โ† Home
Cloud and Data Center Technologies
Unit 1 โ€ข Topic 1.1

Cloud Computing: The Operating Model of Modern IT

Core Definition

Cloud computing is the on-demand delivery of compute, storage, networking, platforms, and software over the internet with measurable, elastic usage and shared infrastructure. It is less a technology and more an operating model: automate everything, scale instantly, and pay only for what you consume.

NIST Characteristics

Self-service, broad access, resource pooling, rapid elasticity, measured service.

Why It Matters

Transforms CapEx-heavy infrastructure into flexible OpEx with automation and global reach.

Design Mindset

Everything is an API. Everything is a service. Everything is automatable.

Essential Cloud Behaviors

Provision resources through APIs, auto-scale based on demand, and expose consumption metrics for governance and cost control.

Autoscaling Self-healing Metered billing

Shared Responsibility

Providers secure the cloud infrastructure; customers secure their data, identities, and configurations.

Identity Data Network policy
Layer Provider Manages Customer Manages
SaaS App, runtime, OS, virtualization, hardware Data, access policies
PaaS Runtime, OS, virtualization, hardware App code, data, access policies
IaaS Virtualization, hardware, facilities OS, runtime, apps, data, access policies

1.2 Roots of Cloud Computing

1950sโ€“1970s: Mainframes introduced time-sharing. Multiple users shared a single large computer through terminals.
1980sโ€“1990s: Client-server and distributed systems created the need for standardized networking, storage, and compute provisioning.
2000s: Virtualization unlocked hardware pooling; utility and grid computing proved pay-per-use consumption.
2010sโ€“Now: Hyperscale clouds, container orchestration, and serverless platforms shifted the developer workflow to automation-first.
๐Ÿงญ

Why History Matters

Every wave of computing reduces the cost of resource sharing. Cloud is the culmination: automation and elasticity at global scale.

1.3 Cloud Service Models (SPI)

๐Ÿงฑ

IaaS

Provision raw compute, storage, and network resources. Maximum control, maximum responsibility.

Examples: EC2, Compute Engine, Azure VMs

๐Ÿงช

PaaS

Managed runtimes for code deployment without managing OS or runtime patches.

Examples: App Engine, Azure App Service, Heroku

๐Ÿงฉ

SaaS

Full applications delivered over the web. Ideal for rapid productivity.

Examples: Office 365, Salesforce, Slack

Deployment Blueprint

Architects map workloads to service models based on data sensitivity, operational maturity, and speed-to-market requirements.

1.4 Cloud Layers & Deployment Types

Public Cloud

Shared infrastructure operated by a provider. Best for elasticity and global reach.

Private Cloud

Single-tenant environment with high governance and regulatory control.

Hybrid Cloud

Mix of on-prem and cloud with portability, data residency, and burst capacity.

Community Cloud

Shared by organizations with common compliance or security requirements.

Reference Architecture Layers

User Interface
API Gateway
Service Orchestration
Resource Pooling
Physical Infrastructure

1.5 - 1.7 Features, Benefits, and Applications

Desired Features

High availability, automated recovery, secure multi-tenancy, and policy-driven governance.

Fault tolerance Autoscaling Audit trails

Benefits vs Challenges

Benefits: agility, elasticity, global scale. Challenges: cost sprawl, shared responsibility, vendor lock-in.

Cost governance Security posture
โš ๏ธ

Exam Focus: Cloud Economics

Know when to choose OpEx over CapEx, how elasticity saves cost, and how reserved capacity lowers steady-state spend.

Typical Cloud Applications

Disaster recovery, streaming media, analytics pipelines, Dev/Test environments, and global e-commerce platforms.

Unit 2 โ€ข Virtualization

Virtualization Fundamentals

2.1 Introduction to Virtualization

Virtualization abstracts physical resources into logical units. It allows consolidation, isolation, and rapid provisioning while maximizing hardware utilization.

2.2 Core Characteristics

  • Partitioning: multiple OS instances on one host.
  • Isolation: one VM cannot impact another.
  • Encapsulation: VM state stored as files.
  • Hardware Independence: move workloads freely.

Virtualization vs Containerization

VMs emulate hardware; containers share the host kernel. Containers are lighter and faster but rely on OS-level isolation.

2.3 Virtualization Methods

Method Mechanism Typical Use
Full Virtualization Binary translation, unmodified guest OS Legacy workloads
Para-virtualization Hypercalls, modified guest OS High-performance workloads
OS-level Virtualization Namespaces and cgroups Containers, microservices
Performance Insight

CPU virtualization is efficient; I/O virtualization is the bottleneck. Use SR-IOV or paravirtual drivers for high throughput.

2.4 Types of Virtualization

Compute Virtualization

VMs and containers delivering logical compute. Supports live migration and autoscaling.

Network Virtualization

SDN and overlay networks (VXLAN) decouple network services from hardware.

Storage Virtualization

Aggregates multiple disks into logical volumes. Enables thin provisioning and snapshots.

Desktop/Application

VDI and app containers provide secure, centralized application delivery.

2.5 Hypervisors and VM Lifecycle

Type 1 Hypervisor

Runs directly on hardware. Preferred for data centers and cloud providers.

Examples: ESXi, Hyper-V, KVM

Type 2 Hypervisor

Runs on top of a host OS. Common for development and testing environments.

Examples: VirtualBox, VMware Workstation

VM Lifecycle

Provision
Configure
Monitor
Scale
Retire

2.6 Clustering and Automation

Cluster Types

High-availability clusters provide failover; load-balancing clusters distribute traffic; HPC clusters maximize throughput.

Auto-scaling Triggers

Scale based on CPU, queue depth, response latency, or custom business metrics.

๐Ÿ› ๏ธ

Infrastructure as Code

Declarative templates (Terraform, CloudFormation) make infrastructure versioned, repeatable, and auditable.