Five Greatest Web Hosting Disasters of All Time

It was a freezing cold morning at my first Internet Service Provider job. I was the first one to the tiny office, and I plugged in the space heater under my boss’s desk.

I went to my folding table, just a few feet away in our one-room office, and began to read email. Suddenly the room erupted in alarms and flashing lights. Everything was down!

My boss arrived to find the office in uproar, and since I had just started and knew almost nothing, I had nothing to do. I watched as he quickly determined the source of the problem: I had plugged the space heater into a power strip which was plugged into a UPS. It had quickly depleted the UPS’s battery and turned the unit off, taking down the main servers also plugged into it.

This kind of thing happens all the time at startup ISPs. During my career there, I took down the servers not once, not twice, but three times with my unmitigated boneheadedness. As I was reflecting on this recently, a thought hit me: what was the biggest Web Hosting disaster in history?

In the following pages, I’ve summarized the Five Greatest Web Hosting Disasters of All Time. Please understand that these numbers are “educated guesses,” as companies tend to play their cards pretty close to the vest when it comes to colossal foul-ups. Here’s what I was able to track down.

Web Hosting DISASTER #5:

All Hotmail Mailboxes Open to Anyone

WHAT HAPPENED: A group of hackers publicized a security hole present in all Hotmail accounts which allowed them to be accessed by anyone with a web browser using the password “eh”. Though the hackers claimed that this was a backdoor situation left as an oversight by Microsoft employees, Microsoft spokespeople denied the claims.

This affected all Hotmail accounts, but apparently didn’t cause any great disruption in Microsoft stock trading, which fell only a point.

COST: $55,000 ($.001 average loss per visitor)
OTHER FACTORS: Canadians loved the password.

Web Hosting DISASTER #4:

Microsoft Router Config Screwup

WHAT HAPPENED: In a flourish of corporate finger-pointing, Microsoft blamed its own technicians for an error which caused a major disruption in the software giant’s Internet connection for nearly 24 hours. The error in question was apparently a botched router configuration. No report was found to indicate the fate of the technicians involved.

Unfortunately for Microsoft, the error came directly on the heels of their launch of a $200 million advertising campaign to bolster their business software’s reputation as compared to companies like Oracle and IBM. The outage affected some of the most popular sites on the Internet at the time, including and

COST: $450,000 ($.05 average loss per visitor + cleanup effort + wasted ad campaign)
OTHER FACTORS: Bill Gates was on vacation.

Web Hosting DISASTER #3:

Microsoft Forgets to Renew Domain

WHAT HAPPENED: Everyone makes mistakes, but not every mistake causes Hotmail to take a dive. When Microsoft forgot to re-up their registration for “” in December 1999, lost its ability to authenticate, locking out 20 million people from 55 million free e-mail accounts. This happened on Christmas Eve, robbing Santa of precious e-mail access to his elves, who all use Hotmail.

Luckily for Microsoft, a Linux user named Michael Chaney paid their bill in an effort to get Hotmail back up. He was eventually reimbursed by the software giant, presented with a $500 check, and some free software. He put the check on eBay, and the funds eventually found their way into the hands of the Sisters of the Road Cafe, an organization which helps the homeless and impoverished in Portland, OR.

SOURCE: DoubleWide
COST: $1,000,000 ($.05 average loss per e-mail address)
OTHER FACTORS: Ruined Christmas.

Web Hosting DISASTER #2:

AOL Down for 19 Hours

WHAT HAPPENED: In August 1996, Steve Case posted a note to AOL’s six million subscribers, starting with the words “Wednesday was a bad day for me.” Mr. Case went on to blame his bad day on some “high capacity switches within the local area network.”

It is unknown why or how these switches brought AOL down, but the time in question was nearly 18 hours and 45 minutes. AOL subscribers were issued a refund for a one-day prorated amount of their monthly fee. This should have come to around $.66 per person ($19.99 divided by 30 days) paid to six million people for a total loss of $3.9 million in reimbursement. A bad day indeed!

COST: $4,000,000 ($.66 average loss per subscriber + cleanup effort)
OTHER FACTORS: Millions of grandmothers couldn’t pass along chain letters.

Web Hosting DISASTER #1:

AT&T Frame Relay Network Goes Down

WHAT HAPPENED: AT&T, at the time the leading supplier of frame relay networks, turned in one of the most catastrophic ISP whoops-a-daisies of all time when their frame relay network took a dive on April 13th, 1998 and stayed down for nearly 26 hours. This left more than 6,000 customers in the dark, mostly institutions like banks, retailers, and manufacturers.

It seems that some bad code got uploaded to a Cisco switch, which caused a data storm that raged across the entire network and shut down all 145 of the switches connected. The problem switch wasn’t at a customer location, but was connected to the network, and managed to take the whole system down.

Based on the enterprise-level client base of AT&T’s frame relay network, this screw-up deserves the title of Greatest Hosting Disaster of All Time.

COST: $60,000,000 ($10,000 average loss per customer)
OTHER FACTORS: Also ruined a developer’s career.

Twitter Facebook