Cloud Computing
Presented By:
Dr. Sukhwinder Sharma
Associate Professor
Department of Data Science and Engineering
Manipal University Jaipur, Jaipur
Syllabus
•Introduction to Cloud Computing: Definition, Characteristics, History, Deployment Models (Public, Private,
Hybrid, Community), Service Models (IaaS, PaaS, SaaS), Cloud Architecture, Cloud Providers and Services,
Cost Benefit Analysis of Cloud Adoption.
•Virtual Machines Provisioning and Migration Services: Virtualization Concepts, Types of Virtualization,
Hypervisors, Creating and Managing Virtual Machines, Containerization (Docker, Kubernetes), High
Availability and Disaster Recovery, Cloud Migration Concepts, Cloud Migration Techniques.
•SLA Management, Cloud Security and AWS Services: Service Level Agreement (SLA), SLA Management
in Cloud, Automated Policy-based Management, Cloud Security Fundamentals, Security Challenges in the
Cloud, Vulnerability Assessment, Security and Privacy, Cloud Computing Security Architecture, Amazon Web
Services (AWS), AWS Services - Identity and Access Management (IAM), and Virtual Private Cloud (VPC).
•Advanced Topics: Serverless Computing, Edge Computing, Managed databases (RDS, NoSQL), Data
Warehousing Solutions (Redshift, BigQuery), AI/ML services in the cloud (AWS SageMaker, Google AI
Platform), Real-world Cloud Computing Case Studies, Discussion on Cloud Adoption in Various Industries.
Serverless Computing
• Serverless computing is an application development and execution model
that enables developers to build and run application code without
provisioning or managing servers or back-end infrastructure.
• Serverless does not mean "no servers." The name notwithstanding, servers
in serverless computing are managed by a cloud service provider (CSP).
• Serverless describes the developer's experience with those servers—they
are invisible to the developer, who doesn't see them, manage them or
interact with them in any way.
• Developers can focus on writing the best front-end application code and
business logic with serverless computing. All they need to do is write their
application code and deploy it to containers managed by a CSP.
• The cloud provider handles the rest—provisioning the cloud infrastructure required
to run the code and scaling the infrastructure up and down on demand as needed—
and is also responsible for all routine infrastructure management and maintenance,
such as operating system updates and patches, security management, capacity
planning, system monitoring and more.
• Moreover, developers never pay for idle capacity with serverless. The cloud provider
spins up and provisions the required computing resources on demand when the
code executes and spins them back down again—called ''scaling to zero''—when
execution stops. The billing starts when execution starts and ends when execution
stops; typically, pricing is based on execution time and resources required.
Available at https://www.ibm.com/think/topics/serverless
• Examples: AWS Lambda, Azure Functions, Google Cloud Functions.
• Benefits:
• Reduced operational complexity.
• Cost efficiency (pay only for execution time).
• Scalability and flexibility.
• Use Cases:
• Event-driven applications (e.g., notifications, image processing).
• Backend for web and mobile applications.
• IoT and real-time data processing.
• Challenges and Considerations
• Cold start latency affecting performance.
• Vendor lock-in due to platform-specific implementations.
• Limited execution time and resource constraints.
• Debugging and monitoring complexities.
Edge Computing
• Computing that happens close to the data source instead of a centralized cloud.
• Reduces latency and bandwidth use by processing data near the user/device.
• Used in IoT, autonomous systems, and real-time applications.
• Advantages:
• Faster response times.
• Improved security (localized data processing).
• Reduced dependency on centralized cloud.
• Use Cases:
• Smart cities (traffic monitoring, surveillance).
• Industrial IoT (predictive maintenance).
• AR/VR applications and autonomous vehicles.
• Challenges and Future Outlook
• Limited computing power at edge nodes.
• Security risks due to distributed architecture.
• Need for standardized protocols for interoperability.
Managed databases (RDS, NoSQL)
• Cloud providers offer managed database services to handle provisioning, scaling, and maintenance.
• Two main types:
• Relational (SQL) databases (AWS RDS, Google Cloud SQL) and
• NoSQL databases (DynamoDB, Firebase, Cosmos DB).
• Benefits:
• Automated backups, scaling, and security.
• High availability and disaster recovery.
• Cost-efficient pay-as-you-go model.
• Use Cases:
• RDS: Traditional business applications, e-commerce platforms.
• NoSQL: Real-time analytics, IoT applications, social media platforms.
• Use Cases: Comparison and Considerations
• SQL (structured) vs. NoSQL (flexible schema).
• Performance trade-offs (RDS better for transactions, NoSQL better for scale).
• Vendor-specific optimizations (e.g., AWS Aurora for SQL, Firebase for mobile apps).
Data Warehousing Solutions (Redshift,
BigQuery)
• Large-scale storage and processing of structured data for analytics.
• Cloud-based solutions like Amazon Redshift and Google BigQuery offer high-speed querying and
scalability.
• Benefits:
• Optimized for complex analytical queries.
• Scalable and cost-efficient for big data workloads.
• Integrated with BI tools (Tableau, Looker).
• Use Cases:
• Business intelligence and reporting.
• Financial analysis, healthcare analytics, customer behavior tracking.
• Challenges and Selection Criteria:
• Cost considerations for storage and queries.
• Performance differences: Redshift (columnar, optimized for batch processing), BigQuery (serverless, faster for ad-
hoc queries).
• Security and compliance requirements.
Redshift
• A fully managed, petabyte-scale data warehouse service by AWS.
• Designed for fast querying and analysis of large datasets.
• Use Cases
• Business Intelligence (BI)
• Big data analytics
• Data lake integration
• Key Features
• Columnar Storage – Efficient storage and faster query performance.
• Massively Parallel Processing (MPP) – Distributes queries across nodes.
• Scalability – Easily scale from a few hundred GBs to petabytes.
• Integration with AWS services – S3, Glue, Athena, and more.
BigQuery
• Google Cloud’s fully-managed, serverless data warehouse.
• Designed for fast SQL queries using the processing power of Google’s
infrastructure.
• Use Cases
• Real-time analytics
• Business intelligence
• Machine learning integration
• Key Features
• Serverless Architecture – No need to manage infrastructure.
• Standard SQL Support – Easy for analysts and developers.
• Real-time Analytics – Streaming data ingestion.
• Built-in ML & AI – BigQuery ML for model training using SQL.
AI/ML services in the cloud (AWS
SageMaker, Google AI Platform)
• Cloud platforms like AWS SageMaker and Google AI Platform offer
powerful services for building, training, and deploying machine learning
(ML) models at scale.
• When to Choose What?
• Choose AWS SageMaker if:
• You're already invested in AWS.
• You want full control with powerful DevOps/MLOps tools.
• You need scalable and customizable training/deployment.
• Choose Google Vertex AI if:
• You're working with big data (BigQuery, GCS).
• You prefer ease of use and automation.
• You're building ML-powered mobile/edge applications.
AWS SageMaker
• AWS SageMaker is a fully managed service by Amazon Web Services that covers
the entire ML workflow—from data labeling to model deployment.
• Key Features:
• Built-in algorithms: Pre-built models like XGBoost, linear learner, etc.
• AutoML (SageMaker Autopilot): Automatically builds, trains, and tunes models.
• Jupyter Notebooks: Integrated notebook instances for data exploration.
• Training & Tuning: Scalable model training with hyperparameter tuning.
• Model Deployment: One-click deployment with auto-scaling endpoints.
• MLOps (SageMaker Pipelines): Manage ML lifecycle with CI/CD tools.
• Inference Options: Real-time, batch, and asynchronous inference.
• Edge Deployment: With SageMaker Neo and AWS IoT Greengrass.
• Pricing:
• Pay-as-you-go based on instance types and usage time.
Google Cloud AI Platform (Vertex AI)
• Google AI Platform has now evolved into Vertex AI, which unifies data
engineering, AutoML, and custom model training/deployment under one roof.
• Key Features:
• Custom and AutoML models: Build your own or let Vertex do it.
• Notebooks: Pre-built notebooks using Jupyter in the cloud.
• Training: Custom training with TensorFlow, PyTorch, scikit-learn, etc.
• Deployment: Scalable model serving with prediction endpoints.
• Pipelines: Managed pipelines for orchestration (Kubeflow-based).
• Model Monitoring: Detect skew and drift in production.
• Deep integration: With BigQuery, Dataflow, GCS, and TensorBoard.
• Edge & Mobile support: TensorFlow Lite, Coral devices.
• Pricing:
• Based on usage (compute, storage, predictions). More granular control over pricing tiers.
Real-world Cloud Computing Case Studies,
Discussion on Cloud Adoption in Various Industries.
• Useful links:
• https://www.fynd.academy/blog/cloud-computing-case-study
• https://www.maropost.com/blog/5-real-world-examples-of-cloud-computing
/
• https://cioindex.com/references/cloud-computing-case-studies/
• https://securetechinstitute.com/cloud-computing-case-studies-real-world-exa
mples-for-beginners/
• https://www.infosys.com/services/cloud-cobalt/case-studies/cobalt-stories.ht
ml
