Cloud Fundamentals

---

What is Cloud? What are its advantages? #

---

What is Cloud?

• Scenario: Imagine you need to setup 100 servers for a website you are building. Can we rent instead of buy?
• Cloud: Rent computing, storage, database, and other services from companies like AWS, Azure, Google Cloud, .. – whenever you need them.

Cloud is On-Demand

• No Upfront Investment: No need to buy servers or data centers
• Faster Start: Launch apps in minutes, not months
• Pay-As-You-Go: Pay only for what you use

Cloud is Scalable

• Elastic - Scale Up/Down: Add or remove resources as needed
• Automatic Scaling: Many services scale based on demand
• Global Reach: Deploy apps closer to your users across the world

Benefits of Cloud

• Faster Delivery: Launch apps in minutes
• Cost-Efficient: Pay only for what you use
• Scalable: Auto-scale to handle traffic spikes
• Global: Deploy close to users, anywhere in the world
• Secure & Compliant: Built-in security and certifications

---

Region vs Zones vs EdgeLocations #

---

Region vs Zones vs Edge Locations

• Scenario: You want your app to be fast, reliable, and globally available. How does the cloud provider structure its infrastructure to help you do that?
• Answer: Using Regions, Zones, and Edge Locations – each plays a different role.

Region

• What It Is: A geographical area (e.g., Mumbai, US-East)
• Use Case: Choose a region close to your users or data for low latency and compliance

Region Advantages

• High Availability: Even if a region is down, you can serve from other regions
• Low Latency: Deploy closer to users for faster response
• Global Footprint: Reach users worldwide without owning infrastructure
• Adhere to government regulations: Keep data in-region to meet compliance laws

Zone

• What It Is: One or more data centers inside a region
• Isolated but Connected: Physically separated but connected with low latency
• Use Case: Deploy across zones for high availability and fault tolerance (while being in same region)

Edge Location

• What It Is: A content delivery endpoint near users
• Key Thing To Remember: Count of Edge Locations is far greater than Count of Regions
• Smaller than AZs: Focused on caching and request routing
• Use Case: Deliver content faster via CDN (e.g., CloudFront)

Edge Location Use Case – Step by Step

• Step 01: A user requests a video or image from your website (e.g., https://yourapp.com/logo.png)
• Step 02: The request is routed to the nearest edge location (a local server near the user)
• Step 03: If the content is cached at the edge, it is served instantly — no need to reach the main server
• Step 04: If not cached, the edge location fetches it from the origin server
• Step 05: The content is delivered to the user and simultaneously cached at the edge for future requests
• Step 06: Next time another user in the same region makes the same request, it’s delivered directly from the edge — faster and more efficient
• Edge locations improve performance, reduce latency, and offload origin servers, especially for static content and media files

What are Multi Regions?

• Multi Regions: Multiple geographically separate cloud regions are logically linked to enable geo-redundancy, disaster recovery, and resilient architecture

Key Benefits of Multi Regions

• High Availability: Data is replicated to multiple regions
• High Durability: Multiple copies of data are stored across both regions

Concept	AWS	Azure	Google Cloud (GCP)
Region	Region E.g., us-east-1	Region E.g., East US, West Europe	Region E.g., us-central1, europe-west1
Zone	Availability Zone E.g., us-east-1a	Availability Zone E.g., East US 2 Zone 1	Zone E.g., us-central1-a
Edge Location	Edge Location (CloudFront POP) E.g., Hyderabad	Point of Presence (POP) used in Azure Front Door / Azure CDN	Edge Location / POP used in Cloud CDN
Multi-Region / Dual-Region	❌ Not explicitly named; use features like Cross-Region Replication, Global DynamoDB	✅ Region Pairs (Dual Regions) Automatically paired for DR, etc.	✅ Multi-Region is natively supported E.g., us, eu, asia - BigQuery, GCS

---

On-premise vs Hybrid Cloud vs Multi Cloud #

---

1. On-Premise

• What It Is: All infrastructure is owned and operated by you
• Where It Runs: In your own data center
• Pros:
  • Full control over hardware and security
  • No dependency on external providers
• Cons:
  • Expensive upfront cost
  • Hard to scale quickly
  • Slower to adopt modern tools

2. Hybrid Cloud

• What It Is: A mix of on-prem + public cloud
• Where It Runs: Some workloads on-prem, some in cloud
• Pros:
  • Flexibility to keep sensitive data on-prem
  • Leverage cloud scale for web apps or backups
  • Step-by-step cloud adoption
• Cons:
  • Complex setup and integration
  • Needs strong network and security architecture
  • Harder to manage consistently

3. Multi-Cloud

• What It Is: Use of 2 or more public cloud providers (e.g., AWS + Azure) with or without on-premises
• Why Use It: Avoid vendor lock-in, optimize cost/performance
• Pros:
  • Use best features from each provider
  • Increase availability and redundancy
  • Competitive pricing and negotiation
• Cons:
  • Higher complexity in architecture and management
  • Requires multi-skilled teams
  • Data sync challenges

---

IaaS vs PaaS vs SaaS vs Serverless #

---

IaaS vs PaaS vs SaaS vs Serverless

• Scenario: You want to build, deploy, or use an application. How much do you want to manage yourself vs how much should the cloud provider handle for you?
• Goal: Choose the right cloud service model based on control, convenience, and responsibility.

IaaS (Infrastructure as a Service)

• Use only infrastructure from cloud provider
  • Ex: Using VM service to deploy your apps/databases
• Cloud provider is responsible for:
  • Hardware, Networking & Virtualization
• You are responsible for:
  • OS upgrades and patches
  • Application Code and Runtime
  • Configuring load balancing
  • Auto scaling
  • Availability
  • etc.. ( and a lot of things!)
• Examples:Amazon EC2, Azure Virtual Machines, Google Compute Engine

PaaS (Platform as a Service)

• Use a platform provided by the cloud
  • Cloud provider is responsible for:
    • Hardware, Networking & Virtualization
    • OS (incl. upgrades and patches)
    • Application Runtime
    • Auto scaling, Availability & Load balancing etc..
  • You are responsible for:
    • Configuration (of Application and Services)
    • Application code (if needed)
• Examples:
  • Compute: AWS Elastic Beanstalk, Azure App Service, Google App Engine
  • Databases: Relational (Amazon RDS, Google Cloud SQL, Azure SQL Database etc)
  • And a lot of others!

SaaS (Software as a Service)

• Centrally hosted software (mostly on the cloud)
  • Offered on a subscription basis (pay-as-you-go)
  • Examples:
    • Email, calendaring & office tools (such as Outlook 365, Microsoft Office 365, Gmail, Google Docs)
    • Customer relationship management (CRM), enterprise resource planning (ERP) and document management tools
• Cloud/Service provider is responsible for:
  • OS (incl. upgrades and patches)
  • Application Runtime
  • Auto scaling, Availability & Load balancing etc..
  • Application code and/or
  • Application Configuration (How much memory? How many instances? ..)
• Customer is responsible for:
  • Configuring the software!

Serverless

• What do we think about when we develop an application?
  • Where to deploy? What kind of server? What OS?
  • How do we take care of scaling and availability of the application?
• What if you don't worry about servers and focus ONLY on code?
  • Enter Serverless
    • Remember: Serverless does NOT mean "No Servers"
• Serverless for me:
  • You don't worry about infrastructure (ZERO visibility into infrastructure)
    • Flexible scaling and automated high availability
  • Most Important: Pay for use
    • Ideally ZERO USAGE => ZERO COST
• You focus on code and the cloud managed service takes care of all that is needed to scale your service/code to serve millions of requests!
  • And you pay for usage and NOT servers!

Serverless Examples

Category	AWS	Azure	Google Cloud
Compute	AWS Lambda, AWS Fargate	Azure Functions, Azure Container Apps	Cloud Functions, Cloud Run
Storage	Amazon S3	Azure Blob Storage	Cloud Storage
Databases and a lot of others	..	..	..

---

What is Virtualization? #

---

What is Virtualization?

• Scenario: In the past, running one app meant using one full physical server—even if the app used only 10% of the CPU. This wasted resources and increased costs.
• Goal: Use one physical server to run multiple virtual systems efficiently.

Virtualization – The Basics

• Definition: Creating virtual versions of computing resources (like virtual machines) on a single physical server
• How It Works:
  • A Hypervisor (software layer) sits on top of the physical server
  • It splits the hardware into multiple Virtual Machines (VMs)
  • Each VM acts like a separate computer with its own OS and applications

Why Use Virtualization?

• Better Utilization: Run many apps on fewer servers
• Cost Efficiency: Lower hardware and maintenance costs
• Flexibility: Quickly create, start, stop, or move VMs
• Isolation: Each VM is independent – safer and easier to manage

Foundation of Cloud Computing

• Public cloud providers (like AWS, Azure, Google Cloud) use virtualization to offer shared infrastructure
• You get a VM (virtual server), not a dedicated physical machine
• Containers, serverless functions and almost everything you see in the cloud are built on top of virtualization layers

---

How can you create Virtual Machines in the cloud? #

---

What is a Virtual Machine (VM)?

• Definition: A software-based (virtual) machine running inside a physical server
• Goal: Run multiple isolated systems on a single physical machine
• Benefit: Full control over OS, software, and security settings while sharing hardware

Cloud VMs are Flexible

• Choose What You Need: Select CPU, RAM, storage, and OS
• Multiple OS Options: Run Linux, Windows, or custom images
• Custom Machine Types: Match VMs to your workload and budget

Cloud VMs are Scalable

• Scale Up/Down: Change number of VMs based on traffic
• Auto Scaling: Add or remove VMs automatically
• Global Deployment: Launch in regions close to users

Cloud VMs are Cost Efficient

• Pay-As-You-Go: Billed by second, minute, or hour
• Reserved/Spot Pricing: Save cost with long-term reservations or using spare capacity
• Release When Not Needed: Reduce waste by releasing unused VMs

Reviewing Important Virtual Machine Concepts

Feature	Explanation	AWS	Azure	Google Cloud
Managed Service	Create VMs	Amazon EC2 (Elastic Compute Cloud)	Azure Virtual Machines	Google Compute Engine
Image	What OS and software should be on the instance?	AMI (Amazon Machine Image)	VM Image	Image
Instance Family	Type of hardware: General, High CPU or High Memory or Storage Optimized	Instance Family	VM Series	Machine Family
Instance Size	Amount of vCPU and memory	Instance Type - t3.micro, m5.large	VM Sizes - B2s,B2ms ..	Machine Type - e2-medium, n2-standard-2
Attached Disks	Block storage volumes for VMs	Elastic Block Store	Managed Disks	Persistent Disks

Networking for VMs

Feature	Explanation	AWS	Azure	Google Cloud
Permanent Internal IP Address	Permanent Internal IP Address that does not change during the lifetime of an instance	Private IP Address	Private IP Address	Internal IP Address
Ephemeral External IP Address	Ephemeral External IP Address that changes when an instance is stopped	Public IP Address	Public IP Address	External or Ephemeral IP Address
Permanent External IP Address	Permanent External IP Address that can be attached to a VM	Elastic IP Address	Static IP Address	Static IP Address
Firewall Rules	Control incoming/outgoing traffic	Security Group	Network Security Group (NSG)	Firewall Rules

Managing Costs

Feature	Explanation	AWS	Azure	Google Cloud
Cheaper temp instances	Create cheaper, temporary instances for non critical workloads	Spot instances	Spot VMs	Spot VMs
Reservations/Usage Discounts	Reserve compute instances ahead of time/Get discounts for using resources for long periods of time	Reserved instances	Reserved instances	Committed use discounts/Sustained use discounts
Reservation based on dollar amounts	Reserve based on monetary commitment ($100 per hour, for example)	AWS Savings Plans	Azure Savings Plan for Compute	-
Managing Budget	Budget Management	Budget alerts	Budget alerts	Budget alerts

---

What are some of the architectural aspects to think about when creating virtual machines in the cloud? #

---

• Scaling
• Resiliency
• Observability
• Tenancy

What is Scaling?

• Scenario: Your app is growing – more users, more traffic, more data. One server is not enough anymore. You need to scale.
• Scaling: Expanding system capacity to handle more load.

Vertical Scaling (Scale Up)

• What It Is: Increase the power of a single machine
  • More CPU, RAM, disk, or network
  • Example: Upgrade a small server to a large one
• Benefits:
  • Simple and fast to implement
  • No code changes or architectural updates
• Limitations:
  • Hardware has physical limits
  • Downtime may be required
  • Higher cost for high-end machines

Horizontal Scaling (Scale Out)

• What It Is: Add more machines or instances to share the load
  • Run multiple copies of the application or database
• Benefits:
  • Improved availability and fault tolerance
  • Scale without downtime
• Challenges:
  • Needs infrastructure like load balancers

Horizontal Scaling in Practice

• Auto Scaling: Automatically add or remove instances based on traffic
• Load Balancing: Evenly distribute requests across all instances
• Deployment Models:
  • Within a single zone
  • Across multiple zones

Resiliency for Virtual Machines and Load Balancing (Cloud Neutral)

• Scenario: You're running your application on virtual machines. What happens if a VM crashes, a zone goes down, or your app faces a sudden surge in traffic?
• Goal: Design your architecture to handle failures gracefully and continue to serve users reliably — that’s resiliency.

What is Resiliency?

• Definition: The ability of a system to provide acceptable performance and recover quickly even when parts of the system fail
• Why It Matters:
  • Failures are inevitable (hardware, software, network)
  • Users expect high uptime and reliability

Building Resilient Architectures

Design Element	Purpose
Auto-Healing Groups / Instance Groups	Automatically detect and replace failed VMs
Load Balancing	Distribute traffic across healthy instances to ensure availability
Multi-Zone Deployment	Avoid single point of failure by spreading instances geographically
Health Checks	Detect and route traffic only to healthy instances
Disaster Recovery Planning	Keep up-to-date VM images and backups in multiple regions or zones

Creating Multiple VMs

Feature	Explanation	AWS	Azure	Google Cloud
VM Templates	Templates to simplify creation of Virtual Machines	Launch Templates	- (ARM Templates)	Instance templates
Auto Scaling	Automatically add or remove VMs based on usage (load, time, events)	Auto Scaling Group (ASG)	Virtual Machine Scale Sets (VMSS)	Managed Instance Groups (MIGs)
Load Balancing	Load Balancing	Elastic Load Balancer	Azure Load Balancer (& others)	Cloud Load Balancing
VM Management	Simplify management (software, OS patches etc) of 1000's of Virtual Machines	Systems Manager	Azure Update Manager	VM Manager

Observability

• Collect metrics like CPU, memory, network usage, disk I/O
• Capture application and system logs for troubleshooting
• Set alerts for failures

Question	AWS	Azure	Google Cloud
How do you monitor metrics around your applications	Amazon CloudWatch	Azure Monitor	Cloud Monitoring
How do you manage application and service logs?	Amazon CloudWatch Logs	Azure Monitor Logs	Cloud Logging
How do you trace requests across applications and services?	AWS X-Ray	Azure Application Insights Distributed Tracing	Cloud Trace

Why Tenancy Matters?

• Scenario: You’re deploying an app with licensing restrictions or regulatory needs. You need control over the hardware it runs on.
• Tenancy: Determines who shares the physical server where your virtual machines (VMs) run.

Shared Tenancy (Default)

• What It Is: Multiple customers’ instances run on the same physical server
• Limitations:
  • Not suitable for strict compliance or licensing needs

Dedicated Hosts

• What It Is: Entire physical server reserved for your use
• Benefits:
  • Required for some licensing models (e.g., Windows Server, SQL Server)
  • Helps meet compliance and audit requirements
• Limitations:
  • More expensive

Feature	Explanation	AWS	Azure	Google Cloud
Host dedicated to one customer	Physical hosts dedicated to one customer	EC2 Dedicated Hosts	Azure Dedicated Hosts	Sole-tenant nodes

---

What are the Compute Options in Cloud? #

---

Category	AWS	Azure	GCP
IaaS	Amazon EC2	Azure VMs	Google Compute Engine
PaaS	AWS Elastic Beanstalk	Azure App Service	App Engine
FaaS - Serverless	AWS Lambda	Azure Functions	Cloud Functions
CaaS - Serverless	AWS Fargate	Azure Container Instances	Cloud Run, GKE Autopilot
CaaS - Kubernetes	Amazon EKS	Azure Kubernetes Service, Azure Red Hat OpenShift	Google Kubernetes Engine
CaaS - Custom	Amazon ECS
Multi-cloud Container services	Amazon EKS Anywhere	Azure Arc-enabled Kubernetes	Google Kubernetes Engine (GKE) Enterprise edition
VMWare	VMware Cloud on AWS	Azure VMware Solution	VMware Engine