SAP-C03 practice question: Fault-tolerant and self-healing architectures
Resilience, Migration, and Business Continuity A free question from our original SAP-C03 practice exam. Try it, then open the answer to see why each option is right or wrong.
A retailer runs its product API on Amazon EC2 instances across three Availability Zones behind an Application Load Balancer. Instances occasionally fail, and the team currently replaces them by hand after an alert.
The team wants failed instances replaced automatically, with the capacity kept across all three zones. The API must keep no session state on the instances, because a user's request may reach any instance.
Which solution meets these requirements?
Choose one answer.
AA. Place the instances in an Auto Scaling group across the three zones with load balancer health checks, and store sessions in ElastiCache so that any instance can serve any request.
BB. Run one large instance with an hourly EBS snapshot, and restore from the latest snapshot by hand when the instance fails.
CC. Place the instances in an Auto Scaling group in one zone with EC2 status checks only, and keep sessions in instance memory with sticky sessions enabled on the load balancer's target group.
DD. Add more instances in each zone and keep the spare ones stopped, and start a stopped instance by hand whenever an alert fires.
Show the answer and explanation
Answer: A
A. Right: Correct. The Auto Scaling group replaces unhealthy instances and keeps capacity spread across the zones, and ElastiCache holds the sessions outside the instances. Any instance can then serve any request.
B. Wrong: A single instance is a single point of failure, and a manual restore from an hourly snapshot leaves an outage and up to an hour of changes at risk.
C. Wrong: One zone cannot survive that zone's failure, and status checks miss application failures that the load balancer health checks detect. Sticky sessions in memory also tie users to failing instances.
D. Wrong: Stopped spares need a person to start them, which keeps the manual replacement the team wants to remove. Stopped instances also take time to become serving capacity.
Explanation
An Auto Scaling group across three zones with load balancer health checks replaces failed instances automatically, and ElastiCache keeps sessions off the instances. A single instance, one-zone group with sticky sessions, and manual stopped spares each fail the automatic recovery or the statelessness need.
Want more on Fault-tolerant and self-healing architectures? The 5 full-length practice sets have 6 more Fault-tolerant and self-healing architectures questions, 375 in all, none of them in the free exam. One payment of $14.99, no subscription. See the practice sets →