A university research department is transitioning its scientific simulation workloads to A… — AWS Certified Cloud Practitioner

A university research department is transitioning its scientific simulation workloads to AWS. These workloads run for approximately $12$ hours at the end of each month, requiring $64$ high-performance compute nodes. The rest of the month, the department only requires a single low-power server to host their results archive. The IT director is evaluating the economic impact of this migration compared to their legacy on-premises hardware.

Which of the following describes the most cost-effective provisioning strategy that illustrates the economic benefit of cloud elasticity for this workload?

Deploying $64$ On-Demand Instances only for the $12$ hours the simulations run and terminating them immediately afterward, while running a single low-power instance continuously.Answer
B
Purchasing a $3$ -year Reserved Instance contract for $64$ instances to guarantee that compute capacity is permanently available and reserved for the monthly simulation runs.
C
Capitalizing the virtual hardware costs upfront as a capital expense (CapEx) to establish a baseline of $64$ active instances, then converting the idle capacity into operating expenses (OpEx) when they are not in use.
D
Statically provisioning $64$ instances to remain active throughout the entire month to ensure the system has the scalability to handle the peak, which represents the AWS definition of elasticity.

Answer

Deploying

64

On-Demand Instances only for the

12

hours the simulations run and terminating them immediately afterward, while running a single low-power instance continuously.

The correct answer demonstrates elasticity by running the high-performance compute nodes only when active (

12

hours a month) and terminating them when finished. This ensures the university only pays for what it uses, avoiding the cost of idle servers.

Step-by-Step Solution

Analyze the workload's resource usage pattern over time.

The workload has a highly fluctuating demand pattern: it requires

64

compute nodes for only

12

hours per month (approx.

1.6\%

of the month) and just a single node for the remaining

98.4\%

of the time.

Identifying the demand pattern is necessary to choose between static provisioning and elastic, dynamic provisioning.

Evaluate the economic feasibility of AWS pricing models for the intermittent portion of the workload.

Using On-Demand instances for the

64

simulation nodes costs only for the

12

hours of actual use, whereas Reserved Instances or static provisioning would charge for the entire month regardless of utilization.

This step determines how to minimize waste for resources that are idle most of the month.

Formulate a hybrid provisioning strategy that matches the workload segments.

Run a single low-power instance continuously for the archive, and spin up

64

On-Demand instances only when the monthly simulation runs, terminating them immediately upon completion.

This strategy leverages cloud elasticity to optimize cost by dynamically aligning compute supply with workload demand.

Key Concept

Cloud elasticity vs. static provisioning, and the pay-as-you-go consumption model.

Estimated Time:2m 0s

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics

More questions from this topic