data center migration at ValueWeb goes badly, leaving thousands of sites offline, some for as long as three days. The migration involved moving several thousand dedicated servers from Affinity Internet’s Miami, Fla. hosting facility to a Hostway data center in Tampa, about 270 miles away.
• July 23: Several generators at 365 Main’s San Francisco data center fail to start when the facility lost grid power Tuesday afternoon, causing an outage that knocked many of the web’s most popular destinations offline for several hours, including CraigsList, Technorati, LiveJournal, TypePad, AdBrite, the 1Up gaming network, Second Life and Yelp, among others.