Question #883
A solutions architect is evaluating the high availability of an application running on Amazon EC2 instances within a private subnet of a VPC. The EC2 instances are managed by an Auto Scaling group configured with a minimum and maximum capacity of 1. The application uses an Amazon RDS for PostgreSQL DB instance. The VPC includes subnets in three Availability Zones and uses a single NAT gateway. The solutions architect must propose a solution to ensure the application operates reliably across multiple Availability Zones. Which solution meets this requirement?
Deploy NAT gateways in all Availability Zones. Update route tables to direct traffic through the respective NAT gateways. Modify the RDS for PostgreSQL DB instance to a Multi-AZ deployment. Reconfigure the Auto Scaling group to distribute instances across all Availability Zones. Set the Auto Scaling group's minimum and maximum capacity to 3.
Replace the NAT gateway with an internet gateway. Migrate the RDS for PostgreSQL DB instance to Amazon DynamoDB. Configure the Auto Scaling group to launch instances in multiple Availability Zones. Adjust the Auto Scaling group's minimum and maximum capacity to 3.
Deploy a NAT instance in each Availability Zone. Convert the RDS for PostgreSQL DB instance to a Single-AZ configuration. Launch additional EC2 instances manually in the remaining Availability Zones. Set the Auto Scaling group's capacity to 3.
Add NAT gateways in all Availability Zones. Update route tables accordingly. Enable point-in-time recovery for the RDS for PostgreSQL DB instance. Configure the Auto Scaling group to span multiple Availability Zones but retain the minimum and maximum capacity at 1.
Explanation
The correct answer is A because it addresses all critical components for high availability:
1. NAT Gateways: Deploying NAT gateways in all AZs ensures EC2 instances in private subnets can access the internet even if one AZ fails.
2. RDS Multi-AZ: Converting the RDS instance to Multi-AZ provides automatic failover, ensuring database availability.
3. Auto Scaling Group: Scaling the ASG to 3 instances across AZs ensures the application remains available even during AZ outages.
Other options fail because:
- B: Replacing NAT with an internet gateway exposes private instances to the public internet, violating security best practices. Migrating to DynamoDB is unnecessary.
- C: Using NAT instances (not managed gateways) reduces reliability. Manually launching instances bypasses Auto Scaling benefits.
- D: Keeping ASG capacity at 1 does not provide redundancy across AZs, leaving the application vulnerable to AZ failures.
Key Points:
- Use NAT gateways per AZ for private subnets.
- RDS Multi-AZ ensures database HA.
- Auto Scaling groups should span AZs with sufficient instances for redundancy.
Answer
The correct answer is: A