Current news events have caused us to ask: "What happens when computers fail?" The recent system failures experienced by the New York Stock Exchange, American Airlines, the Wall Street Journal, Southwest Airlines, and now Delta Airlines, show us that the increasing complexity and interconnectedness of systems is at the very least problematic and at its worst, dangerous.
This warning shot should wake us up and make us take note of the need for computer technicians and software developers who have a solid knowledge of the systems they are using. The number of people who understand legacy code and architecture are quickly dwindling. So, who's to blame when computers fail?
Too many people starting out in computer science studies are exposed to programming first and then the architecture and theory is glossed over. More time needs to be spent covering the basics of computer logic. Too many techs are trying to make the machine a purely organic thinker when at its Root is Boolean logic.
Another issue we are facing is increased obsolesence of devices. Obsolesence is being driven by the corporate world and in its wake we are being rocked by system failures or incompatibilities. While it is good to march forward with tech and try new things – much more time should be spent in abstraction models.
Some of our current system crashes may be the result of too many CEOs who don't understand the technology around them and are pushing the envelope. Alongside these CEOs are numerous technicians who are willing to take on the challenge out of pure competition or maybe out of fear of loosing employment.
In recent news, several large corporations have made the news with their computer system failures:
→ In July 2015 the New York Stock Exchange, United Airlines, and the Wall Street Journal's website, all suffered “minor” computer malfunctions. While these were minor glitches, and only a matter of hours were lost, the resulting news from these glitches made the general public uneasy and whispers of terrorism floated through the Internet. Of course, these types of failures happen every day in all corners of the digital world but because fewer and fewer people understand what is going on behind their screens; these small glitches can create great chaos. What ties these three failures together? These groups are all business operations that rely on massively intricate computer systems. With systems such as these, something as simple as a bit of misplaced text or code can cripple the system.
→ Last August the FAA base of operations in Leesburg, VA suffered a crash due to upgraded software. Several hours were lost before technicians reset the system that tracks high-altitude flights to an earlier program.
Less than three weeks ago, Southwest Airlines was hit with a major systems malfunction that they say originated with a router failure and resulted in a “cascading technology glitch.” From July 20th to July 25th, Southwest Airlines was hobbled and racked up 2,000+ cancellations. They also experienced late arrivals totaling over 8,000.
→ This week, we saw Delta Airlines have their own system crash which resulted in worldwide flight delays. Initially, the problem was reported as cascading failure due to an overnight power outage at its hometown facility in Atlanta, GA. However, a day later, Georgia Power is reporting that a piece of electrical equipment on Delta's own system caused the power outage that only affected Delta's facility.
→ In the case of Southwest Airlines, there apparently was a backup system that was meant to kick in when the main system failed; however, the backup system failed also. There has been no report yet about Delta's use of a backup system. It is possible that they don't have one.
An ever-increasing number of computer systems are interacting with each other in complicated ways. They use algorithms to make changes in themselves and each other as the humans stand back and watch. We are even seeing some instances of behavior between components that shows an almost organic performance. With actions like this, intricate computer systems could start making their own decisions in performance. Some may welcome the self-servicing system; however, I for one would feel better knowing that some competent and level-headed technicians were at the switch instead.