Chapter 2:

Cloud Computing Service

Models

Esprit 2025-2026
[Link]@[Link]
 Course Overview
This lesson will guide you through the fundamental concepts of cloud computing service models, equipping you with a clear
understanding of their definitions, characteristics, and practical applications. We'll cover:

1 2 3

Definition of Service Models Characteristics of Service Comparing Service Models

Models
Understanding what cloud service Analyzing the differences and
models are and their fundamental Exploring the unique features and similarities between IaaS, PaaS, and
role in cloud computing architecture. responsibilities associated with each SaaS to help you choose the right
model. model for various scenarios.
Understanding Cloud Service Models
A cloud service model encapsulates a specific configuration of IT resources provided by a cloud vendor. These models dictate the
level of management and control users have over their cloud environment.

The National Institute of Standards and Technology (NIST) identifies three primary service models, forming the bedrock of cloud offerings:

Infrastructure as a Platform as a Service Software as a Service

Service (IaaS) (PaaS) (SaaS)
Provides virtualized computing Offers a platform for developers to Delivers ready-to-use applications
resources over the internet. build, run, and manage applications. over the internet.

While various specialized models like Network as a Service (NaaS), Desktop as a Service (DaaS), Identity as a Service (IDaaS), and
Storage as a Service (StaaS) exist, they are generally categorized under one of these three foundational NIST models.
Understanding Cloud Service Models

Content Email
CRM
Management
Sales Office
Productivity
SaaS Collaboration
Document
Social Financials
Networks Management
Cloud
Customer Database

Integration
Development
Business
Service Testing Intelligence
Management PaaS Application
Storage Deployment
Platform
IaaS Network Hosting
Compute
Backup/Recovery
The Cloud Consumer Role
The cloud service consumer is a dynamic runtime role taken on by a software program when it interacts with a cloud service. This
role is fundamental to how applications and users leverage cloud capabilities.

Common types of cloud service consumers include diverse software programs and services designed to remotely access and utilize
cloud resources. These can range from mobile applications fetching data to complex enterprise systems relying on cloud
infrastructure.

Understanding the consumer's role is critical for designing and deploying effective cloud solutions, as it defines the interface and
interaction points with the cloud provider's services.
Who Uses Cloud Computing?
Cloud computing is a ubiquitous technology, adopted by organizations of all types, sizes, and industries. Its flexibility and scalability make it suitable for a wide
range of applications and business needs.

Here are some prominent examples of cloud consumers:

Mobile App Companies Websites

Leveraging cloud infrastructure for backend services, data storage, and user From small blogs to large e-commerce platforms, cloud hosting provides
authentication. reliability and performance.

Large Organizations Video Streaming Platforms

Extending IT services, managing big data analytics, and supporting global operations.
Netflix, Hulu, and YouTube rely on cloud for content delivery, encoding, and vast storage.

Online Video Games Developers

Utilizing cloud servers for multiplayer environments and massive online gaming Collaborating in shared environments for rapid development, testing, and deployment.
experiences.
Key Roles and Responsibilities in Cloud Service Models
Each cloud service model primarily targets different user personas and their corresponding responsibilities within an organization.
Understanding these roles is crucial for efficient cloud adoption and management.

IaaS PaaS SaaS

• System Administrators: Manage • Application Developers: Focus on • Organizations providing software
operating systems, virtual machines, writing and deploying code, utilizing applications: Manage user accounts,
and deployed applications. provided tools and runtime subscriptions, and application
• Network Administrators: Configure • environments.
Application Testers: Validate settings for their members or
virtual networks, firewalls, and load application functionality within the • employees.
End Users & Software Application
• balancers.
Storage Administrators: Oversee • PaaS environment.
Application Deployers: Manage the Administrators: Primarily consume
cloud storage solutions and data deployment pipelines and the application with limited
management. configuration of applications. configuration options.

• Application Administrators: Oversee

the application lifecycle and
• performance.
Application End Users: Interact with
the deployed applications directly.
Service Models (1/3) - Infrastructure as a Service (IaaS)
IaaS provides the foundational layer of cloud computing, offering raw computing resources over the internet. It gives consumers
the most control over their infrastructure, making it ideal for those who need highly customizable environments.

Consumer Capabilities Consumer Management & Control

• Provisioning: Ability to provision virtual machines, • Full Control: Over operating systems, applications,
storage volumes, and virtual networks. and deployed software.
• Deployment: Freedom to deploy and run virtually any • Limited Control: Over select networking components
software, including custom operating systems and (e.g., host firewalls, IP addresses).
applications. • No Control: Over the underlying cloud infrastructure
(physical servers, virtualization layer).

IaaS is analogous to renting raw server space in a data center, where you are responsible for everything from the operating system up.
Service Models (2/3) - Platform as a Service (PaaS)
PaaS offers a comprehensive environment for developing, running, and managing applications without the complexity of building
and maintaining the underlying infrastructure. It's a favorite among developers seeking agility.

Consumer Capabilities Consumer Management & Control

• Application Deployment: Deploying consumer- • Full Control: Over the deployed applications and their
created or acquired applications directly onto the cloud hosting environment configurations.
• infrastructure.
Language & Tools: Utilizing programming languages, • No Control: Over the underlying cloud infrastructure,
libraries, services, and tools supported by the cloud including network, servers, operating systems, or
provider for development. storage.

PaaS can be thought of as a fully equipped workshop where you can build your product without worrying about setting up the tools or electri
Service Models (3/3) - Software as a Service (SaaS)
SaaS is the most abstracted cloud service model, delivering ready-to-use applications directly to end-users over the internet. It
simplifies IT management by shifting all infrastructure and software maintenance to the provider.

Consumer Capabilities Consumer Management & Control

• Access and Utilize: Applications and services offered • No Control: Over the underlying cloud infrastructure
on the cloud infrastructure. (network, servers, operating systems, storage, or
• Accessibility: Applications are accessible from individual application components).
various client devices (e.g., web browser, mobile app, • Limited Configuration: Possible user-specific
or dedicated program interface). application configuration settings (e.g., personal
preferences or dashboard customizations).

SaaS is like subscribing to a fully functional software application, where you only focus on using the service, and the provider
handles everything else.
The Cloud Service Stack: A Spectrum of Control
Cloud computing offers various service models, each defining a different level of responsibility between the cloud provider and the
consumer. The choice of model significantly impacts control over the infrastructure, applications, and data.

On-Premise IaaS
Full control over all layers: Networking, Storage, Servers, Infrastructure as a Service. Provider manages Networking,
Virtualization, OS, Middleware, Runtimes, and Storage, Servers, and Virtualization. User manages OS,
Applications. High operational overhead. Middleware, Runtimes, and Applications.

PaaS SaaS
Platform as a Service. Provider manages up to Runtimes. Software as a Service. Provider manages everything. User
User focuses on Applications and data. Ideal for only interacts with the application. Highest abstraction.
developers.

This continuum from on-premise to SaaS represents a shift in responsibility and control, moving more operational burdens to the
cloud provider, allowing users to focus on higher-value tasks.
The Cloud Service Stack: A Spectrum of Control
Applications

Runtimes
SaaS: web application
Middleware

O/S
PaaS: application
deployment Virtualisation

Serveurs
IaaS: virtualisation
plateform Stockage

Réseau

Cloud Computing service models

Infrastructure as a Service (IaaS)

IaaS provides virtualized computing resources over the internet. It is the most flexible cloud service model, giving users
significant control over their operating systems, applications, and middleware, while the provider manages the underlying
infrastructure. This model is akin to renting physical servers, storage, and networking hardware without the responsibility of
maintaining them.
Platform as a Service (PaaS)

PaaS offers a complete development and deployment environment in the cloud, without the need to manage the underlying
infrastructure. This model is designed for developers, providing all the necessary tools and services to build, run, and manage
applications more efficiently. It abstracts away the complexities of infrastructure, allowing users to focus purely on code and
application logic.
Software as a Service (SaaS)
Key Characteristics of IaaS
Resource Virtualization Multi-Tenancy
Resources are distributed as a service, allowing on-demand IaaS generally includes multiple users sharing a single
access to compute, storage, and networking. This piece of hardware, logically isolated from each other. This
virtualization enables rapid provisioning and de- multi-tenant architecture is fundamental to cloud
provisioning of resources, adapting to changing demands. economics, enabling providers to offer services at lower
costs.

Dynamic Scaling Utility Pricing Model

IaaS allows for dynamic scaling, enabling users to It has a variable cost, utility pricing model, meaning you
effortlessly scale resources up or down based on real-time only pay for what you use. This pay-as-you-go approach
demand. This elasticity ensures applications maintain eliminates large upfront capital expenditures, converting
performance while optimizing costs. CapEx to OpEx.
IaaS Billing and Economic
Considerations

Typical Cloud Resources and Billing Rates

The specific billing rates for IaaS resources such as compute instances,
storage, and network egress are highly dynamic and influenced by several
These factors collectively
market factors.
• The cost of hardware continuously changes due to advancements in create a fluid pricing
manufacturing and supply chain dynamics. environment where optimizing
• Hardware becomes more power efficient, leading to lower operational cloud spend requires

costs for cloud providers, which can be passed on to consumers. continuous monitoring and
strategic resource
• The cost of utility power fluctuates, directly impacting the operational
management.
expenses of data centers.
• The cloud marketplace becomes more competitive, driving providers to
offer more attractive pricing models and features.
Key Characteristics of PaaS:
Empowering Developers
Web-Based Development Tools
PaaS offers robust web-based user interface creation tools, simplifying the
development process and enabling real-time collaboration among teams.

Multi-Tenant Architecture
It employs a multi-tenant architecture where multiple concurrent users utilize
the same development application, ensuring efficient resource sharing and
cost-effectiveness.
Virtualization Foundation
PaaS builds on virtualization technology, allowing resources to be easily
scaled up or down dynamically, providing flexibility to meet fluctuating
application demands.
Usage-Based Fees
Usage fees are calculated based on factors such as the number and types of
consumers (developers, testers, administrators) and the time for which the
platform is in use, aligning costs with actual consumption.
PaaS Offerings: Core Functions and Tools
Cloud Service Providers (CSPs) offer a comprehensive suite of tools and functions within their PaaS offerings, designed to support
the entire software development lifecycle. These features enhance productivity and streamline operations.

Runtime Environments Software Development Kit (SDK)

Support for multiple programming languages (e.g., Python, Java, A comprehensive kit of software development tools, libraries,
[Link]) for application development and testing, ensuring and APIs to facilitate efficient application creation and
broad compatibility. integration.

Administration Console Data Store

A centralized console for cloud-based application management, Robust data storage, management, and distribution
monitoring, and configuration, offering a unified control plane. repositories, supporting various database types and ensuring
data availability and integrity.
Key Characteristics of SaaS

Browser Accessibility
SaaS applications are primarily accessible via a web browser, requiring no local
installation and often featuring offline capabilities for enhanced user flexibility.

"One to Many" Model

Software is delivered in a "one to many" model, where a single instance of the
application serves multiple users, enabling efficient updates and maintenance.

No Upgrade Management
Users are not required to handle software upgrades, patches, or security updates, as
these are managed entirely by the cloud provider, ensuring users always have the
latest version.

Centralized Hosting
SaaS solutions are managed from a central location and hosted on remote servers,
ensuring high availability, scalability, and seamless data access from anywhere.
Comparing Service Models: IaaS, PaaS, and SaaS
Understanding the distinctions between IaaS, PaaS, and SaaS is crucial for selecting the most appropriate cloud strategy. Each
model offers a unique balance of control, management, and flexibility, tailored to different organizational needs.

Control Level High (OS, applications, data) Medium (applications, data) Low (application configuration)

Management Responsibility User manages OS, User manages applications, Provider manages everything
middleware, apps data

Typical Use Cases Virtual machines, custom Application development, CRM, email, office
networking, data centers web hosting, APIs productivity suites

Billing Model Pay-as-you-go (compute, Usage-based (platform time, Subscription-based (per

storage, network) users) user, per month/year)

Choosing the right model depends on factors such as required control, development needs, existing infrastructure, and budget.
Often, organizations adopt a hybrid approach, leveraging multiple models to meet diverse operational requirements.
 Comparing Cloud Service Models
Model Suitable No Suitable
IaaS (Infrastructure as a Service) New organizations without significant capital for Applications demanding the highest levels of
hardware investment. performance.
Trial or temporary infrastructure needs for When hosted infrastructure perfectly aligns with
projects. organizational needs.
Rapidly growing businesses requiring scalable
hardware.

PaaS (Platform as a Service Teams with multiple developers or external When some customization of the underlying
parties needing collaboration. software is needed, but not hardware.
Automating testing and deployment services for For applications using specific proprietary
faster releases. languages supported by the platform.
Developing highly portable applications across
different environments.

SaaS (Software as a Service Software with demand spikes, like seasonal When real-time data processing requires
applications. extremely fast processing (often at the application
Short-term software needs, avoiding long-term layer).
commitments. When legislative requirements do not prohibit
Applications requiring significant web or mobile external data hosting.
access.

Note: The "Suitable" column indicates optimal scenarios, while "No Suitable" highlights situations where the model is generally not
recommended, often due to specific technical or regulatory constraints.
 Criteria for comparing cloud providers
Key Criteria for Cloud Provider Selection
Vendor Lock-in & High Availability Security & Privacy
Interoperability • •
Robust incident-reporting Granular access control
Assess the ease of migrating data systems. mechanisms (IAM).
and applications if a change in • Comprehensive backup and • Data integrity validation and
provider is necessary, ensuring disaster recovery solutions. encryption (at rest and in
flexibility and avoiding proprietary transit).
• Redundancy across multiple • Clear policies on data location
dependencies.
availability zones to minimize and residency.
downtime. • Access via secured networks
(VPNs, private links).
• DDoS Protection Services and
WAFs to mitigate threats.

These factors are crucial for ensuring business continuity, data protection, and long-term strategic flexibility when adopting cloud services.
 Criteria
Evaluating CloudforProviders:
comparing cloud providers
Performance, Support,
and Deployment
SLA & Support Supported Hosts / Clients Supported Deployment Models
• Clearly defined performance • Compatibility with a wide range of • Public Cloud: Fully managed by the
guarantees and uptime operating systems (Linux, provider, accessible over the
• commitments.
24/7 technical support availability • Windows).
Support for various programming • internet.
Hybrid Cloud: Combines on-
with specified response times. languages (Python, Java, [Link]). premises infrastructure with public
• Guaranteed minimum outage • Adherence to industry standards • cloud.
Private Cloud: Dedicated cloud
durations and compensation and open-source technologies. infrastructure, managed internally
• clauses.
Service Credits for performance • Integration capabilities with or by a third party.
degradation. existing enterprise systems. • Community Cloud: Shared
infrastructure for specific
communities with common
concerns.

Understanding these aspects helps align cloud services with operational needs and architectural preferences.
 Criteria for comparing cloud providers
Global Reach: Geographic Availability of Cloud Providers

This global distribution enables organizations to deploy resources closer to end-users, reducing latency and enhancing application perform
The number and distribution of regions and availability zones vary significantly among major cloud providers:

Cloud computing infrastructure is distributed worldwide across "regions" and "availability zones."
Each region is a distinct geographical area, while availability zones are isolated locations within a region, designed for redundancy and fau

• AWS: Operates 33 regions globally, with each region typically having between 2-6 availability zones. They continue to expand their glo
• GCP: Has 40 regions and 121 zones across the globe, focusing on highly connected networks for global applications.
• Azure: Boasts the largest global presence with 60+ regions worldwide, available in 140 countries, offering extensive reach.

While GCP offers a region in Hong Kong, access to those resources from mainland China can be
inconsistent due to network restrictions. This highlights the importance of detailed regional assessments for
specific market access requirements.

Notably, geographic availability can also be influenced by regulatory and political factors. For instance, AWS
is currently the only major cloud provider authorized to operate directly in mainland China, navigating
complex local regulations.
 Criteria for comparing cloud providers
Understanding Cloud Pricing Models
While the core pricing models of major cloud providers are similar, the nuances in their approaches can lead to significant cost differences.
Most providers offer:

1 2 3

On-Demand, Pay-as-You-Go Discounts for Committed Usage Usage-Based Serverless

Standard hourly rates for compute Reduced rates for customers who Serverless functions (e.g., AWS Lambda,
resources (CPU, RAM, Disk), offering commit to specific usage levels over a Azure Functions, GCP Cloud Functions)
flexibility without upfront commitments. period (e.g., 1 or 3 years). These vary in are billed based on actual compute
Ideal for unpredictable workloads. structure: power consumed, typically in 100-
• AWS Reserved Instances: Pre-
millisecond increments. This model is
purchase compute capacity annually
highly cost-effective for event-driven,
for significant savings.
intermittent workloads.
• GCP Sustained Use Discounts:
Automatic discounts applied the
longer an instance runs within a
month, without upfront commitment.

Detailed cost analysis and optimization strategies are crucial to manage cloud spending effectively, considering both current and future operational n
 Workshop: Cloud Deployment Models
In this interactive workshop, each group will delve into one of the four primary cloud deployment models:

Public Cloud Private Cloud

Hybrid Cloud Community Cloud

Each group should prepare and present the following:

• Adoption Scenarios: Define specific situations or business needs that warrant the adoption of your assigned deployment model.
• Evaluation Criteria: Identify and explain the key criteria (e.g., elasticity, control level, cost, security, compliance) used to evaluate and choose y
• Success Stories: Provide two or more real-world examples or hypothetical scenarios demonstrating the successful implementation and
benefits of your deployment model.