Cloud economics: CLF-C02 study notes and practice questions
What to know about Cloud economics for the CLF-C02 exam, from the Cloud Concepts domain (24% of the exam guide), then 4 free questions to test it. Part of the CLF-C02 study guide.
Cloud economics questions ask why AWS can cost less than running your own data center, and which choices change the bill.
Key ideas
Fixed vs variable costs. On premises, you pay upfront for servers, buildings, power and cooling whether you use them or not (capital expense). On AWS you pay for what you use each month (operational expense).
Total cost of ownership (TCO) includes more than servers: facilities, power, cooling, hardware refreshes, and the staff time to run it all. AWS shifts much of that to AWS.
Operational savings. Managed services remove work like patching databases, so staff spend time on the business instead.
Right-sizing means choosing the smallest resource that meets the need, and checking again later.
Automation saves money by turning resources off when no one needs them (for example, stopping development servers at night) and creating environments only when needed with CloudFormation.
Licensing
Option
Meaning
License included
The software license is part of the AWS price (for example, Windows Server on EC2)
Bring your own license (BYOL)
Reuse licenses you already own; some licenses are tied to physical cores or sockets, which is why Dedicated Hosts exist
Economies of scale
Because AWS buys and runs hardware for millions of customers, its costs per unit are lower than any single company’s, and it has lowered prices many times.
Common traps
Thinking cloud is always cheaper. It is cheaper when you use its strengths: elasticity, managed services and pay-as-you-go. A server left running all the time at full size can cost more than needed.
Capital vs operational expense. Buying servers is capital expense; paying AWS monthly is operational expense.
Dedicated Hosts vs Dedicated Instances. Only Dedicated Hosts let you see and use the physical sockets and cores that per-socket licenses need.
A company licenses a database product per physical CPU socket and wants to bring those licenses to AWS.
Its license terms require it to show how many sockets and cores each server uses, so it needs visibility of the physical hardware it runs on.
Which Amazon EC2 option meets both requirements?
Choose one answer.
AA. Spot Instances, which use spare EC2 capacity at a discounted price
BB. Dedicated Hosts, which are physical servers fully dedicated to your use
CC. Dedicated Instances, which run on hardware dedicated to your AWS account
DD. License-included Amazon Machine Images, with the license in the hourly price
Show the answer and explanation
Answer: B
A. Wrong: Spot Instances use spare EC2 capacity at a discount, but AWS can interrupt them and the customer does not choose the physical hardware. They give no visibility of sockets or cores.
B. Right: Correct. Dedicated Hosts are physical servers fully dedicated to one customer. They support bring-your-own-license for per-socket and per-core licenses and show the number of sockets and cores. Learn more: Amazon EC2 Dedicated Hosts (opens in a new tab)
C. Wrong: Dedicated Instances run on hardware dedicated to one account, which is a real option. However, they give no visibility of sockets or cores and only partial BYOL support, so they miss the license-reporting requirement.
D. Wrong: License-included AMIs bundle the software license into the instance price, so the company would pay AWS for licenses it already owns. That does not reuse its per-socket licenses or show hardware use.
Explanation
Dedicated Hosts are physical servers dedicated to one customer. They let the company bring its existing per-socket licenses and see the sockets and cores on each host. Dedicated Instances isolate hardware but hide sockets and cores, while Spot Instances and license-included images do not meet the requirement.
A company's development and test environments run around the clock, but its engineers use them only during weekday working hours.
The finance lead wants to cut the monthly AWS bill by automating how these environments are created and run, rather than relying on manual reviews.
Which TWO actions achieve this? (Choose two.)
Choose two answers.
AA. Use a schedule to stop development and test instances automatically outside working hours.
BB. Increase the size of every instance so that engineers need to tune performance less often.
CC. Turn off AWS CloudTrail in every account to save on log storage.
DD. Create test environments from AWS CloudFormation templates and delete each stack when testing ends.
EE. Buy three-year Reserved Instances for test servers that run only during the working day.
Show the answer and explanation
Answers: A, D
A. Right: Correct. An automated schedule stops instances outside working hours and starts them again in the morning. The company stops paying for compute nobody is using, with no manual effort each day.
B. Wrong: Incorrect. Larger instances cost more per hour, so this raises the bill. Tuning performance less often does not reduce cost.
C. Wrong: Incorrect. CloudTrail records API activity for auditing and security, and its storage cost is small. Turning it off to save money removes visibility for very little saving.
D. Right: Correct. CloudFormation templates build an environment the same way each time, and deleting the stack when testing ends stops charges for resources nobody is using.
E. Wrong: Incorrect. A Reserved Instance is billed at its discounted rate for every hour of its term, whether or not it runs. Committing for three years to servers that run only in working hours pays for idle time.
Explanation
Scheduling stops and starts non-production instances automatically, and CloudFormation lets teams create and remove environments on demand, so charges stop when the work stops. Larger instances and disabling audit logs raise cost or remove visibility, and reserved capacity that sits idle wastes money.
A company keeps 80 TB of product images on a file server that it runs on Amazon EC2.
Its administrators spend most of their time adding disks, planning capacity and copying backups.
The company wants to store the images without buying or managing storage servers, and to pay only for the data it stores.
Which option is the best fit?
Choose one answer.
AA. Larger Amazon EBS volumes attached to the same EC2 instance, with more disks added as needed
BB. Amazon S3, which needs no capacity bought in advance and charges for what is used
CC. EC2 instance store volumes on the same instances, which give fast local disks
DD. A second file server on EC2 in another Region, copied from the first
Show the answer and explanation
Answer: B
A. Wrong: Larger EBS volumes still attach to one instance that the company runs and manages. Its administrators would keep adding disks and planning capacity, so this does not remove the effort.
B. Right: Correct. Amazon S3 is an object storage service. Customers do not have to purchase a predetermined amount of storage in advance, and they pay for what they store, so no storage servers need to be managed. Learn more: What is Amazon S3? (opens in a new tab)
C. Wrong: Instance store is temporary block storage attached to the host, meant for buffers, caches and scratch data. It is not a durable home for product images, and the company would still run the servers.
D. Wrong: A second EC2 file server in another Region doubles the servers the company must run and copy backups for. It does not remove the management work the scenario describes.
Explanation
Amazon S3 is designed for storing large amounts of data as objects without the customer provisioning capacity or running storage servers, and it charges for what is used. EBS, instance store and a second EC2 file server all keep the company responsible for the disks and servers it runs.
B. Wrong: Stopping guessing about capacity is about matching resources to demand instead of forecasting it. It does not describe replacing upfront purchases with usage-based spending.
C. Wrong: Going global in minutes is about deploying applications in many Regions quickly, for latency and reach. It does not address up-front spending or costs that follow usage.
D. Wrong: Speed and agility is about provisioning resources quickly so projects move faster. It does not describe the shift from fixed purchases to usage-based costs.
Explanation
The benefit is trading fixed expense for variable expense. The start-up avoids large upfront purchases and pays for IT as it uses it, so costs follow usage. The other three are real cloud benefits, but each addresses speed, reach or capacity planning instead.
Want more on Cloud economics? The 5 full-length practice sets have 24 more Cloud economics questions, 325 in all, none of them in the free exam. One payment of $14.99, no subscription. See the practice sets →