BABassel AasiCloud & DevOps Engineer
Training / Certifications / AWS Cloud Practitioner / Cloud Service Models and Deployment Models

AWS Cloud Practitioner (CLF-C02) · Lesson 2 of 26

Cloud Service Models and Deployment Models

Compare IaaS, PaaS and SaaS by what you manage versus what the provider manages, and identify cloud, hybrid and on-premises deployment models.

Domain 3: Cloud Technology and Services (34%)Task 3.1, 1.1About 4:55CaptionsStatic view for reduced motion
Lesson 2: Service models and deployment models Illustrative example How much do you manage?Service models:IaaS, PaaS and SaaSWhere does it run?Deployment models:cloud, hybrid, on-premisesOn-premisesYour own data centerApplicationDataRuntimeOperating systemVirtualizationServers and storageNetwork and data centerIaaSExample: Amazon EC2ApplicationDataRuntimeOperating systemVirtualizationServers and storageNetwork and data centerPaaSExample: AWS Elastic BeanstalkApplicationDataRuntimeOperating systemVirtualizationServers and storageNetwork and data centerSaaSExample: web-based emailApplicationDataRuntimeOperating systemVirtualizationServers and storageNetwork and data centerYou still control your dataand who can access itYou manageAWS (or the SaaS provider) managesMore control, more to manageLess to manage, less controlCloudAdd to cartEverything runs in AWS(built there or migrated)HybridOn premisesAWSConnected by a private link(VPN or AWS Direct Connect)On-premisesYour own data center(“private cloud”)If the question says……think“Full control of the operating system”IaaS (e.g. Amazon EC2)“Just deploy code, no servers to manage”PaaS (e.g. Elastic Beanstalk)“Ready-to-use software in a browser”SaaS“Some systems stay on premises, linked to AWS”Hybrid deployment“Everything runs in AWS”Cloud deployment“Virtualization in our own data center”On-premises (private cloud)Cloudall in AWSHybridon premises + AWSOn-premisesown data center

Tip: turn your phone sideways or use full screen for larger labels.

Captions match the narration word for word. Press Space to play or pause and the arrow keys to change scenes.

Learning objective

Compare IaaS, PaaS and SaaS by what you manage versus what the provider manages, and identify cloud, hybrid and on-premises deployment models.

Exam mapping

Domain 3: Cloud Technology and Services (34%). Task statement 3.1, 1.1 in the official exam guide.

Key takeaways

  • IaaS: AWS manages the data center, network, servers, storage and virtualization; you manage the operating system and everything above it (for example, Amazon EC2).
  • PaaS: the provider also manages the operating system and runtime; you bring your code and data (for example, AWS Elastic Beanstalk).
  • SaaS: the provider runs the complete application; you manage your data, settings and user access.
  • From IaaS to PaaS to SaaS, you manage less but also control less.
  • Cloud deployment: everything runs in the cloud. Hybrid: cloud resources connected to on-premises infrastructure. On-premises (private cloud): everything in your own data center.

Exam tips

How the exam asks about this

Questions describe a situation and expect you to recognize the concept behind it.

If the question mentions…Think
Full control of the operating system and installed softwareIaaS (for example, Amazon EC2)
Upload code; the service handles servers, patching and scalingPaaS (for example, AWS Elastic Beanstalk)
Complete, ready-to-use application run by a vendorSaaS
Keep some systems in our data center and connect them to AWSHybrid deployment
Run every part of the application in AWSCloud deployment
Virtualization and resource management tools in our own data centerOn-premises (private cloud)

Common traps

  • Amazon VPC stands for virtual private cloud, but it's a network inside AWS. Using it is still a cloud deployment, not an on-premises “private cloud”.
  • Serverless services such as AWS Lambda run your code without servers to manage. That's not SaaS: SaaS is finished software you use, not code you write.
  • Hybrid means on-premises plus cloud, connected together. It doesn't just mean using two cloud providers.
  • Even with SaaS, your data and who can access it remain your responsibility.

Practice

Exam-style questions

Answer each one before you open the explanation.

Question 1. A company needs to run a legacy application that requires a specific, customized operating system configuration. It doesn't want to buy hardware. Which cloud service model fits best?

  1. SaaS
  2. PaaS
  3. IaaS
  4. On-premises deployment
Show answer

C. IaaS. IaaS, such as Amazon EC2, gives full control of the operating system without owning hardware. PaaS and SaaS hide the operating system from you.

Question 2. Developers want to upload their web application code and have AWS handle capacity provisioning, load balancing, scaling and health monitoring. Which service matches this?

  1. Amazon EC2
  2. AWS Elastic Beanstalk
  3. Amazon S3
  4. AWS Outposts
Show answer

B. AWS Elastic Beanstalk. Elastic Beanstalk is a platform service: you provide the code, and it provisions and manages the environment that runs it.

Question 3. A company keeps its customer database in its own data center for regulatory reasons and runs its website on AWS, connected to that database over a private link. Which deployment model is this?

  1. Cloud
  2. Hybrid
  3. On-premises
  4. SaaS
Show answer

B. Hybrid. Connecting cloud resources to infrastructure that stays on premises is a hybrid deployment.

Question 4. In which service model does the customer manage the fewest layers of the stack?

  1. IaaS
  2. PaaS
  3. SaaS
  4. On-premises
Show answer

C. SaaS. With SaaS, the provider runs the whole application. The customer manages its data, settings and user access.

Question 5. A company runs its application on Amazon EC2. Which TWO of these does the company manage? (Select TWO.)

  1. The guest operating system, including patches
  2. The physical servers
  3. The application installed on the instance
  4. Cooling in the data center
  5. The virtualization layer (hypervisor)
Show answer

A. The guest operating system, including patches and C. The application installed on the instance. EC2 is IaaS: AWS manages the physical servers, data center and hypervisor; the customer manages the operating system and everything installed on it.

Transcript

Lesson transcript

The full narration, scene by scene. The captions file contains the same text with timings.

Scene 1 · 0:00

Lesson 2: Service models and deployment models. Welcome back. In lesson one, we defined cloud computing. Now we'll answer two questions that come up again and again on the exam. First: how much of the technology stack do you manage, and how much does the provider manage? That's the service model. Second: where do your workloads run: in the cloud, in your own data center, or both? That's the deployment model.

Scene 2 · 0:28

The stack behind every application. Every application sits on a stack of layers. At the bottom are the network and the data center building, then servers and storage, virtualization, the operating system, the runtime your code needs, your data, and finally the application itself. When the online store runs in its own data center, it manages every one of these layers, from power and cooling all the way up to its own code.

Scene 3 · 0:58

IaaS: rent the infrastructure. With infrastructure as a service, or IaaS, AWS manages the physical layers: the data center, the network, the servers and storage, and the virtualization that turns them into virtual machines. You manage everything from the operating system up. Amazon EC2 is the classic example. The store can rent virtual servers in minutes and still keep full control of the operating system and software.

Scene 4 · 1:25

PaaS: bring your code. With platform as a service, or PaaS, the provider also manages the operating system and the runtime. You bring your application and your data, and the platform handles provisioning, patching and scaling. AWS Elastic Beanstalk works this way: the store uploads its code, and Elastic Beanstalk runs it. You give up some control over the servers, but there's far less to manage.

Scene 5 · 1:52

SaaS: use finished software. With software as a service, or SaaS, you use a finished application that the provider runs completely, usually in a web browser. Think of web-based email or customer relationship management software. The store just signs in and uses it. There are no servers, patches or code to manage. You do still control your own data, and who in your company can access it.

Scene 6 · 2:20

More control, or less to manage?. Here's the pattern to remember. Moving from left to right, the provider manages more and you manage less. IaaS gives you the most control and the most responsibility. SaaS gives you the least to manage, but also the least control. PaaS sits in the middle. The right choice depends on how much control a workload really needs.

Scene 7 · 2:45

Deployment model 1: cloud. Now, where do workloads run? There are three deployment models. In a cloud deployment, every part of the application runs in the cloud. Applications are either built there or migrated there from existing infrastructure. If the store moves its website, database and file storage to AWS and closes its server room, that's a cloud deployment.

Scene 8 · 3:09

Deployment model 2: hybrid. A hybrid deployment connects cloud resources with infrastructure that isn't in the cloud, usually an existing data center, over a private network link. Companies choose hybrid when legacy systems are hard to move, when rules require certain data to stay in their own facilities, or while they migrate step by step. The store might keep its warehouse system on premises and run its website on AWS.

Scene 9 · 3:38

Deployment model 3: on-premises. In an on-premises deployment, sometimes called a private cloud, everything runs in the company's own data center, using virtualization and resource management tools. It provides dedicated resources and direct control over the hardware. But the company still buys capacity up front and maintains everything itself, so most of the benefits of cloud computing don't apply.

Scene 10 · 4:03

How the exam asks about this. On the exam, look for the clues. Full control of the operating system points to IaaS. Deploying code without managing servers points to PaaS. Ready-to-use software points to SaaS. Some systems on premises, connected to the cloud, means hybrid. Everything in the cloud is a cloud deployment. And virtualization in your own data center is on-premises, or private cloud.

Scene 11 · 4:28

Recap: service and deployment models. To recap: IaaS, PaaS and SaaS describe how much the provider manages, from just the infrastructure, to the platform, to the whole application. Cloud, hybrid and on-premises describe where workloads run. You'll use both ideas again in the shared responsibility lesson, because what you must secure depends on the service you use. Next, we'll look at the economics of the cloud.

Download captions (WebVTT)

Sources and accuracy

Checked against AWS documentation and the official CLF-C02 exam guide on October 18, 2026. AWS changes often, so confirm details in the current documentation. Diagrams and numbers in the animation are illustrative.

This free course is independent and isn't affiliated with or endorsed by Amazon Web Services. AWS and AWS Certified Cloud Practitioner are trademarks of Amazon.com, Inc. or its affiliates.

← Lesson 1Cloud Computing and Its Six Advantages
Lesson 3Arrives

Preparing a team for AWS certification?

I run hands-on AWS training for teams, from Cloud Practitioner basics to architecture, built around your own environment.

Book a consultation Email me