Examkin

Soru

Zorluk: OrtaCost-Optimized Compute Selection and Purchasing Strategies

A company is designing a cost-effective architecture for a new application workload. The workload consists of the following components:

* A web application backend with a constant, predictable baseline of 1010 Amazon EC2 instances running 24/724/7, which scales up to 3030 instances during peak hours.
* An offline data processing pipeline that runs daily for 66 hours, performs idempotent tasks, and can tolerate interruptions.
* A metadata store running on Amazon DynamoDB that receives highly unpredictable, spiky write requests from mobile clients.

Which two strategies should the solutions architect recommend to minimize costs while meeting these requirements? (Choose two.)

  1. Purchase EC2 Instance Savings Plans to cover the baseline of 1010 EC2 instances.Cevap
  2. Use Spot Instances to run the offline data processing pipeline.Cevap
  3. C
    Purchase Compute Savings Plans to cover the Amazon DynamoDB metadata store writes.
  4. D
    Migrate the offline data processing pipeline to AWS Lambda to run continuously for the 66-hour duration.
  5. E
    Configure the Amazon DynamoDB table with Provisioned Capacity mode to handle the spiky write requests.

Cevap

Purchase EC2 Instance Savings Plans to cover the baseline of 1010 EC2 instances, and use Spot Instances to run the offline data processing pipeline.
The correct options are to purchase EC2 Instance Savings Plans for the 1010 baseline EC2 instances, and to use Spot Instances for the interruptible, 66-hour offline data processing pipeline. EC2 Instance Savings Plans offer significant savings for predictable, continuous 24/724/7 compute. Spot Instances offer the lowest cost for batch processing jobs that can tolerate interruptions.

Adım Adım Çözüm

1
Analyze the baseline web application requirements.
The baseline of 1010 EC2 instances runs 24/724/7 continuously and predictably, making it eligible for commitment-based discounts.
EC2 Instance Savings Plans provide up to 72%72\% discount for a steady state commitment.
2
Analyze the offline data processing pipeline requirements.
The pipeline runs for 66 hours daily, performs idempotent tasks, and tolerates interruptions.
Spot Instances are ideal for fault-tolerant, interruptible workloads, providing the highest possible cost savings (up to 90%90\%).
3
Analyze the database/metadata store requirements.
DynamoDB receives unpredictable, spiky write requests.
On-Demand capacity mode should be used to avoid over-provisioning or throttling costs. Compute Savings Plans do not apply to DynamoDB.

Anahtar Kavram

Selecting cost-optimized compute purchasing options (Savings Plans, Spot Instances) based on workload predictability and fault tolerance.
Bu soruyu puanla

Konular

Cost-Optimized Compute Selection and Purchasing Strategies