Which steps should the solutions architect take to meet these requirements?
(Choose three.)
Create multiple read replicas and put them into an Auto Scaling group.
Create multiple read replicas in different Availability Zones.
Create an Amazon Route 53 hosted zone and a record set for each read replica with a TTL and a weighted routing policy.
Create an Application Load Balancer (ALB) and put the read replicas behind the ALB.
Configure an Amazon CloudWatch alarm to detect a failed read replicas. Set the alarm to directly invoke an AWS Lambda function to delete its Route 53 record set.
Configure an Amazon Route 53 health check for each read replica using its endpoint.
Explanations:
Creating multiple read replicas in different Availability Zones ensures high availability and distributes read traffic across multiple instances. This setup improves the scalability and availability of the application.
Using Amazon Route 53 with a weighted routing policy for each read replica allows traffic to be directed to different replicas based on the specified weight. This helps balance the load among the replicas.
Configuring an Amazon Route 53 health check for each read replica ensures that traffic is only directed to healthy replicas. If a replica fails, Route 53 can stop routing traffic to it.
Creating multiple read replicas and putting them into an Auto Scaling group is not applicable to Amazon RDS for PostgreSQL. Auto Scaling groups do not support RDS instances.
An Application Load Balancer (ALB) cannot be used to distribute traffic to RDS read replicas because ALBs are designed for HTTP/HTTPS traffic, not database connections.
While CloudWatch alarms can detect failure in read replicas, invoking an AWS Lambda function to delete a Route 53 record set is unnecessary and would not resolve the underlying issue of read replica failure.
I opine that the answer is:
Create multiple read replicas in different Availability Zones.