You are the lead enterprise architect for a retail company. Traditionally, you have planned for additional capacity in anticipation of increase load on your environment during busy shopping periods and event - typically centered around holidays.

Your company has made the decision to move its environment to Azure. However, instead of deploying additional capacity into your cloud environment (i.e. more VMs), you want to take advantage of solutions such as VM availability sets and scale sets in order to guarantee maximum availability and performance under increased load. The majority of your environment supporting your customers and operations consists of front-end/web servers, application servers, and database servers.

How would you leverage availability and scale sets across these three groups of servers to guarantee system uptime and performance under increase load? Would you use one, another, or both to accomplish this? Through availability and scale sets, can you describe what the benefit would be over simply deploying more VMs to the environment?