Examkin

Question

Difficulty: MediumAWS Pricing Models

A research institute is planning a project that will last exactly one year. As part of this project, they need to run a critical data analysis workload on an Amazon EC2 instance for 8 hours every Friday. The analysis cannot be interrupted once it begins. Which EC2 pricing model is the most cost-effective for this workload?

  1. On-Demand InstancesAnswer
  2. B
    Reserved Instances
  3. C
    Spot Instances
  4. D
    Savings Plans

Answer

On-Demand Instances are the most cost-effective model because the workload runs for only 416416 hours total over the year, making any 1-year commitment far more expensive, while Spot Instances cannot guarantee the required uninterrupted execution.
The correct answer is On-Demand Instances. Because the workload runs for only 8 hours a week, the total run time over the year is 416416 hours. Paying the standard hourly rate for On-Demand is far cheaper than committing to a 1-year Reserved Instance or Savings Plan, which charges for all 8,7608,760 hours of the year. Additionally, On-Demand Instances guarantee that the compute resources will not be reclaimed or interrupted, ensuring the analysis runs successfully to completion.

Step-by-Step Solution

1
Calculate the total hours of compute required for the workload over the course of the project.
The workload runs for 8 hours per week for 52 weeks, which equals 8×52=4168 \times 52 = 416 hours.
Determining the total compute hours helps compare actual usage against commitment-based pricing options.
2
Evaluate the cost of commitment-based models (Reserved Instances and Savings Plans) over a 1-year period.
A 1-year commitment requires paying for all hours in a year (24×365=8,76024 \times 365 = 8,760 hours). Even with a typical discount of 30% to 72%, paying for 8,7608,760 hours is significantly more expensive than paying for only 416416 hours of On-Demand compute.
Evaluating commitment models prevents the misconception that a long-term project automatically warrants a long-term billing commitment.
3
Evaluate the viability of Spot Instances based on the workload requirements.
Spot Instances can be terminated by AWS with a 2-minute warning if capacity is needed elsewhere. Because the analysis must complete without interruption, Spot Instances are disqualified.
Ensuring the technical requirements of the application (no interruption) are met before selecting a pricing model.
4
Select the most cost-effective and compliant pricing option.
On-Demand Instances satisfy the zero-interruption requirement and are the most cost-effective choice since the instance is only billed when running (416416 hours total).
Choosing the final pricing model that matches both the business budget constraints and system stability requirements.

Key Concept

Selecting the optimal Amazon EC2 pricing model based on workload patterns, predictability, and duration.
Rate this question

Topics

AWS Pricing Models

More questions from this topic