Don’t Let Your Site Get Sacked On Super Bowl Sunday

Super Bowl Sunday is fast approaching. Companies pay millions of dollars to get that coveted 30 second commercial spot—and that doesn’t account for the cost to create the ads. For many big name companies, Super Bowl Sunday breeds big traffic numbers. It’s imperative that their websites and apps can stand up to the traffic blitz that this global event implies.

Surprisingly, not all companies are prepared. Last year, 13 websites suffered outages during the big game. Think about that: that’s 13 of the world’s largest companies and their websites couldn’t tackle the volume of traffic sent their way.

Learn more about how to prep your site for a Super Bowl traffic blitz.

Let’s face it; one of the biggest killers of applications and websites is success. We see it all the time. Super Bowl Sunday is one example. But high traffic events knock over websites every day. Whether it’s a celebrity photographed donning a new pair of leggings from a relatively unheard of designer or a tech guru endorses a new device— BAM!—those sites can be taken down.

When Kate Middleton and Prince William brought Prince George home from the hospital, the baby was swaddled in an Aden + Anais blanket. Once the photos surfaced, more than 10,000 orders were placed nearly simultaneously, causing the Aden + Anais site to suffer a royal crash, resulting in a significant missed business opportunity.

The problem is, forecasting a massive traffic spike is tricky, especially around marketing campaigns or big events. Oftentimes, the ones that are most successful go better than planned and the flood of traffic is immense. The spikes can be so severe that they appear like an application layer DDoS attack. Most companies size for daily traffic and can scale up for peak loads, but if those peek loads are exceeded, kiss your site goodbye.

One common mistake is tuning an application at the page level, but that often doesn’t account for how a specific site will respond to increased or unpredictable load and what the customer will experience. Another common problem is fine-tuning the data tier and not the application tier, which is typically what goes down when a site gets knocked over. Putting a load balancer in front of that application tier can help better distribute the load.

An application and traffic management solution (like Lagrange Systems’ CloudMaestro) is also an option. It segments workloads and augments them with dynamic scaling — meaning application server capacity can scale to new servers before or as a massive traffic spike hits. It’s fluid—increasing the number of servers when the demand is there, and decreasing them when it’s gone.

In the end, people aren’t fast enough. An infrastructure that automatically scales for you on Super Bowl Sunday or for any high traffic event is critical.

This is a guest post written and contributed by Jay Smith, CTO and Co-Founder of Lagrange Systems, a Rackspace Cloud Tools partner. Lagrange Systems makes CloudMaestro™, a SaaS SDN software solution designed for true and automatically provisioned high availability.


Please enter your comment!
Please enter your name here