Someday we’ll all tell our grandkids about what we were doing during the great Gmail outage of August 11th, 2008. Well, okay, probably not–Google’s e-mail service was down for only a couple of hours, which is relatively brief as Internet outages go. But when one of the world’s most popular mail systems goes missing even briefly, zillions of people are inconvenienced and want to share their frustration. In a weird way, it’s a huge compliment: If Gmail wasn’t essential, nobody would care if it went away.
For a dozen years or so now, the Internet has been a mainstream communications medium, and its history has been pockmarked with examples of big-time services choking for extended periods–often a lot longer than today’s Gmail blip. The most famous examples of unplanned downtime have a lot in common: They usually last longer than anyone expected and get blamed on cryptic technical glitches. Almost always, angry consumers announce they’re done with the service in question; almost always, the service eventually recovers.
Oh, and one more thing: The biggest and most embarrassing failures all seem to happen during the summer months. Maybe technology, like human beings, just doesn’t work quite as hard when the weather’s hot and there are distractions like baseball games, picnics, and vacations to contemplate.
Now that Gmail’s back, it’s worth recapping a few other outages that made headlines when they happened–and since the ones that follow are in alphabetical order, they begin with maybe the most famous one of all (hint: it involved a company whose initials are A.O.L.)…
Who? America Online, the company that was synonymous with online services in the mid-1990s.
When? August 1996.
How long? 19 hours.
Why? AOL CEO Steve case blamed it on problems with the installation of “”high capacity switches within the local area network.“
Upshot? The service’s 6.3 million users received refunds for the downtime.
Saving grace? Steve Case’s artful and honest work as chief apologist made him even more famous as AOL’s public face.
Any other problems? AOL suffered multiple embarrassing outages in the era (here’s another one), then pretty much got over having ‘em–though a 5-hour e-mail meltdown in 2006 felt like old times all over again.
Who? eBay, then as now the world’s biggest auction site.
When? June 1998.
How long? Almost a day–reportedly the biggest outage in online history at the time.
Why? “[The] problem resulted from a failure in the software that was used to list items for sale and update bids. He insisted it was not a problem with the site being strained to capacity and said it was not clear why it had failed.” Matters were complicated further by the fact that eBay hadn’t quite finished implementing a backup system.
Upshot? eBay lost an estimated $3-$5 million in sales and was forced to offer refunds, extend auctions, and waive fees by way of apology to sellers whose auctions were affected.
Saving grace? Some eBay users who had bid early on items in auctions that were scheduled to end during the time the service turned out to be down got bargains.
Any other problems? This outage was one of several catastrophic ones that eBay suffered in 1998 and 1999.
Who? MobileMe, Apple’s brand-new service for synching information on iPhones, Macs, and PCs.
When? July 2008
How long? Several weeks of at least sporadic issues for some users.
Why? Apple’s MobileMe blog mentioned multiple technical gremlins, including “a lot more traffic to our servers than we anticipated,” a bug “that was preventing MobileMe IMAP mail folders from syncing correctly between the web app and Mac OS X Mail or Outlook,” and “a serious problem with one of our mail servers.”
Upshot? Apple extended all MobileMe accounts by a month; Steve Jobs said that the MobileMe launch wasn’t “up to Apple’s standards” in a widely-published internal memo and put a new Apple exec in charge of the service.
Saving grace? None to date, although it’s been a rare opportunity to see Apple eat crow.
Any other problems? As I write this, it appears that the worst of MobileMe’s woes may behind it, but grumbling continues; on June 29th, an Apple blog post promised an update later in the week, but none have been published.
Who? MSN Messenger, the instant-messaging service from the world’s largest software company.
When? July 2001.
How long? For some users, about a week.
Why? “[An] extremely rare set of circumstances occurred when one of our database servers had a disc controller fail.“
Upshot? Some unhappy users decamped to AIM, Yahoo Messenger, and other IM services.
Saving grace? The outage didn’t help people feel better about Microsoft’s scary HailStorm initiative for serving as a one-stop repository for all of consumers’ data and probably helped to kill it.
Any other problems? MSN Messenger also went down in 2003 and 2005.