A company runs a critical online transaction processing (OLTP) application on AWS that use… — AWS Certified Solutions Architect - Associate

A company runs a critical online transaction processing (OLTP) application on AWS that uses an Amazon RDS for PostgreSQL DB instance. The database size is $800\text{ GB}$ and requires a Multi-AZ deployment for high availability. The workload is characterized by a steady baseline of $2,500$ write IOPS and $1,500$ read IOPS. Every Sunday, a batch data archiving job runs for $4$ hours, generating an additional peak of $12,000$ read IOPS. The database currently uses Provisioned IOPS SSD (`io2`) storage provisioned at $16,000$ IOPS to ensure performance during the Sunday peak. A solutions architect needs to redesign the database storage and capacity configuration to minimize costs while maintaining high availability and performance. Which configuration is the most cost-effective way to meet these requirements?

A
Modify the primary DB instance to use General Purpose SSD (gp3) storage with $16,000$ provisioned IOPS in a Multi-AZ configuration. Purchase a Compute Savings Plan to apply discounts to the database instance and provisioned storage costs.
Modify the primary DB instance to use General Purpose SSD (gp3) storage with $4,000$ provisioned IOPS. Create a Single-AZ RDS Read Replica using gp3 storage with $12,000$ provisioned IOPS to handle the Sunday archiving workload, and route the archiving queries to the replica.Answer
C
Modify the primary DB instance to use General Purpose SSD (gp3) storage with $4,000$ provisioned IOPS. Create an RDS Read Replica in a different Availability Zone with gp3 storage with $12,000$ provisioned IOPS. Configure the replica as the primary failover target to eliminate the need for the Multi-AZ deployment.
D
Migrate the PostgreSQL database to an Amazon DynamoDB table configured in Provisioned Capacity Mode. Set the write capacity units (WCUs) to $2,500$ and read capacity units (RCUs) to $16,000$ to accommodate the Sunday peak, and enable DynamoDB Auto Scaling.

Answer

Modify the primary DB instance to use General Purpose SSD (gp3) storage with

4,000

provisioned IOPS. Create a Single-AZ RDS Read Replica using gp3 storage with

12,000

provisioned IOPS to handle the Sunday archiving workload, and route the archiving queries to the replica.

The correct configuration uses gp3 storage for the Multi-AZ primary instance, sized only for the OLTP baseline of

4,000

IOPS, and delegates the Sunday read-heavy batch workload to a Single-AZ Read Replica provisioned with

12,000

IOPS. This leverages gp3's low-cost structure and avoids doubling the provisioning cost of the peak

12,000

IOPS over both Multi-AZ instances.

Step-by-Step Solution

Analyze database workload requirements

Identified a steady baseline of

4,000

IOPS (

2,500

writes +

1,500

reads) and a weekly read-only peak of

12,000

IOPS.

Understanding the baseline vs. peak requirements is necessary to size storage and allocate capacity cost-effectively.

Select the optimal storage class

Selected gp3 storage over io2 storage.

gp3 storage provides cost-efficient baseline performance and allows provisioning IOPS independently of storage size, significantly reducing storage costs compared to io2.

Design database replication and capacity allocation

Allocated

4,000

IOPS to the Multi-AZ primary database and offloaded the

12,000

IOPS Sunday archiving workload to a Single-AZ Read Replica.

Since the archiving job is read-only, using a Single-AZ Read Replica avoids paying for peak IOPS on both primary and standby instances in the Multi-AZ configuration.

Key Concept

Cost-optimized database capacity planning involves matching storage types (like gp3 vs. io2) to workload requirements, utilizing read replicas to offload read-heavy batch processes from Multi-AZ deployments, and understanding the billing scope of compute savings options.

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics