Skip to content

📘 Services Provided by Public Clouds & Comparison of Top Public Cloud Providers

Objectives

By the end of this chapter you will be able to:

  • Describe the core service models (IaaS, PaaS, SaaS, CaaS, FaaS) used by public cloud providers.
  • List and explain the major service categories offered by clouds (compute, storage, DB, networking, security, serverless, containers, analytics, AI/ML, CDN, migration, monitoring).
  • Compare the leading public cloud providers (AWS, Microsoft Azure, Google Cloud Platform) at a high level — strengths, enterprise fit, and hybrid/multicloud strategies.
  • Know where to look (official docs) and what hands-on exercises to do next.

Quick context

Public cloud is dominated by a few large providers — often referenced as the “big three” (AWS, Microsoft Azure, Google Cloud). This dominance affects available features, global reach, pricing pressure, and partner ecosystems.

Big Three

Cloud service-models (short & practical definitions)

  • IaaS (Infrastructure as a Service)
    Low-level building blocks (VMs, block storage, virtual networks). You manage OS, runtime, apps; provider manages hypervisor, physical hosts, networks, and storage.

  • PaaS (Platform as a Service)
    Managed runtime/platform for apps (platform hides OS/patching). Examples: managed app-hosting, managed databases, managed integration services.

  • SaaS (Software as a Service)
    Fully managed applications delivered over the web (e.g., Office 365, Salesforce). No infra management.

  • CaaS (Containers as a Service)
    Managed container hosting/orchestration platforms (hosted Kubernetes or container services).

  • FaaS / Serverless (Function as a Service)
    Short-lived functions invoked by events; provider manages scaling and server management (e.g., AWS Lambda).

Practical note

Vendor APIs/CLIs/SDKs let you automate IaaS/PaaS/FaaS operations (create VM, attach storage, create bucket, or deploy function).

📘 Core Service Categories in Public Cloud

Compute (Running Applications)

  • What it is: Virtual machines, containers, or serverless functions that run your applications.

  • Why it matters: Every app needs compute power (CPU + memory).

  • Examples:

    • AWS → EC2, Lambda
    • Azure → Virtual Machines, Azure Functions
    • GCP → Compute Engine, Cloud Functions

  • Watch out for:

    • Choosing wrong instance size = higher costs.
    • Data on temporary disks is lost if VM stops.

Storage (Saving Data)

  • What it is: Services to store files, application data, or databases.

  • Types:

    • Object Storage → for files, images, backups (S3, Blob, Cloud Storage).
    • Block Storage → attached to VMs, like virtual hard drives (EBS, Managed Disks, Persistent Disk).
    • File Storage → shared network drives (EFS, Azure Files, Filestore).

  • Why it matters: Different apps need different storage types (databases need block storage, websites need object storage).

  • Watch out for:

    • Data transfer (egress) costs.
    • Access permissions — misconfigured storage often causes data leaks.

Databases

  • What it is: Managed relational (SQL) or non-relational (NoSQL) databases.

  • Why it matters: Almost all apps need structured data storage.

  • Examples:

    • Relational: RDS (AWS), Azure SQL Database, Cloud SQL (GCP).
    • NoSQL: DynamoDB, Cosmos DB, Firestore.
    • Analytics: Redshift, Synapse, BigQuery.

  • Watch out for:

    • Costs can grow quickly with scale.
    • Replication/backup setup is critical for DR.

Networking & Content Delivery

  • What it is: Services that connect cloud resources and deliver apps globally.

  • Why it matters: Without networking, your VM is isolated.

  • Examples:

    • VPC (all providers) → private cloud networks.
    • Load Balancers → distribute traffic.
    • CDN → caches content near users (CloudFront, Azure CDN, Cloud CDN).

  • Watch out for:

    • Misconfigured firewalls can expose resources publicly.
    • Cross-region data transfer adds cost.

Identity & Security

  • What it is: Tools to control who can access what in your cloud.

  • Why it matters: Security is the #1 risk in cloud.

  • Examples:

    • IAM (AWS), Azure Active Directory, IAM (GCP).
    • Key Management: AWS KMS, Azure Key Vault, Cloud KMS.

  • Watch out for:

    • Over-permissive policies (“*” access).
    • Misconfigured public buckets.

Serverless & Event-Driven

  • What it is: Running code without managing servers.

  • Why it matters: Great for automating small tasks, cost-efficient.

  • Examples:

    • AWS Lambda, Azure Functions, GCP Cloud Functions.

  • Use case:
    Process image when it’s uploaded to storage.

  • Watch out for:

    • Execution time limits.
    • Cold-start latency.

Containers & Orchestration

  • What it is: Run applications in lightweight, portable containers.

  • Why it matters: Industry standard for microservices.

  • Examples:

    • AWS EKS, Azure AKS, Google GKE (all are managed Kubernetes).

  • Watch out for:

    • Cluster cost (idle nodes still cost money).
    • Networking/storage setup for stateful apps.

Monitoring & Logging

  • What it is: Services to track performance, errors, and logs.

  • Why it matters: Without monitoring, you don’t know when systems fail.

  • Examples:

    • AWS CloudWatch, Azure Monitor, Google Cloud Operations Suite.

  • Watch out for:

    • High logging costs with verbose apps.
    • Need alerts (not just dashboards).

Big Data & Machine Learning

  • What it is: Platforms to process massive datasets and build AI models.

  • Why it matters: Modern apps rely on analytics & AI.

  • Examples:

    • AWS SageMaker, Azure ML, Google Vertex AI.
    • AWS EMR, Azure Synapse, Google BigQuery.

  • Watch out for:

    • Data transfer costs between services.
    • Lack of data governance.

Migration & Hybrid Tools

  • What it is: Services to move apps/data into cloud or connect on-prem data centers.

  • Why it matters: Few companies start from scratch in cloud.

  • Examples:

    • AWS DMS (Database Migration Service), Azure Migrate, GCP Migrate for Compute.

  • Watch out for:

    • Large data transfers take time/cost.
    • Need cutover planning to avoid downtime.

Cost Management & Marketplace

  • What it is: Tools to control spending + app marketplace.

  • Why it matters: Cloud can get expensive if not managed.

  • Examples:

    • Cost Explorer (AWS), Cost Management (Azure), Billing Reports (GCP).
    • Marketplaces with pre-built apps (VM images, SaaS).

  • Watch out for:

    • Hidden egress/network charges.
    • Marketplace apps may have extra licensing costs.

Key Takeaway

  • All cloud providers offer these 11 categories, even if names differ.
  • Focus on the concepts, not the brand names. Once you understand categories, you can map them across AWS, Azure, and GCP.

How clouds expose functionality (APIs, CLI, SDKs)

  • RESTful APIs + JSON payloads are the canonical programmatic interface.
  • CLIs (aws/az/gcloud) provide scriptable tooling for labs and automation.
  • SDKs (Python/Go/Java/Node) are used inside applications and CI/CD pipelines.
  • Infrastructure-as-Code (Terraform, CloudFormation, ARM templates) enables reproducible infra and policy-as-code.

Top public cloud providers: short comparison & positioning

Provider-at-a-glance (high-level)

Provider Strengths / Differentiator Enterprise fit / Common use-cases Hybrid / Edge offering
AWS (Amazon Web Services) Largest breadth of services and global regions; mature ecosystem; many instance types and specialized hardware Startups → enterprise, feature-rich production systems, AI/ML infrastructure, broad partner ecosystem Outposts, Local Zones, Snow family for edge & data transfer
Microsoft Azure Deep enterprise integration (Active Directory, Windows Server, SQL Server), strong hybrid story Enterprises with Microsoft stack, large corp migrations, SaaS integrations Azure Arc, Azure Stack for on-prem hybrid
Google Cloud Platform (GCP) Data, analytics and ML-first: strong data services and scale; network performance Data analytics platforms, ML/AI workloads, Kubernetes-native apps Anthos for hybrid/multicloud, edge solutions

Note: Each provider also offers multi-region redundancy, compliance frameworks, and managed enterprise support tiers.

Example provider strengths explained (practical)

  • AWS: unmatched service variety — good when you need a very specific managed service (e.g., managed message brokers, streaming, specialized instances). AWS docs are the canonical reference for its service behaviors.
  • Azure: excellent if your environment is Windows-heavy or you want frictionless Active Directory + identity integration and vendor support agreements. Azure’s storage (Blob) and VM offerings are comparable to others and integrate tightly with Microsoft ecosystem tools.
  • GCP: invests heavily in data platforms (BigQuery), low-latency networking, and Kubernetes. Its Compute and data services are optimized for analytics and large-scale data processing.

How similar are services between providers?

  • Functionally most categories exist across providers (compute, object storage, managed DBs, serverless, managed kubernetes, CDN) — they differ mainly in API/CLI ergonomics, SLA/pricing, extra features, and regional availability. Google provides an official service comparison/mapping which is useful to study when migrating or comparing features.

Summary

  • Public clouds provide the same core categories (compute, storage, networking, IAM, DB, serverless, containers), but each provider packages and names these services differently. Understanding the category matters more than memorizing names.
  • Start practical learning with one provider or choose provider based on your institution's priorities; then learn how to translate concepts across vendors.
  • Next chapter will give a service-name mapping table (AWS / Azure / GCP) and deeper explanations of common services so students can map terms across providers.