SAP-C02 Question 468
Select 3A company is migrating a mission-critical application to AWS and has strict uptime requirements of 99.99%. The application consists of a multi-tier architecture with an Application Load Balancer (ALB), Amazon EC2 instances in an Auto Scaling group, and an Amazon RDS database. To meet their Service Level Agreement (SLA) for uptime, which combination of actions should the Solutions Architect recommend?
- A
Deploy the EC2 instances in multiple Availability Zones (AZs) within the same region.
- B
Configure the RDS database with Multi-AZ deployment.
- C
Use AWS Trusted Advisor to monitor and resolve high-risk issues related to availability.
- D
Enable Auto Scaling to ensure the application can handle traffic spikes.
- E
Deploy a read replica for the RDS database in a different region for disaster recovery.
Show answer and explanation
Correct answers: A, B, D
Explanation
To meet a strict SLA of 99.99%, high availability and fault tolerance must be ensured across all tiers of the architecture. Deploying EC2 instances across multiple AZs and configuring the RDS database with Multi-AZ deployment ensures redundancy and failover capabilities. Auto Scaling ensures the application can dynamically handle changes in traffic, contributing to availability. While Trusted Advisor and read replicas are useful for monitoring and disaster recovery, they do not directly guarantee the specified SLA.
- A. Correct.
Deploying EC2 instances in multiple Availability Zones ensures high availability and fault tolerance in case one AZ goes down. This directly contributes to meeting the uptime SLA of 99.99%.
- B. Correct.
Configuring the RDS database with Multi-AZ deployment ensures automatic failover in case of a primary database failure. This aligns with the uptime SLA requirements.
- C. Incorrect.
While AWS Trusted Advisor provides valuable recommendations, it does not actively ensure 99.99% uptime. It is a monitoring tool, not a direct measure to meet an SLA.
- D. Correct.
Enabling Auto Scaling allows the application to handle traffic spikes by scaling resources dynamically, improving availability and resilience.
- E. Incorrect.
Deploying a read replica in a different region is useful for disaster recovery and latency improvements but does not directly contribute to achieving 99.99% uptime within a single region.