Please enable JavaScript.
Coggle requires JavaScript to display documents.
Cloud Service Models & Provider Selection - Coggle Diagram
Cloud Service Models & Provider Selection
Provider Comparison (same services)
Amazon Web Services (AWS)
IaaS: EC2
PaaS: Elastic Beanstalk
FaaS: Lambda
Storage: S3
Strengths: largest service catalog, mature, global reach, big ecosystem
Weaknesses: complex pricing, steep learning curve, many services to choose from
Best for: startups to large enterprises needing breadth and maturity
Microsoft Azure
IaaS: Azure Virtual Machines
PaaS: Azure App Service
FaaS: Azure Functions
Storage: Blob Storage
Strengths: integration with Microsoft products (Windows, Active Directory, Microsoft 365), strong hybrid cloud (Azure Arc)
Weaknesses: console/documentation can be complex, some services less mature than AWS
Best for: enterprises already using Microsoft stack, hybrid scenarios
Google Cloud Platform (GCP)
IaaS: Compute Engine
PaaS: App Engine / Cloud Run
FaaS: Cloud Functions
Storage: Cloud Storage
Strengths: data analytics (BigQuery), AI/ML, Kubernetes (GKE), per-second billing, discounts
Weaknesses: smaller catalog and market share, fewer enterprise features/regions
Best for: data-driven, AI and container-native workloads
Comparison axes
Service equivalence: EC2 = Azure VM = Compute Engine; Lambda = Azure Functions = Cloud Functions
Pricing: pay-as-you-go, reserved/committed discounts, spot/preemptible instances, free tier or credits
Global footprint: number of regions and availability zones
Ecosystem: marketplace, partners, community, documentation
Specialty: AWS = breadth, Azure = Microsoft/hybrid, GCP = data/AI
Cloud Service Models & Provider Selection
Four Service Models
IaaS - Infrastructure as a Service
Definition: virtualized compute, storage and network rented on demand
Customer manages: OS, middleware, runtime, apps, data
Provider manages: servers, virtualization, storage, network, data center
Examples: AWS EC2, Azure Virtual Machines, Google Compute Engine
Use cases: lift-and-shift, test/dev environments, hosting, backup, HPC
Pros: maximum control and flexibility, pay-as-you-go
Cons: more administration, security of OS/apps is on the customer
PaaS - Platform as a Service
Definition: managed platform to build, deploy and run applications without handling infrastructure
Customer manages: application code and data
Provider manages: OS, runtime, middleware, scaling, patching, infrastructure
Examples: AWS Elastic Beanstalk, Azure App Service, Google App Engine, Heroku
Use cases: web/API development, CI/CD, managed databases
Pros: faster development, built-in scaling, less ops work
Cons: less control, vendor lock-in, runtime restrictions
SaaS - Software as a Service
Definition: complete application delivered over the Internet, usually by subscription
Customer manages: configuration, users, own data
Provider manages: everything else (full stack)
Examples: Microsoft 365, Google Workspace, Salesforce, Zoom, Slack
Use cases: email, CRM, collaboration, HR, accounting
Pros: no installation, automatic updates, access from anywhere
Cons: limited customization, dependence on provider, data privacy concerns
FaaS - Function as a Service (Serverless)
Definition: execute small event-driven functions; the provider allocates resources and scales automatically
Customer manages: function code only
Provider manages: servers, OS, runtime, scaling, availability
Billing: per invocation and execution time (scale to zero)
Examples: AWS Lambda, Azure Functions, Google Cloud Functions
Use cases: REST APIs, file/image processing, IoT events, chatbots, scheduled tasks
Pros: no server management, automatic scaling, pay only for execution
Cons: cold starts, execution time limits, hard to debug, lock-in
Responsibility gradient
On-premises: customer manages everything
IaaS: customer manages from OS upward
PaaS: customer manages app + data
FaaS: customer manages function code
SaaS: provider manages everything
Provider Comparison (same services)
Amazon Web Services (AWS)
IaaS: EC2
PaaS: Elastic Beanstalk
FaaS: Lambda
Storage: S3
Strengths: largest service catalog, mature, global reach, big ecosystem
Weaknesses: complex pricing, steep learning curve, many services to choose from
Best for: startups to large enterprises needing breadth and maturity
Microsoft Azure
IaaS: Azure Virtual Machines
PaaS: Azure App Service
FaaS: Azure Functions
Storage: Blob Storage
Strengths: integration with Microsoft products (Windows, Active Directory, Microsoft 365), strong hybrid cloud (Azure Arc)
Weaknesses: console/documentation can be complex, some services less mature than AWS
Best for: enterprises already using Microsoft stack, hybrid scenarios
Google Cloud Platform (GCP)
IaaS: Compute Engine
PaaS: App Engine / Cloud Run
FaaS: Cloud Functions
Storage: Cloud Storage
Strengths: data analytics (BigQuery), AI/ML, Kubernetes (GKE), per-second billing, discounts
Weaknesses: smaller catalog and market share, fewer enterprise features/regions
Best for: data-driven, AI and container-native workloads
Comparison axes
Service equivalence: EC2 = Azure VM = Compute Engine; Lambda = Azure Functions = Cloud Functions
Pricing: pay-as-you-go, reserved/committed discounts, spot/preemptible instances, free tier or credits
Global footprint: number of regions and availability zones
Ecosystem: marketplace, partners, community, documentation
Specialty: AWS = breadth, Azure = Microsoft/hybrid, GCP = data/AI
Criteria for Choosing a Provider
Cost
Pricing model and transparency
Total cost of ownership (egress/bandwidth fees, support plans)
Discounts: reserved, committed use, spot
Free tier / credits for testing
Security & Compliance
Certifications: ISO 27001, SOC 2, GDPR, PCI-DSS
Encryption at rest and in transit, IAM, key management
Shared responsibility clarity
Performance & Reliability
SLA / uptime guarantee (e.g. 99.9% and above)
Latency and network quality
Redundancy, backup and disaster recovery
Scalability & Flexibility
Auto-scaling and elasticity
Range of instance types and services
Service Portfolio & Technology Fit
Supports needed models (IaaS, PaaS, FaaS, SaaS)
AI/ML, big data, containers, IoT, DevOps tools
Compatibility with existing stack and languages
Data Location & Sovereignty
Regions close to users (latency)
Local legal requirements on data residency
Vendor Lock-in & Portability
Open standards, APIs, Kubernetes
Multi-cloud / hybrid support, exit strategy
Support & Reputation
Support plans, response times, documentation, training
Provider stability, customer references, community
Sustainability
Renewable energy use and carbon footprint
Four Service Models
IaaS - Infrastructure as a Service
Definition: virtualized compute, storage and network rented on demand
Customer manages: OS, middleware, runtime, apps, data
Provider manages: servers, virtualization, storage, network, data center
Examples: AWS EC2, Azure Virtual Machines, Google Compute Engine
Use cases: lift-and-shift, test/dev environments, hosting, backup, HPC
Pros: maximum control and flexibility, pay-as-you-go
Cons: more administration, security of OS/apps is on the customer
PaaS - Platform as a Service
Definition: managed platform to build, deploy and run applications without handling infrastructure
Customer manages: application code and data
Provider manages: OS, runtime, middleware, scaling, patching, infrastructure
Examples: AWS Elastic Beanstalk, Azure App Service, Google App Engine, Heroku
Use cases: web/API development, CI/CD, managed databases
Pros: faster development, built-in scaling, less ops work
Cons: less control, vendor lock-in, runtime restrictions
SaaS - Software as a Service
Definition: complete application delivered over the Internet, usually by subscription
Customer manages: configuration, users, own data
Provider manages: everything else (full stack)
Examples: Microsoft 365, Google Workspace, Salesforce, Zoom, Slack
Use cases: email, CRM, collaboration, HR, accounting
Pros: no installation, automatic updates, access from anywhere
Cons: limited customization, dependence on provider, data privacy concerns
FaaS - Function as a Service (Serverless)
Definition: execute small event-driven functions; the provider allocates resources and scales automatically
Customer manages: function code only
Provider manages: servers, OS, runtime, scaling, availability
Billing: per invocation and execution time (scale to zero)
Examples: AWS Lambda, Azure Functions, Google Cloud Functions
Use cases: REST APIs, file/image processing, IoT events, chatbots, scheduled tasks
Pros: no server management, automatic scaling, pay only for execution
Cons: cold starts, execution time limits, hard to debug, lock-in
Responsibility gradient
On-premises: customer manages everything
IaaS: customer manages from OS upward
PaaS: customer manages app + data
FaaS: customer manages function code
SaaS: provider manages everything
Criteria for Choosing a Provider
Cost
Pricing model and transparency
Total cost of ownership (egress/bandwidth fees, support plans)
Discounts: reserved, committed use, spot
Free tier / credits for testing
Security & Compliance
Certifications: ISO 27001, SOC 2, GDPR, PCI-DSS
Encryption at rest and in transit, IAM, key management
Shared responsibility clarity
Performance & Reliability
SLA / uptime guarantee (e.g. 99.9% and above)
Latency and network quality
Redundancy, backup and disaster recovery
Scalability & Flexibility
Auto-scaling and elasticity
Range of instance types and services
Service Portfolio & Technology Fit
Supports needed models (IaaS, PaaS, FaaS, SaaS)
AI/ML, big data, containers, IoT, DevOps tools
Compatibility with existing stack and languages
Data Location & Sovereignty
Regions close to users (latency)
Local legal requirements on data residency
Vendor Lock-in & Portability
Open standards, APIs, Kubernetes
Multi-cloud / hybrid support, exit strategy
Support & Reputation
Support plans, response times, documentation, training
Provider stability, customer references, community
Sustainability
Renewable energy use and carbon footprint