Confronting the final disaster

It was an ordinary, cloudless morning in England on September 1, 1859. This was the heyday of the Victorian amateur scientist, before science became institutional. One such amateur astronomer, Richard Carrington, was doing what he often did, which was monitor the sun.

Using a telescope to project an image of the sun onto a white screen, Carrington was sketching the sunspots when suddenly he saw two beads of astonishingly bright, white light. As Carrington later wrote, “the brilliancy was fully equal to that of direct sun-light.” Startled by what he was observing, he recorded that he “hastily ran to call some one to witness the exhibition with me, and on returning within 60 seconds, was mortified to find that it was already much changed and enfeebled.”

The entire event took about five minutes, start to finish. Nobody had ever seen the like before. Nobody knew what it meant or what it was. Soon they would, and the knowledge changed our understanding of the sun and our world thereafter.


Source: NASA

In the field of data protection we deal with many kinds of data loss events. They can be mechanical (a failed hard drive), operational (whoops! I deleted it again), facilities-based (why is it raining in the data center? Oh, the pipe broke), or environmental (fires, floods, space invaders).  We plan for all of these -- perhaps not space invaders – but have you ever stopped to consider what would happen if there was an event that wiped out not only your data and your data center, but along with it destroyed even the means to begin bringing things back?

It’s entirely possible and it could happen at any time.

What Richard Carrington saw that day was a Coronal Mass Ejection (CME), a gigantic burst of solar energy that is fired into space, sometimes directly at the earth.  In 1859, it took 17 hours for the CME that Carrington observed to reach the planet.

The effects were noticed all over the earth. Night skies lit up like day. Gold miners in the Rocky Mountains awoke thinking it was morning. People in the Northeast Region of the United States reported being able to read newspapers at night by the light of the flare.  And the telegraphs lit on fire.

In 1859, there was very little electronic infrastructure, and the only electronic form of communication were the telegraph lines. The energy from Carrington’s flare (now referred to as the Carrington Event) created havoc as it coursed through the system. Sparks burst forth from telegraph equipment, setting paper on fire and harming some operators. So much power soaked the lines that even when telegraph batteries were disconnected there was sufficient charge to send messages.    

As it turns out, the Carrington Event was the largest such flare in the 160-year recorded history of solar events. Even particle records in Arctic ice show us that it was the largest solar flare event in at least 500 years, and twice the size of the next largest.

So, what’s to worry about?  Plenty! Though an extreme example, a Carrington-sized flare could largely destroy all electric generating and transmission capability across a wide region of the Earth along with data communications.

If you haven’t noticed lately, pretty much everything depends on electricity and electronic forms of communication.  Step through your typical day mentally and add up all the things, large and small, that need electrical power. Wake up to the electric alarm clock, brush your teeth with water pumped by electricity, put on your electric coffee pot, open the electric refrigerator to grab some milk, check your email  on your iPhone.  Now imagine that you can’t do any of those things, and consider that you’ve probably only been awake at that point for about five minutes.  

Now think about your data center. No power? Well, there’s the generator and that will keep us going for a few days. But wait, the generator itself might be ruined. And in any case, it runs on fuel, and how are we going to get fuel if no fuel pumps are working?

Not to worry. We recover from storms and electrical outages pretty quickly. Same thing, right?

Not even close. In the case of a major CME, electrical transformers will be blown. You can’t fix them, they have to be replaced.  They cost a lot of money so few are purchased for stand-by and, in the case of the United States, transformers are mainly built overseas. In most cases, you have to order them two years in advance.  

In short, we are woefully unprepared for a major CME event. But then, there really isn’t much you can do on an individual or corporate basis. Preparation would require a major commitment from both governments and electric utilities.

Meanwhile, we are faced with more manageable “catastrophes” at our day jobs. While not cosmic in scale, they are critical enough to warrant our attention.  These more practical concerns are an important part of what this blog will focus on. We’ll look at different techniques and technologies, review specific data loss events that come up in the news, and evaluate the “what could have been” in terms of disaster readiness. We will examine events from 50,000 feet as well as right down to the bytes on the hard drive. I hope it is both entertaining and informative, and I invite readers to share their experiences and opinions in the comments section or with me directly at

In the meantime, if I’ve made you nervous about a possible solar flare, you can always get a quick view of sun activity here.  Everything normal right now! Phew! 

Copyright © 2012 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
Shop Tech Products at Amazon