Uptime is very important with any computer system, because a solution which is not available cannot be relied upon to provide accurate information in a timely and routine manner - usually resulting in the loss of time, revenue, or data. Highly-available systems and websites rely on eliminating single points of failure, and on reacting quickly to any network, hardware or software issues which may occur. In this paper, we present a cost-effective website failover solution, based on using a Content Delivery Network as a proxy, and asynchronous replication of data across servers hosted on distant locations. We outline several scenarios where this approach is favorable to load balancers, failover IP addresses, and other solutions currently used to design highly-available systems. At the end, we present a case study where we show the dramatic reduce of expenses with the proposed solution.