If an Availability Zone fails, how can the company remain compliant with the SLA?
Add a target tracking scaling policy with a short cooldown period.
Change the Auto Scaling group launch configuration to use a larger instance type.
Change the Auto Scaling group to use six servers across three Availability Zones.
Change the Auto Scaling group to use eight servers across two Availability Zones.
Explanations:
While adding a target tracking scaling policy may help maintain instance count during normal operation, it does not guarantee compliance with the SLA if an Availability Zone fails. The minimum instance requirement needs to be built into the Auto Scaling group configuration.
Changing the Auto Scaling group launch configuration to use a larger instance type does not address the requirement for maintaining a minimum of four instances during an Availability Zone failure. It may improve performance, but it does not ensure compliance with the SLA.
Changing the Auto Scaling group to use six servers across three Availability Zones ensures that even if one Availability Zone fails, there will still be at least four instances remaining operational (2 in one AZ and 2 in another), thus meeting the SLA requirements.
While using eight servers across two Availability Zones may provide redundancy, if one Availability Zone fails, only four instances would remain operational, which does not meet the SLA requirement of having a minimum of four instances.
I appraise that the answer is:
Change the Auto Scaling group to use six servers across three Availability Zones.