UNIT V

Cloud Computing

In-Depth Study Notes

Cloud Fundamentals • Virtualisation • Services & Platforms

Topics Covered

1. Definition of Cloud Computing

2. Characteristics of Cloud

3. Cloud Deployment Models

4. Cloud Service Models

5. Driving Factors and Challenges of Cloud

6. Virtualisation

7. Load Balancing

8. Scalability and Elasticity

9. Replication
10. Monitoring

11. Cloud Services & Platforms: Compute Services

12. Storage Services

13. Application Services

★ Points to Remember

★ Important Questions (Part-A & Part-B)

 1. Definition of Cloud Computing

NIST Definition: "Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access
to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and
services) that can be rapidly provisioned and released with minimal management effort or service provider
interaction." — NIST SP 800-145

In Simple Terms:
Cloud computing is the delivery of computing services — including servers, storage, databases, networking,
software, analytics, and intelligence — over the Internet ("the cloud") to offer faster innovation, flexible resources,
and economies of scale. Users pay only for what they use.

Key Aspects:
• Resources are delivered as services over the Internet.
• Eliminates the need for on-premises physical hardware and infrastructure.
• Follows a pay-as-you-go or subscription pricing model.
• Provides self-service, on-demand access without human interaction from the provider.
• Resources appear unlimited from the user's perspective (resource pooling).

■ Cloud computing shifts computing from a product to a service model — just like electricity from a power grid.

2. Characteristics of Cloud Computing

NIST identifies 5 essential characteristics of cloud computing:

On-Demand Self-Service A consumer can provision computing capabilities such as server time and network
storage automatically without requiring human interaction with each service
provider. Users get what they need, when they need it.

Broad Network Access Capabilities are available over the network and accessed through standard
mechanisms promoting use by heterogeneous thin or thick client platforms (e.g.,
mobile phones, tablets, laptops, workstations).

Resource Pooling The provider's computing resources are pooled to serve multiple consumers using a
multi-tenant model, with different physical and virtual resources dynamically
assigned according to consumer demand. The customer generally has no control or
knowledge over the exact location of the provided resources.

Rapid Elasticity Capabilities can be elastically provisioned and released, in some cases
automatically, to scale rapidly outward and inward commensurate with demand. To
the consumer, the capabilities available for provisioning often appear to be
unlimited.
Measured Service Cloud systems automatically control and optimize resource use by leveraging a
metering capability. Resource usage can be monitored, controlled, and reported —
providing transparency for both the provider and consumer. Users pay only for what
they consume.

Additional Characteristics:
• Multi-tenancy — Multiple users share the same physical infrastructure while data remains isolated.
• High availability and fault tolerance — Services are designed to remain operational despite failures.
• Geographic distribution — Data centers across multiple regions provide low latency and redundancy.
• Automatic updates — Software and security patches are applied by the provider.

3. Cloud Deployment Models

A cloud deployment model defines where the cloud infrastructure is located and who manages it. There are four
main deployment models:

Public Cloud

Description: Infrastructure is owned and operated by a third-party cloud service provider and delivered over the
Internet. Examples: AWS, Microsoft Azure, Google Cloud Platform. Resources are shared among multiple
organizations.

Advantages: Cost-effective, highly scalable, no maintenance, global reach.

Disadvantages: Less control, security & compliance concerns, vendor lock-in.

Private Cloud

Description: Cloud infrastructure is used exclusively by a single organization. It can be hosted on-premises or by
a third party. Examples: VMware, OpenStack, IBM Cloud Private.

Advantages: Greater control, enhanced security, customizable, compliance-friendly.

Disadvantages: Higher cost, requires in-house expertise, limited scalability.

Hybrid Cloud

Description: Combination of public and private clouds, bound by technology that enables data and application
portability. Sensitive workloads on private, scalable workloads on public. Example: Azure Arc, AWS Outposts.

Advantages: Flexibility, optimized workload placement, cost balance.

Disadvantages: Complex management, integration challenges, potential latency.

Community Cloud
Description: Infrastructure shared by several organizations with common concerns (e.g., security, compliance,
mission). Managed by the organizations or a third party. Example: Government cloud, healthcare cloud.

Advantages: Cost sharing, collaborative, tailored to community needs.

Disadvantages: Limited scalability compared to public cloud.

 4. Cloud Service Models
Cloud service models define the type of service offered and the division of management responsibilities between
the provider and the consumer. The three fundamental models form the "Cloud Pyramid."

IaaS — Infrastructure as a Service

What it provides: Provides virtualized computing infrastructure over the Internet. The provider manages physical
hardware; the user manages OS, middleware, and applications.

Examples: Virtual Machines, Storage, Networking. Platforms: AWS EC2, Google Compute Engine, Azure VMs,
DigitalOcean

Advantages: Full control over OS and software stack, Highly flexible and scalable, Pay for actual usage

Best for: Developers, IT admins needing full infrastructure control.

PaaS — Platform as a Service

What it provides: Provides a platform and environment to allow developers to build applications and services.
Provider manages infrastructure + runtime; user manages applications and data.

Examples: Development tools, Databases, Middleware, Runtime environments. Platforms: Google App Engine,
AWS Elastic Beanstalk, Heroku, Azure App Service

Advantages: Faster development, No infrastructure management, Built-in scalability and DevOps tools

Best for: Developers building applications without managing underlying infrastructure.

SaaS — Software as a Service

What it provides: Delivers software applications over the Internet on a subscription basis. Provider manages
everything; user just uses the application through a browser or thin client.

Examples: Complete applications: email, CRM, ERP. Platforms: Gmail, Microsoft 365, Salesforce, Zoom,
Dropbox

Advantages: No installation needed, Accessible from anywhere, Automatic updates, Low upfront cost

Best for: End users and businesses needing ready-to-use software.

Responsibility Comparison: IaaS → User manages: OS, Middleware, Apps, Data | PaaS → User manages:
Apps, Data | SaaS → User manages: Nothing (just data input). As you move from IaaS→PaaS→SaaS, provider
responsibility increases.

5. Driving Factors and Challenges of Cloud

Driving Factors (Why Organizations Adopt Cloud):

Cost Reduction Eliminates capital expenditure (CapEx) on hardware. Shifts to operational

expenditure (OpEx). Pay-as-you-go model reduces waste.

Scalability Resources can be scaled up or down instantly based on demand without upfront
investment.

Agility & Speed Provision new resources in minutes. Faster time-to-market for applications and
services.

Global Reach Deploy applications in multiple geographic regions instantly, improving user
experience worldwide.

Business Continuity Built-in redundancy, backup, and disaster recovery capabilities minimize downtime
risk.

Innovation Access to cutting-edge services like AI/ML, IoT, and Big Data without building from
scratch.

Remote Work Enables distributed teams to access resources and collaborate from anywhere.

Sustainability Large-scale data centers are more energy-efficient than many small on-premises
deployments.

Challenges and Concerns:

Security & Privacy Data breaches, unauthorized access, data loss. Shared infrastructure increases
attack surface.

Compliance & Legal GDPR, HIPAA, and other regulations may restrict where data can be
stored/processed.

Vendor Lock-in Proprietary APIs and services make it difficult to migrate between providers.

Downtime & Reliability Even major providers experience outages. SLAs guarantee uptime but don't
eliminate risk.

Latency Network dependency introduces latency, critical for real-time applications.

Data Transfer Costs Egress (outbound) data transfer fees can be significant for data-intensive
applications.

Limited Control Users have less control over infrastructure and underlying hardware.

Skill Gap Cloud adoption requires new skills; shortage of trained cloud professionals.

Shadow IT Employees using unauthorized cloud services create security and compliance risks.
 6. Virtualisation

Definition: Virtualisation is the process of creating a software-based (virtual) representation of hardware,

storage, network, or operating systems. It allows a single physical machine to run multiple virtual machines
(VMs) simultaneously, each with its own OS and applications.

How Virtualisation Works:

A software layer called the Hypervisor (also called Virtual Machine Monitor — VMM) is installed on top of the
physical hardware. The hypervisor allocates hardware resources (CPU, RAM, storage) to each virtual machine.
Each VM is isolated from others and believes it has dedicated hardware.

Types of Hypervisors:

Type Description Examples

Type 1 (Bare-Metal) Runs directly on physical hardware. VMware ESXi, Microsoft Hyper-V,
No host OS needed. Better Xen, KVM
performance.

Type 2 (Hosted) Runs on top of a host OS. Easier to VMware Workstation, VirtualBox,
set up, less efficient. Parallels Desktop

Types of Virtualisation:

Server Virtualisation Multiple virtual servers on one physical server. Most common form. Improves server
utilization.

Desktop Virtualisation Virtual Desktop Infrastructure (VDI) — desktops hosted centrally and accessed
remotely.

Network Virtualisation Combines hardware and software resources into a virtual network. Example: VLAN,
SDN.

Storage Virtualisation Pools physical storage from multiple devices into a single virtual storage device.
Example: SAN.

Application Applications run in isolated environments without being installed on the local
Virtualisation machine. Example: Docker containers.

OS Virtualisation Containerization — multiple isolated user-space instances (containers) on a single

kernel. Example: Docker, Kubernetes.

Benefits of Virtualisation:
• Improved hardware utilization (from ~5-15% to 60-80%).
• Reduced hardware costs — fewer physical servers needed.
• Faster server provisioning — spin up a VM in minutes.
• Improved disaster recovery — VM snapshots and live migration.
 • Isolation — problems in one VM do not affect others.
• Test and development — safe sandbox environments.
• Foundation for cloud computing — cloud platforms are built on virtualisation.

■ Virtualisation is the FOUNDATION of cloud computing. Without virtualisation, multi-tenant cloud services and dynamic
resource allocation would be impossible.

7. Load Balancing

Definition: Load balancing is the process of distributing incoming network traffic or computational workloads
across multiple servers or resources to ensure no single server becomes overwhelmed, maximizing throughput,
minimizing response time, and avoiding overload.

Load Balancing Algorithms:

Algorithm How It Works Best For

Round Robin Requests distributed sequentially to Servers with equal capacity.

each server in turn.

Weighted Round Robin Servers assigned weights; Servers with different capacities.
higher-weight servers get more
requests.

Least Connections New requests go to the server with Long-lived connections.

the fewest active connections.

IP Hash Client IP determines which server Applications requiring sticky

handles the request (session sessions.
persistence).

Random Requests assigned randomly to Simple, equal-capacity

servers. environments.

Resource Based Routes based on real-time server Dynamic, varied workloads.

health (CPU, RAM usage).

Types of Load Balancers:

• Hardware Load Balancers — Dedicated physical devices (e.g., F5 BIG-IP). High performance, expensive.
• Software Load Balancers — Software running on commodity hardware (e.g., HAProxy, Nginx, Apache).
• Cloud Load Balancers — Managed services by cloud providers (e.g., AWS ELB, Azure Load Balancer, GCP
Load Balancer).
• DNS Load Balancing — Using DNS to distribute traffic across multiple IP addresses.
• Global Server Load Balancing (GSLB) — Distributes traffic across geographically dispersed data centers.

Benefits of Load Balancing:

• High Availability — If one server fails, traffic is redirected to healthy servers.
• Improved Performance — Requests are distributed, reducing response time.
• Scalability — New servers can be added to the pool without downtime.
• Session Persistence — Sticky sessions ensure users connect to the same server.
• Health Monitoring — Load balancers continuously check server health.
• SSL Termination — Offloads SSL decryption from backend servers.
 8. Scalability and Elasticity

Scalability:

Definition: Scalability is the ability of a system to handle increasing workload by adding resources, without
affecting performance. It refers to the system's capacity to grow in a planned, predictable manner to meet
long-term demand growth.

Type Description Example

Vertical Scaling (Scale Up) Adding more power (CPU, RAM) to Upgrading a server from 8GB to
an existing server. 32GB RAM.

Horizontal Scaling (Scale Out) Adding more servers/nodes to Adding more web servers behind a
distribute load. load balancer.

Diagonal Scaling Combination of both vertical and Upgrading servers AND adding
horizontal scaling. more nodes.

Elasticity:

Definition: Elasticity is the ability of a system to automatically provision and de-provision resources in real time
based on actual demand. It is dynamic and responsive — resources scale up during peak demand and scale
down (and release resources) when demand drops, optimizing cost.

Scalability vs. Elasticity — Key Differences:

Aspect Scalability Elasticity

Focus Handling growth Handling fluctuating demand

Automation Can be manual or automated Automatic (real-time)

Direction Primarily scaling up/out Both scaling up AND down

Cost Model May over-provision resources Pay only for what is used

Timeframe Long-term capacity planning Short-term, real-time adjustment

Example Adding servers for planned growth Auto-scaling during Black Friday
traffic spike

Auto-Scaling in Cloud:
• Cloud providers offer Auto-Scaling services (AWS Auto Scaling, Azure Scale Sets, GCP Autoscaler).
• Policies define when to scale (CPU > 80% → add instance; CPU < 20% → remove instance).
• Works with load balancers to distribute traffic across scaled instances.
• Ensures optimal performance during traffic spikes and cost savings during low traffic.
 • Supports scheduled scaling for predictable patterns (e.g., scale up every weekday morning).

9. Replication

Definition: Replication is the process of copying and maintaining identical data or services across multiple
locations, servers, or data centers to ensure availability, fault tolerance, consistency, and performance.

Types of Replication:

Synchronous Data is written to primary and replica simultaneously. Write operation completes
Replication only after all copies confirm. Guarantees zero data loss. Higher latency.

Asynchronous Data is written to primary first, then replicated to copies with a delay. Lower latency,
Replication risk of minor data loss if failure occurs before sync.

Semi-Synchronous Primary waits for at least one replica to confirm receipt, then continues. Balance
between safety and performance.

Replication Strategies:

Strategy Description Example

Master-Slave (Primary-Replica) One master handles writes; replicas MySQL Replication

handle reads.

Master-Master Multiple nodes can handle reads Cassandra, CouchDB

and writes.

Geo-Replication Data replicated across Azure Geo-Redundant Storage

geographically distributed data (GRS)
centers.

Snapshot Replication Periodic snapshots of data are Database backups, VM snapshots

replicated.

Transaction Replication Each transaction is replicated SQL Server Replication

individually to subscribers.

Benefits of Replication:
• High Availability — System remains operational even if one node fails.
• Disaster Recovery — Data preserved in geographically separate locations.
• Improved Read Performance — Read operations distributed across replicas.
• Load Distribution — Queries distributed across multiple replica nodes.
• Data Durability — Reduces risk of permanent data loss.
 10. Monitoring in Cloud

Definition: Cloud monitoring is the process of reviewing, observing, and managing the operational workflow and
processes within a cloud-based IT infrastructure. It involves continuous collection, analysis, and reporting of
metrics, logs, and events to ensure optimal performance, availability, and security.

What is Monitored:
• Infrastructure Metrics — CPU usage, memory utilization, disk I/O, network throughput.
• Application Performance — Response times, error rates, transaction volumes, user activity.
• Availability / Uptime — Whether services and endpoints are reachable.
• Security Events — Unauthorized access attempts, suspicious patterns, configuration changes.
• Cost and Usage — Cloud spending, resource consumption, billing anomalies.
• Logs — Application logs, system logs, audit logs for debugging and compliance.
• SLA Compliance — Ensuring service levels meet agreed-upon targets.

Types of Cloud Monitoring:

Infrastructure Monitors physical and virtual hardware: servers, VMs, containers, networks.

Application Tracks application performance (APM) — response times, throughput, errors.

Database Query performance, connection counts, deadlocks, storage utilization.

Network Bandwidth, packet loss, latency, firewall rules, traffic patterns.

Security (SIEM) Security Information and Event Management — detects threats and compliance
violations.

Cost Tracks spending against budgets, identifies waste, forecasts future costs.

Popular Cloud Monitoring Tools:

Tool Provider Key Features

CloudWatch AWS Metrics, logs, alarms, dashboards,

auto-scaling triggers.

Azure Monitor Microsoft Full-stack monitoring, Log Analytics,

Application Insights.

Cloud Monitoring Google Metrics, uptime checks, alerting,

dashboards.

Prometheus Open Source Time-series metrics, alerting,

Kubernetes-native.
Grafana Open Source Visualization dashboards, integrates
with multiple sources.

Datadog Third Party APM, infrastructure, logs, real-user

monitoring.

New Relic Third Party Full-stack observability, AI-driven

insights.

Key Monitoring Concepts:

• Alerting — Notifications triggered when metrics breach thresholds (email, SMS, webhook).
• Dashboards — Visual representation of key metrics in real-time.
• Log Aggregation — Centralized collection of logs from all components.
• Tracing — Tracking a request as it travels through distributed services (distributed tracing).
• Observability — The ability to understand a system's internal state from its external outputs (logs, metrics,
traces).
• SLA / SLO / SLI — Service Level Agreements, Objectives, and Indicators define and measure service quality.

11. Cloud Compute Services

Compute services form the backbone of cloud platforms, providing processing power to run applications, perform
calculations, and execute workloads.

Types of Compute Services:

Virtual Machines (VMs)

Description: Emulate complete computers with dedicated OS instances. Full control over OS, software, and
configuration. Examples: AWS EC2, Azure Virtual Machines, GCP Compute Engine.

Key Features: Choose OS (Linux/Windows) | Wide instance type selection | Persistent storage | Full root/admin
access

Use Cases: Legacy applications, custom configurations, long-running workloads.

Containers

Description: Lightweight, portable units that package application code and dependencies. Share host OS kernel
but are isolated. Much faster startup than VMs. Examples: AWS ECS/EKS, Azure AKS, GCP GKE (Kubernetes).

Key Features: Faster than VMs | Consistent across environments | Microservices architecture | Docker +
Kubernetes

Use Cases: Microservices, CI/CD pipelines, scalable web applications.

Serverless / FaaS
 Description: Execute code in response to events without managing servers. Provider handles infrastructure.
Billed per execution. Examples: AWS Lambda, Azure Functions, GCP Cloud Functions.

Key Features: No server management | Auto-scales to zero | Pay per millisecond of execution | Event-driven

Use Cases: Event-driven apps, APIs, data processing, automation tasks.

Batch Computing

Description: Run large-scale parallel and batch computing jobs. Ideal for high-performance computing (HPC).
Examples: AWS Batch, Azure Batch, GCP Dataflow.

Key Features: Parallel processing | Schedule jobs | Spot/preemptible instances for cost savings

Use Cases: Scientific simulations, rendering, genomics, financial modeling.

Comparison — VMs vs Containers vs Serverless:

Feature Virtual Machines Containers Serverless

Isolation Full OS isolation Process isolation Function isolation

Startup Time Minutes Seconds Milliseconds

Resource Use High (full OS) Low (shared kernel) Minimal (per invocation)

Portability Less portable Highly portable Provider-dependent

Billing Per hour/second (running) Per resource (running) Per execution (100ms)

Management User manages OS + App User manages App + User manages only code
Container
 12. Cloud Storage Services
Cloud storage provides scalable, durable, and accessible data storage over the Internet. Different storage types
are optimized for different use cases.

Object Storage

How it works: Stores data as objects (files + metadata + unique ID) in a flat namespace. Highly scalable,
accessed via APIs (HTTP/REST). Not suitable for frequent small modifications.

Examples: AWS S3, Azure Blob Storage, GCP Cloud Storage.

Use Cases: Unstructured data, media files, backups, static website hosting, big data analytics.

Block Storage

How it works: Stores data in fixed-size blocks. Behaves like a physical hard drive. Low latency, high
performance. Directly attached to compute instances.

Examples: AWS EBS, Azure Managed Disks, GCP Persistent Disk.

Use Cases: Databases, VM boot volumes, high-performance applications requiring raw storage.

File Storage

How it works: Shared file system accessed over a network using standard protocols (NFS, SMB/CIFS). Multiple
instances can access the same file system simultaneously.

Examples: AWS EFS, Azure Files, GCP Filestore.

Use Cases: Shared content repositories, home directories, media workflows, lift-and-shift applications.

Archive Storage

How it works: Low-cost storage for infrequently accessed data. Very low per-GB cost but higher retrieval latency
(minutes to hours). Ideal for long-term retention.

Examples: AWS S3 Glacier, Azure Archive Storage, GCP Coldline/Archive.

Use Cases: Regulatory archiving, backup retention, rarely accessed historical data.

In-Memory / Cache

How it works: Stores data in RAM for extremely fast access (microsecond latency). Used as a caching layer in
front of databases to reduce load and improve response times.

Examples: AWS ElastiCache (Redis/Memcached), Azure Cache for Redis, GCP Memorystore.

Use Cases: Session caching, real-time leaderboards, fast data lookup, database caching.
Property Object Block File Archive

Access Method HTTP/REST API Direct I/O NFS/SMB API (with delay)

Latency Medium Very Low Low High (retrieval

delay)

Cost Low Higher Medium Very Low

Scalability Virtually unlimited Limited by instance Moderate Virtually unlimited

Best For Media, backups Databases, VMs Shared filesystems Long-term archives

13. Cloud Application Services

Cloud application services are managed services that provide common application capabilities as pre-built,
scalable components, allowing developers to focus on application logic rather than infrastructure management.

Messaging & Queuing Services

Description: Enable asynchronous communication between application components by placing messages in a

queue. Decouples producers and consumers for resilient, scalable architectures.

Platforms: AWS SQS/SNS, Azure Service Bus/Event Grid, GCP Pub/Sub.

Use Cases: Microservices communication, task queuing, event-driven architecture.

API Gateway

Description: Managed service that acts as a front door for APIs. Handles authentication, rate limiting, caching,
routing, SSL termination, and monitoring for all API calls.

Platforms: AWS API Gateway, Azure API Management, GCP Apigee.

Use Cases: RESTful APIs, mobile backends, microservices exposure.

Email & Notification Services

Description: Cloud-managed services for sending emails, SMS, and push notifications at scale without managing
email servers or notification infrastructure.

Platforms: AWS SES/SNS, Azure Communication Services, SendGrid, Twilio.

Use Cases: Transactional emails, marketing campaigns, alert notifications.

Content Delivery Network (CDN)

Description: Distributed network of servers that caches content close to end users to reduce latency and improve
load times. Routes requests to nearest edge server.

Platforms: AWS CloudFront, Azure CDN, GCP Cloud CDN, Cloudflare.

Use Cases: Static asset delivery, video streaming, global web applications.
Identity & Authentication Services

Description: Managed services for user identity management, authentication (MFA), authorization, single sign-on
(SSO), and social login without building from scratch.

Platforms: AWS Cognito, Azure Active Directory B2C, Google Identity Platform, Auth0.

Use Cases: User registration/login, SSO, OAuth/OIDC implementation.

AI / ML Services

Description: Pre-built AI models and ML platforms available as APIs. Includes vision, language, speech,
recommendation engines, and custom model training without ML expertise.

Platforms: AWS SageMaker/Rekognition, Azure AI/Cognitive Services, GCP Vertex AI/Vision API.

Use Cases: Image recognition, NLP, chatbots, fraud detection, recommendation systems.

Database Services (DBaaS)

Description: Fully managed database services where the cloud provider handles provisioning, patching,
backups, replication, and scaling. Both relational and NoSQL options available.

Platforms: AWS RDS/DynamoDB, Azure SQL Database/Cosmos DB, GCP Cloud SQL/Firestore.

Use Cases: Web applications, analytics, real-time data, global distributed data.

Search Services

Description: Managed full-text search and analytics engines that allow complex search queries over large
datasets with features like autocomplete, faceted search, and ranking.

Platforms: AWS OpenSearch, Azure Cognitive Search, GCP Cloud Search, Elastic Cloud.

Use Cases: E-commerce product search, log analytics, knowledge bases.

 ★ Points to Remember

Cloud Computing Basics:

✦ NIST defines 5 essential characteristics: On-demand self-service, Broad network access, Resource pooling,
Rapid elasticity, Measured service.
✦ Cloud follows a Pay-as-you-go or subscription model — CapEx replaced by OpEx.
✦ There are 4 deployment models: Public, Private, Hybrid, Community.
✦ There are 3 service models: IaaS (infrastructure), PaaS (platform), SaaS (software).
✦ In IaaS, user manages OS + App. In PaaS, user manages only App. In SaaS, user manages nothing.
✦ Cloud = Virtualisation + Internet + Automation + Pay-per-use.

Virtualisation:
✦ Hypervisor = VMM = software layer that creates and manages VMs.
✦ Type 1 hypervisor runs on bare metal (faster). Type 2 runs on host OS (easier to set up).
✦ Virtualisation improves server utilization from ~10% to ~70-80%.
✦ Containerization (Docker/Kubernetes) is OS-level virtualisation — lighter than full VMs.
✦ VM Snapshot = instant backup of a VM state at a point in time.

Load Balancing, Scalability & Elasticity:

✦ Load balancing distributes traffic across multiple servers — prevents single point of failure.
✦ Round Robin = simplest algorithm. Least Connections = best for long-lived sessions.
✦ Vertical Scaling (Scale Up) = adding more power to one machine.
✦ Horizontal Scaling (Scale Out) = adding more machines — preferred in cloud.
✦ Elasticity = AUTOMATIC and BIDIRECTIONAL scaling (up AND down). Scalability can be planned.
✦ Auto-scaling policies trigger based on CPU/memory/request count thresholds.

Replication & Monitoring:

✦ Synchronous replication = zero data loss, higher latency.
✦ Asynchronous replication = possible minor data loss, lower latency.
✦ Geo-replication = data in multiple geographic regions for disaster recovery.
✦ The 3 pillars of observability are: Metrics, Logs, and Traces.
✦ CloudWatch (AWS), Azure Monitor (Azure), Cloud Monitoring (GCP) are provider-native tools.

Compute, Storage & App Services:

✦ Object Storage: Files as objects, accessed via API. Best for unstructured data (AWS S3).
✦ Block Storage: Like a hard drive for VMs. Low latency. Best for databases (AWS EBS).
✦ File Storage: Shared filesystem over network. Multiple VMs can access (AWS EFS).
✦ Archive Storage: Cheapest, retrieval takes time. Best for compliance archives.
✦ Serverless/FaaS: No server management, pay per execution, auto-scales to zero.
✦ CDN brings content closer to users by caching at edge servers globally.
✦ API Gateway manages security, rate limiting, and routing for all API calls.
✦ SaaS examples: Gmail, Microsoft 365, Salesforce. IaaS: AWS EC2. PaaS: Heroku, App Engine.
 Important Questions — Part-A (2 Marks)
Part-A questions require concise answers of 2–3 sentences or a definition.

1 Define Cloud Computing. (NIST Definition)

.
2 List the five essential characteristics of cloud computing.
.
3 What is a Cloud Deployment Model? Name the four types.
.
4 Define IaaS. Give two examples.
.
5 What is SaaS? Give three real-world examples.
.
6 Define Virtualisation. What is a Hypervisor?
.
7 Differentiate between Type 1 and Type 2 Hypervisors.
.
8 What is Load Balancing? Why is it needed?
.
9 Define Scalability. What is Vertical Scaling?
.
1 What is Elasticity in cloud computing?
0
.
1 Define Replication in cloud computing.
1
.
1 What is Synchronous Replication? How does it differ from Asynchronous?
2
.
1 What is Cloud Monitoring? Name two monitoring tools.
3
.
1 Define Object Storage. Give one example.
4
.
1 What is Serverless Computing / Function as a Service (FaaS)?
5
.
1 What is a Content Delivery Network (CDN)?
6
.
1 What is Multi-tenancy in cloud computing?
7
.
1 List any four driving factors for cloud adoption.
8
.
1 What are the main challenges/concerns in cloud computing?
9
.
2 Differentiate between Public Cloud and Private Cloud.
0
.
2 What is an API Gateway? Name one cloud API Gateway service.
1
.
2 What is Block Storage? How does it differ from File Storage?
2
.
2 Define Round Robin load balancing algorithm.
3
.
2 What is Geo-Replication?
4
.
2 What are the three pillars of Observability?
5
.

Important Questions — Part-B (13 / 16 Marks)

Part-B questions require detailed answers with diagrams, comparisons, and examples. Expected length: 2–4 full pages
per question.

1 Explain Cloud Computing with its definition, characteristics, and advantages. How does it differ from
. traditional computing?

2 Discuss the Cloud Deployment Models (Public, Private, Hybrid, Community) with neat diagrams,
. advantages, and disadvantages of each.

3 Explain the Cloud Service Models — IaaS, PaaS, and SaaS — with examples, diagrams, responsibility
. comparison, and use cases.

4 What is Virtualisation? Explain the types of virtualisation, types of hypervisors, and the benefits of
. virtualisation in cloud computing.

5 Explain Load Balancing in cloud computing. Discuss various load balancing algorithms with examples and
. types of load balancers.

6 Discuss Scalability and Elasticity in cloud computing. Compare vertical vs horizontal scaling. Explain
. Auto-Scaling with an example.

7 Explain Replication in cloud computing. Discuss synchronous vs asynchronous replication, replication

. strategies, and its importance for fault tolerance.
8 Explain Cloud Monitoring — what is monitored, types of monitoring, key concepts (metrics, logs, traces,
. alerting), and popular monitoring tools.

9 Describe the Cloud Compute Services: Virtual Machines, Containers, Serverless Computing, and Batch
. Computing. Compare them with a detailed table.

1 Explain the different types of Cloud Storage Services (Object, Block, File, Archive, In-Memory). Compare
0 them and discuss their appropriate use cases.
.

1 Discuss Cloud Application Services — API Gateway, CDN, Messaging/Queuing, Identity Services, AI/ML
1 Services, and DBaaS — with examples.
.

1 What are the driving factors and challenges of cloud computing? Discuss in detail with real-world examples.
2
.

1 Write a detailed note on Cloud Services and Platforms provided by AWS, Microsoft Azure, and Google
3 Cloud Platform. Compare their offerings.
.

1 Explain the concept of Virtualisation and its role in enabling Cloud Computing. How do hypervisors, VMs,
4 and containers relate to cloud services?
.

1 (a) Explain the difference between Scalability and Elasticity with examples. (b) Explain Replication types
5 and their role in high availability.
.

1 Write short notes on: (a) Load Balancing Algorithms (b) Cloud Monitoring Tools (c) Serverless Computing
6 (d) Object vs Block Storage
.

■ Unit V — Cloud Computing | Study Notes | All topics as per syllabus | Practice diagrams for deployment/service models,
virtualisation architecture, and load balancing for best exam results.