A financial services company runs a transaction processing workload in a dedicated member … — AWS Certified Solutions Architect - Professional

A financial services company runs a transaction processing workload in a dedicated member account of an AWS Organization. The architecture consists of the following components:

- An Amazon ECS cluster running tasks on AWS Fargate across two Availability Zones (AZs) for stateless APIs.
- A fleet of Amazon EC2 instances ( $c6i.2xlarge$ instances: $8\text{ vCPU}$ , $16\text{ GiB}$ memory) in an Auto Scaling group across two AZs for message processing.
- Outbound traffic to external payment gateways routed through redundant NAT Gateways (one per AZ).

Performance monitoring over the past $30\text{ days}$ reveals the following utilization metrics:

Resource Type	Current Specification	Average CPU Utilization	Peak Memory Utilization
ECS Fargate Tasks	Dynamic	$10\%$	$15\%$
EC2 Instances	$c6i.2xlarge$ ( $8\text{ vCPU}$ , $16\text{ GiB}$ memory)	$40\%$	$13.5\text{ GiB}$

The company wants to optimize resource costs while maintaining high availability, ensuring zero application downtime during configuration changes, and securing a $3\text{-year}$ discount commitment. Which of the following strategies should a Solutions Architect recommend to optimize resource costs while meeting all performance, high availability, and operational requirements?

Downsize the Fargate task CPU and memory configurations. Migrate the EC2 instances to $m6i.xlarge$ ( $4\text{ vCPU}$ , $16\text{ GiB}$ memory) using a rolling update in the Auto Scaling group. Purchase a 3-year Compute Savings Plan in the management account of the AWS Organization.Cevap
B
Downsize the Fargate task CPU and memory configurations. Migrate the EC2 instances to $c6i.xlarge$ ( $4\text{ vCPU}$ , $8\text{ GiB}$ memory) using a rolling update in the Auto Scaling group. Purchase a 3-year EC2 Instance Savings Plan for the $c6i$ family in the member account.
C
Downsize the Fargate task CPU and memory configurations. Migrate the EC2 instances to $m6i.xlarge$ ( $4\text{ vCPU}$ , $16\text{ GiB}$ memory). Replace the redundant NAT Gateways with a single NAT Gateway in one Availability Zone to eliminate hourly NAT charges. Purchase a 3-year EC2 Instance Savings Plan for the $m6i$ family in the management account.
D
Downsize the Fargate task CPU and memory configurations. Migrate the EC2 instances to $c6i.xlarge$ ( $4\text{ vCPU}$ , $8\text{ GiB}$ memory). Create a centralized NAT Gateway in a shared services VPC and share it with the member account using AWS Resource Access Manager (RAM). Purchase a 3-year Compute Savings Plan in the management account.

Cevap

Downsize the Fargate task configurations, migrate the EC2 instances to

m6i.xlarge

(

4\text{ vCPU}

16\text{ GiB}

memory) using a rolling update, and purchase a 3-year Compute Savings Plan in the management account of the AWS Organization.

The correct strategy involves downsizing the Fargate tasks to match their low average utilization (

10\%

CPU,

15\%

memory). For the EC2 fleet, because the peak memory utilization is

13.5\text{ GiB}

, downsizing to

c6i.xlarge

(

8\text{ GiB}

memory) is not feasible as it would lead to out-of-memory errors. The best approach is to switch to a general-purpose

m6i.xlarge

instance (

4\text{ vCPU}

16\text{ GiB}

memory), which accommodates both the CPU and memory requirements while lowering cost. To apply discounts to both Fargate and EC2, a Compute Savings Plan must be purchased in the management account. A rolling update (instance refresh) guarantees zero downtime.

Adım Adım Çözüm

Analyze Fargate utilization metrics and downsize configurations.

The Fargate tasks run at

10\%

CPU and

15\%

memory, meaning they are over-provisioned. Downsizing their configurations directly reduces compute costs.

Aligns resource limits with actual workload demands to eliminate waste.

Determine the optimal sizing for the EC2 instances based on peak memory requirements.

The

c6i.2xlarge

instances run at

40\%

average CPU (

3.2\text{ vCPU}

) and

13.5\text{ GiB}

peak memory. Changing to

c6i.xlarge

(

8\text{ GiB}

memory) would cause out-of-memory failures. Transitioning to

m6i.xlarge

(

4\text{ vCPU}

16\text{ GiB}

memory) provides sufficient CPU and memory capacity at a lower cost.

Ensures that performance is not degraded during peaks while optimizing instance costs.

Select the correct Savings Plan type to cover the hybrid Fargate and EC2 environment.

A Compute Savings Plan covers both EC2 instances (regardless of family or region) and AWS Fargate tasks across all member accounts when purchased in the management account. An EC2 Instance Savings Plan would not cover the Fargate tasks and lacks flexibility across families.

Maximizes the discount coverage for the entire organization's compute footprint.

Anahtar Kavram

Compute Resource Right-Sizing and Savings Plans Optimization

Soru

Cevap

Adım Adım Çözüm

Anahtar Kavram

Sık Yapılan Hatalar

Konular

Bu konudan daha fazla soru