A blizzard rages outside as I write this, and the governor of Massachusetts has banned travel on the roads. Many of us from Akamai's Cambridge headquarters will spend today at home, and possibly tomorrow.
But Akamai will continue to run. Being spread across the globe makes that a given. It illustrates the power of redundancy.
As a journalist who conducted frequent site visits to see the differences and similarities of large enterprise IT security shops, I found that redundancy was a constant element -- multiple systems spread out across the world. If one data center went down during a natural disaster, for example, others could pick up the slack.
One exec I talked to noted that most of his company's workforce operates remotely. With people working from home all over the country, the operation continued to run smoothly during Superstorm Sandy in 2012, even though the company's headquarters is in New Jersey -- one of the states hardest hit during the storm.
Thanks to redundant systems, all the workforce needed was power and an Internet connection. Some lost both in the storm, but there were enough workers spread out and with power that the work got done and customer needs were met.
Such is the case at Akamai. We have redundancy across servers, server racks, data centers, cities, countries, and even continents. "If your users have power and an Internet connection, they can get to our servers," my colleague Michael Smith once told me.
Another important element is in how we handle mapping -- which allows us to quickly route traffic around the Internet's trouble spots. I looked around and found some good reading on the subject.