Day 1 | Basics of Cloud Computing & Fundamentals of Azure

This foundational chapter establishes the core mental models, architectural terminology, and structural prerequisites required before diving into any major cloud platform (AWS, Azure, or GCP). Without understanding these baseline concepts—such as the evolution from traditional data centers to public cloud, virtualization, APIs, and high availability—navigating cloud management consoles becomes overwhelming.

1. Why Do We Need the Cloud? (The Evolution of Infrastructure)

To understand why cloud computing revolutionized the software industry, we must examine how applications were deployed and managed historically.

The Pre-Cloud Era (Traditional Data Centers)

If you look back 10–15 years, companies like Google, Yahoo, or mid-scale enterprises maintained their own physical infrastructure.

• The Hardware: Companies procured physical servers from hardware vendors like IBM or HP.
• The Setup: System administrator teams placed these servers into physical racks (similar to wardrobes in a home), wired them with high-data-transfer cables, and connected them via switches and routers.
• The Location: This entire physical setup was housed in a dedicated room or building known as a Data Center.

Real-Life Analogy: The Corporate Server Room vs. Renting Office Space

Analogy: Imagine starting a restaurant 15 years ago. Instead of renting a fully equipped commercial kitchen, you had to buy the building, install industrial stoves, build refrigeration units, run plumbing, and hire a 24/7 electrical maintenance crew just to cook a single meal. This is equivalent to building a private data center.

Cloud computing is like moving into a modern food hall: the infrastructure, electricity, plumbing, and security are already built and maintained by the property owner. You simply rent the exact stove space you need and start cooking immediately.

Operational Challenges of Traditional Data Centers

Maintaining physical data centers created massive overhead for organizations:

1. 24x7 Power Supply: If electricity failed, the servers went down, rendering applications inaccessible.
2. Administrative Overhead: Smaller companies could not afford large dedicated system administrator teams to manage physical hardware.
3. Continuous Maintenance: Hardware degraded, cables frayed, operating systems required patching, and security vulnerabilities demanded constant attention.
4. Capital Expenditure (CapEx): Buying physical servers upfront required huge financial investments, regardless of whether traffic was high or low.

2. Core Mental Model: Private vs. Public vs. Hybrid Cloud

Cloud computing emerged to solve the physical and financial bottlenecks of traditional data centers by abstracting hardware management into software-driven services.

Cloud Deployment Models Compared

Multicloud vs. Hybrid Cloud

• Hybrid Cloud: Mixing private infrastructure (on-premises) with public cloud infrastructure (e.g., Azure).
• Multicloud: Utilizing multiple public cloud providers simultaneously (e.g., running some workloads on AWS and others on Azure).

3. What is Cloud Computing?

Cloud computing is the on-demand delivery of computing services—including servers, storage, databases, networking, and software—over the internet ("the cloud") on a pay-as-you-go pricing model.

Instead of owning physical hardware or data centers, companies can rent access to anything from a single virtual machine to artificial intelligence infrastructure from cloud providers like Microsoft Azure.

4. Virtualization: The Magic Behind Multi-Tenancy

When hardware vendors like IBM sell servers, they arrive with massive enterprise configurations (e.g., hundreds of CPU cores and hundreds of gigabytes of RAM).

The Problem of Undersutilization

If a developer requests a single server to test an application, allocating an entire physical machine with 512 GB of RAM and 1,000 CPU cores is wasteful and inefficient.

The Solution: Hypervisors and Virtual Machines (VMs)

To solve this, cloud providers and system administrators use virtualization.

• Hypervisor: A specialized software layer installed directly on top of physical server hardware.
• Virtual Machine (VM): A logical (software-based) isolation of a physical server.

Note: When you request a small virtual machine with 1 CPU and 2 GB of RAM on Azure or AWS, the cloud platform does not ship you a physical computer. Instead, it carves out a tiny, logically isolated slice of a massive physical server managed by a hypervisor.

5. Programmatic Access: APIs vs. User Interfaces (UI)

Interacting with cloud platforms and applications can be accomplished through multiple interfaces depending on automation needs.

1. User Interface (UI)

• What it is: Visual dashboards accessed via web browsers (Chrome, Firefox) where users click buttons, fill out forms, and navigate menus.
• Use Case: Great for beginners learning Azure or manual, one-off resource creation.

2. Command Line Interface (CLI)

• What it is: Text-based terminal commands used to execute administrative tasks.

3. Application Programming Interface (API)

• What it is: Programmatic endpoints exposed by applications or cloud platforms that allow software scripts (written in Python, Shell, etc.) to interact with systems.
• Why it matters: If an organization needs to provision 1,000 virtual machines instantly, clicking buttons in the UI 1,000 times is impractical. Instead, engineers write scripts that send automated requests to Azure's APIs to provision resources instantly.

6. Global Infrastructure: Regions and Availability Zones

Cloud providers operate physical infrastructure spread across the entire globe to ensure low latency and high reliability.

Geographical Hierarchy

1. Geography / World Map: High-level regional divisions (e.g., North America, Europe, Asia-Pacific).
2. Regions: Specific geographic areas containing multiple data centers (e.g., East US, West US, South India).
3. Availability Zones (AZs): Physically separate data centers within a single region equipped with independent power, cooling, and networking.

Why Multiple Regions and Availability Zones Matter

If Microsoft placed all US servers in a single data center in Alaska, a localized power outage or natural disaster would shut down operations for thousands of companies worldwide, causing massive financial losses.

To prevent this:

• Cloud providers distribute data centers across multiple regions (e.g., US East and US West).
• Within each region, multiple Availability Zones are spaced miles apart so that if one data center experiences a failure, neighboring zones remain fully operational.

7. Load Balancing: Distributing Traffic for High Availability

When millions of users access an application simultaneously, a single server cannot handle the influx of requests.

How Load Balancers Work

A Load Balancer sits in front of multiple application replicas (Virtual Machines) deployed across different Availability Zones.

Key Responsibilities of a Load Balancer

1. Traffic Distribution: Routes incoming traffic evenly (or via weighted algorithms like 20%/80%) across multiple servers.
2. Health Checks: Constantly monitors the health of backend virtual machines. If VM 1 crashes or its Availability Zone goes offline, the load balancer automatically stops sending traffic to it and reroutes 100% of traffic to healthy instances.
3. Application Protection: Prevents single servers from crashing under sudden traffic spikes.

8. Essential Cloud Computing Features

Cloud computing introduces architectural capabilities that traditional physical data centers could never achieve easily.

1. Scalability

The ability of an infrastructure to handle growing amounts of work by adding resources.

• Manual Scaling: Human engineers manually add or remove servers based on anticipated traffic.
• Auto-Scaling: Automated systems dynamically provision or terminate virtual machines based on real-time CPU utilization or incoming request volume.

2. Elasticity (Dynamic Scaling)

Often used interchangeably with auto-scaling, elasticity refers to the cloud's ability to automatically expand or shrink infrastructure resources dynamically as demand changes, ensuring you only pay for what you use.

3. High Availability (HA)

The characteristic of a system that aims to ensure an agreed level of operational performance (uptime) for a higher-than-normal period. Popular applications like Instagram and Facebook maintain high availability architectures so they rarely experience downtime.

4. Disaster Recovery (DR)

The strategy, policies, and procedures involved in maintaining or recovering vital technology infrastructure following a natural or human-induced disaster.

• Example: Storing user databases in US East while maintaining an automated, synchronized backup copy in a completely different region (such as Singapore). If US East suffers a catastrophic failure, systems failover to the backup region.

Summary and Key Takeaways

• Physical Data Centers vs. Cloud: Traditional setups required buying physical servers, racks, and managing 24/7 power and cooling. Public clouds abstract this physical management into software-driven services.
• Deployment Models: Choose between Private Cloud (high security/legacy compliance), Public Cloud (global scale, pay-as-you-go), and Hybrid Cloud (combining on-premises with cloud).
• Virtualization: Hypervisors logically partition massive physical servers into isolated, smaller Virtual Machines (VMs) to maximize hardware efficiency.
• APIs and UIs: While User Interfaces are great for manual learning, APIs enable developers and DevOps engineers to automate infrastructure provisioning programmatically using scripts.
• Regions & Availability Zones: Cloud resources are distributed globally into isolated regions and multi-tenant Availability Zones to guarantee fault tolerance and disaster recovery.
• Load Balancers: Vital traffic directors that distribute incoming requests across server replicas and automatically route around failed instances using health checks.
• Elasticity and Auto-Scaling: Cloud architectures scale dynamically up or down based on live traffic demands, preventing application slowdowns during traffic spikes.
• Disaster Recovery Plans: Critical business continuity measures involving multi-region data replication to protect against catastrophic regional failures.