Examkin

Question

Difficulty: MediumMulti-AZ, Multi-Region Architectures and Disaster Recovery (DR)

An online auction platform operates its primary application in the eu-west-1 Region. The architecture consists of an Application Load Balancer (ALB), an Auto Scaling group of Amazon EC2 instances, and an Amazon RDS for MySQL database. The company needs to design a disaster recovery (DR) strategy in the eu-central-1 Region. The DR plan must support a Recovery Point Objective (RPO) of 15 minutes and a Recovery Time Objective (RTO) of 2 hours while minimizing ongoing running costs. Which strategy should a solutions architect recommend to meet these requirements?

  1. A
    Deploy a scaled-down but running application tier in the secondary region, consisting of an Application Load Balancer and an Auto Scaling group with a minimum capacity of 1. Configure a cross-region Amazon RDS Read Replica in the secondary region. In a disaster recovery event, promote the Read Replica and scale out the Auto Scaling group.
  2. B
    Configure Amazon RDS Multi-AZ replication to replicate the database to the secondary region. Set up an Amazon Route 53 latency routing policy to automatically fail over and direct user traffic to the secondary region during an outage.
  3. Configure a cross-region Amazon RDS Read Replica in the secondary region. Store the application tier infrastructure as AWS CloudFormation templates in the secondary region, maintaining the Auto Scaling group with a desired capacity of zero. In a disaster recovery event, promote the Read Replica to a standalone DB instance and deploy the CloudFormation templates to provision the compute resources.Answer
  4. D
    Create snapshots of the Amazon RDS DB instance every 15 minutes, copy them to the secondary region, and immediately transition them to Amazon S3 Glacier Flexible Retrieval to minimize storage costs. In a disaster recovery event, initiate a standard retrieval of the snapshot to restore the database.

Answer

Configure a cross-region Amazon RDS Read Replica in the secondary region. Store the application tier infrastructure as AWS CloudFormation templates in the secondary region, maintaining the Auto Scaling group with a desired capacity of zero. In a disaster recovery event, promote the Read Replica to a standalone DB instance and deploy the CloudFormation templates to provision the compute resources.
The correct strategy represents a Pilot Light architecture. By utilizing a cross-region Amazon RDS Read Replica, database updates are constantly replicated to the DR region, satisfying the 15-minute RPO. Maintaining the compute resources (ALB and EC2 Auto Scaling group) as CloudFormation templates and keeping the Auto Scaling group desired capacity at zero minimizes running costs since there are no active computing charges. In the event of a disaster, provisioning the compute stack and promoting the DB replica can easily be completed within the 2-hour RTO.

Step-by-Step Solution

1
Analyze RPO requirements.
The target RPO is 15 minutes. To prevent losing more than 15 minutes of transactional data, replication to the target region must be continuous. An RDS cross-region Read Replica uses asynchronous replication and satisfies this requirement.
Choosing a backup-and-restore option with manual snapshot copies is operationally heavy and risks violating RPO if replication/copies are delayed.
2
Analyze RTO and cost requirements.
The target RTO is 2 hours. Deploying compute resources using AWS CloudFormation templates and scaling up an Auto Scaling group takes minutes, easily fitting within the 2-hour window. This allows us to keep the Auto Scaling group capacity at zero in the DR region.
This is a Pilot Light strategy. Keeping the application load balancer and compute instances offline when not in use minimizes running costs.
3
Evaluate and eliminate sub-optimal strategies.
Warm Standby is rejected because keeping active compute resources running continuously in the secondary region is more expensive. S3 Glacier is rejected because retrieval delays violate the RTO. Multi-AZ database configurations are rejected because they do not span multiple AWS regions.
Ensuring the selected solution is both technically viable and the most cost-efficient choice among the options.

Key Concept

Disaster Recovery (DR) Strategies: Pilot Light vs. Warm Standby
Estimated Time:2m 0s
Rate this question

Topics

Multi-AZ, Multi-Region Architectures and Disaster Recovery (DR)

More questions from this topic