An IT associate has a web application running on multiple servers, configured to scale based on CPU usage. The current configuration has a minimum size of 1 and a maximum size of 5 servers, designed…

Increasing the minimum instance count to 2 ensures that at least two instances are always running. With instances spread across 3 AZs, if one AZ goes down, the remaining two AZs will each likely still have at least one instance, maintaining application availability instead of potentially leaving only one instance in a single AZ which could fail, or none if that single instance was in the failed AZ.

Changing the scaling metric might optimize for performance under load but does not directly address the issue of maintaining availability during an AZ outage when operating at minimum capacity.

Decreasing the number of Availability Zones to 2 would reduce redundancy and further decrease availability, making the application more susceptible to outages if one of the two remaining AZs fails.

Increasing the maximum server count addresses scaling for high demand, but it does not guarantee availability during low utilization or when an AZ fails if the minimum instance count remains at 1.