Question

Difficulty: MediumAWS Pricing Models

A research institute runs a compute-intensive weather modeling simulation once a month. The simulation is a batch processing job designed to be fully fault-tolerant, frequently saving checkpoints to Amazon S3 so it can resume if interrupted. The simulation does not have a strict completion deadline, and the institute wants to minimize computing costs. Which Amazon EC2 pricing model is the most cost-effective choice for this workload?

  1. Spot InstancesAnswer
  2. B
    On-Demand Instances
  3. C
    Reserved Instances
  4. D
    Dedicated Hosts

Answer

Spot Instances
Spot Instances are the most cost-effective choice because they allow users to take advantage of unused EC2 capacity at steep discounts (up to 90%). This is ideal for fault-tolerant, flexible batch workloads that can be interrupted and resumed without loss of progress.

Step-by-Step Solution

1
Analyze the workload characteristics from the scenario.
The simulation workload is run once a month, is fully fault-tolerant, frequently saves progress, and has no strict completion deadline.
Understanding the tolerance for interruptions and the frequency of the workload helps narrow down the most cost-effective pricing model.
2
Evaluate the EC2 pricing options against the workload needs.
Spot Instances are ideal for fault-tolerant batch jobs, Reserved Instances are for 24/7 steady-state workloads, and On-Demand is for non-interruptible workloads.
Selecting the option that matches the fault-tolerance and non-continuous nature of the simulation ensures the lowest cost.

Key Concept

AWS Pricing Models
Rate this question