Examkin

Soru

Zorluk: ZorCost-Optimized Compute Selection and Purchasing Strategies

An e-commerce company runs its operations on AWS. The order-processing microservices run continuously on containerized infrastructure using AWS Fargate, requiring a predictable baseline capacity of 16 vCPU16\text{ vCPU} and 64 GB64\text{ GB} of memory, and scaling up during peak hours. In addition, an Amazon RDS for PostgreSQL database instance runs continuously to support the orders. Lastly, a memory-intensive inventory reconciliation batch processing script runs on EC2 instances once daily at midnight. The script takes approximately 3 hours3\text{ hours} to complete, is state-saved, can survive interruptions, and uploads several terabytes of output data directly to Amazon S3. Which combination of strategies should the Solutions Architect recommend to achieve the most cost-effective architecture? (Select TWO.)

  1. Purchase a Compute Savings Plan to cover the continuous baseline capacity of the AWS Fargate order-processing microservices.Cevap
  2. Launch EC2 Spot Instances to run the daily inventory reconciliation batch processing workload.Cevap
  3. C
    Purchase a Compute Savings Plan to cover the continuous baseline capacity of both the AWS Fargate microservices and the Amazon RDS for PostgreSQL database instance.
  4. D
    Migrate the 3-hour inventory reconciliation script to run as a single AWS Lambda function to eliminate compute costs when not running.
  5. E
    Route the output data uploads from the inventory reconciliation EC2 instances to Amazon S3 through a NAT Gateway to reduce data transfer charges.

Cevap

Commit to a Compute Savings Plan to cover the continuous baseline capacity of the Fargate order-processing microservices, and run the fault-tolerant, daily 3-hour inventory reconciliation script on EC2 Spot Instances.
Purchasing a Compute Savings Plan provides discount coverage for Fargate container tasks that run continuously, while using EC2 Spot Instances for the daily inventory reconciliation batch script yields maximum cost savings for an interruption-tolerant, short-running workload.

Adım Adım Çözüm

1
Analyze the workload characteristics of the order-processing microservices.
The order-processing microservices run continuously on AWS Fargate with a predictable baseline capacity.
Since the workload is continuous and running on Fargate, Compute Savings Plans are the most appropriate purchasing model to reduce baseline costs.
2
Analyze the workload characteristics of the inventory reconciliation batch script.
The reconciliation script runs for 3 hours daily, can survive interruptions, is state-saved, and is fault-tolerant.
Because the batch workload is interruption-tolerant and runs for a short duration, EC2 Spot Instances provide the highest cost savings (up to 90% discount) without risking service availability.
3
Evaluate the database and data transfer components for optimization.
The RDS instance is continuous, and large data transfers to S3 occur daily.
Identify that Compute Savings Plans do not apply to RDS (which requires RDS Reserved Instances), and data transfers to S3 should use a free S3 Gateway VPC Endpoint instead of a NAT Gateway.

Anahtar Kavram

Selecting the most cost-effective compute hosting models and purchasing strategies based on workload resilience, predictability, and duration.
Bu soruyu puanla

Konular

Cost-Optimized Compute Selection and Purchasing Strategies