The Summer Olympics kick off this weekend in London and this one promises to be a highly a technical one as any that’s ever been launched, whether it’s showing the latest Twitter sentiment by color on the London Eye, meeting unprecedented demand for WiFi inside the Olympic complex or handling what’s expected to be insane traffic on the official Olympic web site.
When it comes to the latter, the web site organizers worked hard to be prepared. They didn’t want to face the utter embarrassment of having the web site collapse under the burden of heavy traffic, and in fact the web site has been using Soasta, a cloud-based load-testing service to test the site’s ability to handle heavy traffic loads under a variety of conditions.
They began running load tests with Soasta six months before the game, according to Paul Bunnell, Lead Web Architect for London 2012. Of course, the web site designers wanted to prevent bottlenecks, but they also wanted to preven full-scale meltdowns, he explained.
“The application, infrastructure and all the third-party integration points needed to go through capacity planning and stress testing with different load patterns to ensure reliability and performance 24/7 especially during popular events such as the Opening Ceremonies and 100-meters final. The testing teams were spread across the globe and utilized CloudTest’s ability to provide real-time results to adapt the testing…,” Bunnell said
He knows that the Olympic web site will be one of the most popular on Earth during the 3 week period in which the games are running and he knew it was imperative to make sure the site runs smoothly.
In addition to load testing, according to CIO.com, the site will make extensive use of Content Delivery Networks to help distribute the load including Akamai, which it will rely on specifically to help deliver video content around the globe.
Even with all this preparation, there are of course no guarantees because you can’t ever account for the entirely unexpected. No matter how much you test, and how careful you are, it’s impossible to account for every contingency. If you doubt that, consider what happened to Amazon Web Services earlier this month when lightning caused an outage, and hidden bugs in software designed to protect the service caused its backup systems to fail.
When a company like Amazon, Google or Netflix, which live and die, by staying up and being reliable, can’t escape from bugs and disruptions out of their control; the Olympic site can’t be completely sure either.
But give them credit, they did the tough work up front to do what they could to prevent a full-scale disaster from happening. What happens come Sunday is anybody’s guess, but nobody can say the organizers didn’t do the hard work early on to do everything within their control to ensure things run smoothly throughout the games.