Software errors: Funny, costly, deadly

What’s the cost of a software outage? Sometimes, the cost is you get to be the butt of internet jokes. Other times, you lose millions of dollars. And every once in a while, the cost can be much higher than lost revenue and brand value, such as endangering lives.

Our latests roundup of software snafus from around the web contains incidents from all three categories, and shows once again why your IT teams should be working to avoid such surprises by using Service Virtualization to test your apps from conception to deployment.

Outage affects 911 service from some cell phones nationwide

AT&T customers nationwide, possibly in the millions, were unable to dial 911 for a brief period one recent night, and the company isn’t saying why.

According to the AP, One Arkansas sheriff’s office shared a message from AT&T that blamed “a service outage that is impacting the ability to deliver AT&T Mobility wireless 911 calls in your area.” The Federal Communications Commission was trying to determine how many people were affected. The company has 147 million customers.

AT&T was very sorry, but hasn’t said what caused the glitch.

Another week, another airline outage

If it seems like you read a lot about airline systems crashing, it’s not your imagination, writes John Pletz of Crain’s Chicago Business.

Pletz writes that there were 17 system outages at U.S. airlines last year, and 16 the year before. The years 2011 through 2014  averaged about six a year. An so far in the first three months of 2017, we’ve seen seven big outages for airline computer systems, affecting scheduling, ticketing and booking.

Each major error can cost an airline up to $100 million. On top of that, the problems are getting the attention of folks in the U.S. Congress. Senators Ed Markey of Massachusetts and Richard Blumenthal of Connecticut wrote letters of inquiry to United and Delta in January.

“In a world where consumers can find, purchase and check in for flights from their smartphones, IT failures should not be grounding entire fleets,” they wrote.

Pletz’s piece puts major blame on airline tech infrastructure for causing what he called a “giant technology hairball that’s choking airlines just as they return to growth mode.” He quoted Chicago consultant Joanna Parke as saying a hodgepodge of old and new technology — exactly the kind of situation where Service Virtualization can help — is causing problems.

“The infrastructure supporting the growth is a tangled mess of new and legacy systems that are constantly interacting with and pulling data from myriad other systems. When something goes wrong, quickly isolating and fixing the issue is difficult. … The fix is untangling the mess, tackling it piece by piece.

“The process will be multiyear and cost hundreds of millions. … It would cost more to scrap and replace old systems. And it’s not feasible to hit pause when more passengers than ever are traveling the skies.”

DMV glitch makes Denver drivers taller

A system glitch at the Colorado DMV turned about 12,000 drivers in Denver into giants. You read that right.

The error, by Marquis ID System, was caused by humans who failed to program a conversion of people’s heights. Instead of translating a 71-inch person to 5-foot-11, that driver was listed at an NBA-worthy 7-foot-1.