Examkin

Soru

Zorluk: ZorCost-Optimized Compute Selection and Purchasing Strategies

A logistics company operates a fleet tracking platform on AWS. The application architecture consists of:
1. An API ingest layer hosted on Amazon ECS with AWS Fargate that requires a continuous, predictable baseline of 1616 vCPUs and 32 GB32\text{ GB} of memory 24/724/7.
2. A batch processing workload on Amazon ECS with AWS Fargate that analyzes historical telemetry data. These tasks are triggered daily, can tolerate interruptions, and do not have strict completion SLAs.
3. An Amazon DynamoDB table used to store real-time location metrics, which experiences sudden, highly unpredictable spikes in write traffic.

The company wants to optimize its compute and database costs over a 33-year period.

Which two strategies should a solutions architect recommend to achieve the most cost-effective architecture? (Select TWO.)

  1. Purchase a 33-year Compute Savings Plan to cover the baseline API ingest layer running on AWS Fargate.Cevap
  2. Configure the batch processing workload on ECS to run using the Fargate Spot capacity provider.Cevap
  3. C
    Purchase a 33-year Compute Savings Plan to cover the read and write capacity of the Amazon DynamoDB table.
  4. D
    Configure the Amazon DynamoDB table with Provisioned Capacity Mode and enable Auto Scaling to handle the location metrics.
  5. E
    Migrate the batch processing workload to AWS Lambda functions and configure provisioned concurrency to run them continuously.

Cevap

The correct strategies are purchasing a 33-year Compute Savings Plan for the baseline API ingest layer and configuring the batch processing workload to use the Fargate Spot capacity provider.
The correct strategies are purchasing a 33-year Compute Savings Plan for the baseline API ingest layer and configuring the batch processing workload to use the Fargate Spot capacity provider. A Compute Savings Plan provides significant savings for baseline, predictable compute usage on AWS Fargate. Fargate Spot is the most cost-effective option for non-urgent, interruptible batch processing tasks because it offers up to a 70%70\% discount without requiring any long-term commitments.

Adım Adım Çözüm

1
Analyze the API ingest layer's compute requirements on AWS Fargate.
The workload requires a continuous, steady-state baseline of 1616 vCPUs and 32 GB32\text{ GB} of memory 24/724/7.
For baseline, predictable compute usage over a long duration, a 33-year Compute Savings Plan offers the highest discount (up to 66%66\%) and applies directly to AWS Fargate compute usage.
2
Evaluate the batch processing workload's operational profile.
The batch tasks can tolerate interruptions, run daily, and lack strict completion SLAs.
ECS Fargate Spot is the most cost-effective purchasing option for interruptible workloads, providing up to a 70%70\% discount compared to standard On-Demand Fargate pricing without requiring long-term commitments.
3
Determine the cost-optimization strategy for the Amazon DynamoDB table.
The table experiences sudden, unpredictable spikes in write traffic, making it unsuitable for provisioned capacity or compute savings plans.
For unpredictable, spiky database workloads, DynamoDB On-Demand capacity mode is the most cost-effective choice because it charges per request, avoiding over-provisioning costs or throttling. Compute Savings Plans do not apply to DynamoDB.

Anahtar Kavram

Selecting cost-optimized compute hosting models and purchasing strategies (Compute Savings Plans, Fargate Spot) while understanding their scopes and limits.
Bu soruyu puanla

Konular

Cost-Optimized Compute Selection and Purchasing Strategies