Moving from single-homed to multi-homed environments can be challenging. The reasons that enterprises move are varied. Some move for performance reasons. Some move to avoid potential outages. Some move because of actual outages. The complexity of your application space has a lot to do with how easy or hard such a move is. A great example of a forward thinking company that wanted performance and 100% availability is our friends at Spir. Spir Communications, an advertising publications company, was experiencing dramatic growth. The company’s website properties, an important part of its revenue, received millions of visits each month. A fast paced industry combined with a company achieving great success – a natural challenge for an operations team.
Hosted in a single data center, their digital services sometimes suffered from slow load times, congestion, and even occasional service outages of up to 30 minutes. It was clearly impacting their bottom line. Spir needed to be able to scale their online classified ads while providing the fast-loading pages customers expected. They set the goal of 100% uptime and top performance. To meet this goal, Spir developed a multi-homed strategy they called Silo. Each Silo consisted of several load-balanced front-end servers and a local read-only database server. Silos were distributed across multiple data centers to improve application availability, and additional Silos could be deployed to meet demand.
To learn how Spir not only gained 100% uptime using Cedexis to load balance their Silo architecture, but further extended performance gains by utilizing a CDN – go check out this case study.