MediaFile

By Laura Isensee

A majority of Google users from California to Taiwan found themselves without access to Google’s popular email service on Tuesday.

Google has a diagnosis: The outage, which lasted more than an hour and a half, was a “Big Deal.”

The company outlined what went wrong on its blog.

“We took a small fraction of Gmail’s servers offline to perform routine upgrades,” Ben Treynor, vice president of engineering and site reliability czar, wrote.

But the company “slightly underestimated” the load that placed on other servers — called request routers — that direct web queries to the right Gmail server for response.

So those servers became overloaded, pushing the load to the remaining request routers, causing more to become overloaded. And “within minutes nearly all of the request routers were overloaded,” Treynor said.

A minor panic spread across the Internet on Thursday after Google suffered what appeared to be a temporary service outage.

Reports cropped up throughout the Web that Google’s search engine, as well as popular services like Gmail and Google Analytics were running slowly or not at all in various parts of the world on Thursday morning.

And the incident caused ripples that slowed down other Web sites.

Because many online firms have woven javascript applications like Google Analytics – which provides analysis of a site’s traffic – into their Web sites, the Google outage impacted their own sites. Imad Mouline, the CTO of Web performance monitoring firm Gomez said his company noticed that certain sites took between two times and four times longer to load because of links to Google applications on Thursday morning.