Cloud Computing & IoT PDF

Cloud Computing & IoT – Cours 1 – Chapitre 1 : Introduction au Cloud Computing Pr Meriem Belguidoum Faculté NTIC...

Cloud Computing & IoT – Cours 1 – Chapitre 1 : Introduction au Cloud Computing Pr Meriem Belguidoum Faculté NTIC Département TLSI M2 SDSI [email protected] Université Constantine 2 2024/2025 semestre 3  Plan Some History What is Cloud Computing? Why using cloud computing? The features of Cloud Computing Cloud deployment models Cloud service models Comparison between some Cloud providers Cloud computing & IoT Université Constantine 2 © Belguidoum Meriem 2  History Era Year Milestone Description Time-sharing Early form of shared computing, 1960s on Mainframes maximizing use of expensive resources. Virtual Foundation for running multiple isolated 1970s Machines environments on a single physical machine. (VMs) Rise of the Precursors Internet & ASPs Hosted software applications offered over 1990s (Application the web, a precursor to SaaS. Service Providers) Highlighted the limitations of traditional IT Late 1990s - Dot-Com Boom and the need for more flexible, scalable Early 2000s & Bust solutions. Université Constantine 2 © Belguidoum Meriem 3  History Amazon Marks the beginning of modern cloud launches AWS 2006 computing with on-demand services for (Amazon Web storage and computing power. Services) Google App Modern Engine & Major players enter the market, increasing Cloud Era 2008-2010 Microsoft Azure competition and driving innovation. Launch Further accelerates cloud adoption due to Mobile & Big 2010s the need for scalability, cost-effectiveness, Data Explosion and data processing capabilities. Serverless, The Cloud The cloud continues to evolve, enabling AI/ML, Edge Today & Present new possibilities and transforming Computing Beyond industries. Emerge Université Constantine 2 © Belguidoum Meriem 4  Real world use cases Industry Use Case Description Companies like Amazon, Alibaba, and Shopify Website and leverage the cloud to manage millions of online E-commerce platform hosting transactions, dynamically scaling resources to accommodate traffic surges. Learning Moodle, Canvas, and Blackboard utilize the cloud Education Management to deliver online learning platforms, offering Systems (LMS) students and educators anytime, anywhere access. Services like Netflix, Spotify, and Disney+ rely on Video and music the cloud to stream multimedia content to millions Entertainment streaming of global subscribers, ensuring high-quality service and minimal buffering. The cloud enables healthcare providers to securely Medical data store and analyze large volumes of medical data, Healthcare storage and including images and patient records, facilitating analysis personalized medicine and research. Online banking Financial institutions use the cloud to power Finance and algorithmic secure and reliable online banking platforms and trading execute complex trading algorithms in real-time. Université Constantine 2 © Belguidoum Meriem 5  What is Cloud Computing ? Definition The National Institute of Standards and Technology (NIST) defines cloud computing as: “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.”… Université Constantine 2 © Belguidoum Meriem 6  Why using Cloud Computing ? Before the cloud, businesses relied heavily on expensive and complex on-premises infrastructure. Managing servers, storage, software, and security required dedicated IT teams and expertise. This traditional approach came with limitations: High Costs: Significant upfront investments in hardware, software, maintenance, and upgrades. Complexity: Managing intricate infrastructure demanded specialized technical skills. Lack of Flexibility: Difficulty adapting quickly to changing needs and peak demands. Underutilized Resources: Servers often sat idle, wasting energy and resources. Université Constantine 2 © Meriem Belguidoum 7  Why using Cloud Computing ? Some challenges in real-world use cases Use Case Challenge Cloud Solution Benefits Elasticity: Cloud -Cost-efficiency: Pay Handling surges in website resources can be only for resources traffic and orders during peak Scaling for Peak scaled up or down on used. seasons (holidays, sales) Demand demand, -Improved without over-provisioning (E-Commerce) automatically performance and resources during off-peak adjusting to changing customer experience times. traffic patterns. during peak times. Cloud-based Collaboration - Enhanced Platforms: Tools like productivity and Google Workspace teamwork Enabling seamless collaboration and Microsoft 365 -Streamlined Global Team for remote teams working offer file sharing, real- workflows and Collaboration across different locations and time document version control. time zones. editing, video - Access to work conferencing, and from anywhere, centralized anytime. communication. Université Constantine 2 © Belguidoum Meriem 8  Why using Cloud Computing ? - Accelerated AI/ML Services: Cloud innovation in AI- Analyzing large datasets for providers offer pre- powered applications. insights and developing trained AI/ML models, Cost-effective access Leveraging AI & advanced AI applications scalable computing to powerful computing Big Data without investing heavily in resources, and specialized resources. expensive infrastructure tools for data analysis and - Faster time-to-market and specialized expertise. model development. for AI-driven products and services. - Enhanced business resilience and data Disaster Recovery as a protection. Safeguarding critical Service -Reduced risk of data Disaster business operations from (DRaaS): Replicate data loss and financial Recovery & disruptions caused by and applications in impact from Business natural disasters, geographically diverse downtime. Continuity cyberattacks, or hardware data centers, ensuring - Cost-effective failures. rapid recovery and compared to building minimal downtime. and maintaining on- premises disaster recovery solutions. Université Constantine 2 © Belguidoum Meriem 9  Why using Cloud Computing? Before the cloud, businesses relied heavily on expensive and complex on-premises infrastructure. Managing servers, storage, software, and security required dedicated IT teams and expertise. This traditional approach came with limitations: High Costs: Significant upfront investments in hardware, software, maintenance, and upgrades. Complexity: Managing intricate infrastructure demanded specialized technical skills. Lack of Flexibility: Difficulty adapting quickly to changing needs and peak demands. Underutilized Resources: Servers often sat idle, wasting energy and resources. Université Constantine 2 © Belguidoum Meriem 10  The benefits of Cloud computing Cloud computing has become a catalyst for innovation and transformation, offering a wide range of benefits for businesses of all sizes: Cost Savings: Reduce capital expenditure on hardware, software, and IT staff, and pay only for resources used. Increased Agility and Flexibility: Quickly scale resources up or down based on demand, adapt to changing market conditions, and launch new products and services faster. Enhanced Collaboration: Enable teams to work together on projects from anywhere with internet access, improving productivity and innovation. Improved Disaster Recovery: Implement robust and cost-effective disaster recovery solutions, minimizing downtime and data loss. Increased Security: Leverage advanced security features offered by cloud providers, often surpassing the security measures of on- premises infrastructure. Université Constantine 2 © Belguidoum Meriem 11  The features of Cloud computing Cloud computing emerged as a solution to these challenges by offering a new paradigm for accessing IT resources: On-demand Self-Service: Access and manage computing resources as needed, without human interaction with the service provider. Broad Network Access: Resources available over the network and accessed through various client devices (laptops, smartphones, tablets). Resource Pooling: Provider's computing resources are pooled to serve multiple clients, with different physical and virtual resources dynamically assigned according to consumer demand. Rapid Elasticity: Resources can be provisioned and released, in some cases automatically, to rapidly scale up or down based on demand. Measured Service: Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer regarding resource consumption. In essence, cloud computing provides a flexible, cost-effective, and accessible alternative to traditional on-premises IT infrastructure. Université Constantine 2 © Belguidoum Meriem 12  Cloud deployment models Four principal types : Public, Private, Hybrid and community Université Constantine 2 © Belguidoum Meriem 13  Public Cloud Public Cloud : is a cloud environment owned and operated by a third-party provider (like AWS, Azure, Google Cloud). Resources (servers, storage) are shared among multiple clients (multi- tenant environment), making it a cost-effective and scalable solution. Like renting an apartment in a large building. You share resources like electricity and plumbing with other tenants, but you have your own private space. Best for: Websites, applications with fluctuating demand, testing and development environments. Université Constantine 2 © Belguidoum Meriem 14  Public Cloud Scenario: A fast-growing social media startup needs a scalable and cost-effective solution to handle unpredictable user growth and traffic spikes. Why Public Cloud? Rapid Scalability: To accommodate millions of users joining the platform. Cost-Effectiveness: Pay only for the resources they use, avoiding large upfront infrastructure investments. Global Reach: Deploy their platform in multiple regions worldwide for low latency and better user experience. Example Solution: Amazon EC2: For flexible virtual servers to handle web traffic and application logic. Amazon S3: For scalable storage of user data (images, videos, etc.). Amazon RDS: For managed databases to store user information and application data. Université Constantine 2 © Belguidoum Meriem 15  Private Cloud Private Cloud: A dedicated cloud infrastructure exclusively for a single organization. It can be hosted on-premises (in the organization's data center) or by a third-party provider. Private clouds offer enhanced security and control over data but can be more expensive than public clouds. like owning your own house. You have complete control over the property and its resources, but you're responsible for maintenance and upkeep. Best for: Organizations with strict security and compliance requirements, sensitive data handling, mission-critical applications. Université Constantine 2 © Belguidoum Meriem 16  Private Cloud Scenario: A large financial institution needs to process and store sensitive customer financial data under strict regulatory compliance requirements (e.g., PCI DSS for payment card information). Why Private Cloud? Enhanced Security & Control: Maintain complete control over their data and infrastructure to meet stringent security and compliance standards. Customization: Tailor their cloud environment to their specific needs and integrate with existing legacy systems. Example Solution: They build a private cloud within their own data centers using: VMware vSphere: To create and manage a virtualized server and storage infrastructure. Dedicated Hardware: Invest in high-performance servers, storage arrays, and networking equipment. Compliance-Specific Security tools Implement firewalls, intrusion detection systems, and data encryption to meet regulatory requirements. Université Constantine 2 © Belguidoum Meriem 17  Hybrid Cloud Hybrid Cloud: A combination of public and private cloud environments, connected securely. This model allows organizations to choose the best deployment model for each workload, balancing cost-effectiveness, security, and control. like having a house with a connection to a nearby apartment building. You can use your own resources or easily access shared amenities when needed Best for: Organizations with varying workloads, legacy applications needing integration, or seasonal demand fluctuations. Université Constantine 2 © Belguidoum Meriem 18  Hybrid Cloud Scenario: A retail company with an existing on-premises data center wants to leverage the agility of the cloud for specific workloads, like website hosting and data analytics, while keeping sensitive customer data securely stored in their private cloud. Why Hybrid Cloud? Flexibility: Choose the best environment (public or private) for each application and workload. Cost Optimization: Use the public cloud for less sensitive, scalable workloads, and reserve the private cloud for critical data and applications. Example Solution: They adopt a hybrid approach: Public Cloud: Utilize Microsoft Azure for: Website Hosting to deploy their e- commerce platform on Azure App Service for scalability and global reach and Data Analytics to Leverage Azure Data Lake and Azure Machine Learning for analyzing customer data and gaining insights. Private Cloud: Maintain their own private cloud for Customer Data and Core Business Applicationsto Host their inventory management system and other critical applications on their private cloud for maximum control and security. Université Constantine 2 © Belguidoum Meriem 19  Community Cloud Community Cloud: A collaborative cloud infrastructure shared by multiple organizations with common interests, such as shared security requirements, compliance regulations, or specialized needs. Resources are pooled and managed jointly by the participating organizations or a third-party provider. like several neighbors sharing a community garden. They each have their own plot, but they share the costs, maintenance, and resources for the collective benefit of the community. Best for: Research institutions collaborating on projects, government agencies sharing data, or industry groups with specific compliance requirements. Université Constantine 2 © Belguidoum Meriem 20  Community Cloud Scenario: A consortium of research universities wants to collaborate on a large- scale genomics research project requiring significant computing power and secure data sharing among researchers across different institutions. Why community Cloud? Resource Sharing: Pool resources (computing power, storage, specialized software) to reduce costs and increase efficiency. Collaboration & Data Sharing: Establish secure data sharing mechanisms and collaborative workflows for researchers across institutions. Specialized Services: use cloud services tailored to research needs, such as high-performance computing clusters and genomic data analysis platforms. Example Solution: They create a community cloud using Shared Infrastructure (Member universities contribute resources to build a shared cloud environment) Federated Identity Management to implement systems that allow researchers from different institutions to securely access shared resources and data using their existing university credentials. Specialized Research Platforms to facilitate data analysis and collaboration. Université Constantine 2 © Belguidoum Meriem 21  Community Cloud Community Cloud: A collaborative cloud infrastructure shared by multiple organizations with common interests, such as shared security requirements, compliance regulations, or specialized needs. Resources are pooled and managed jointly by the participating organizations or a third-party provider. like several neighbors sharing a community garden. They each have their own plot, but they share the costs, maintenance, and resources for the collective benefit of the community. Best for: Research institutions collaborating on projects, government agencies sharing data, or industry groups with specific compliance requirements. Université Constantine 2 © Belguidoum Meriem 22  Cloud deployment models Feature Public Cloud Private Cloud Hybrid Cloud Community Cloud Third-party Shared by provider (like AWS, A single Mix of public and multiple Ownership Azure, Google organization private organizations with Cloud) common interests Shared among Securely Restricted to the Dedicated to one Access many clients connects public community organization (multi-tenant) & private members Most cost- Varies based on Shared costs Cost Most expensive effective usage among members Université Constantine 2 © Belguidoum Meriem 23  Cloud deployment models Less control over Strongest security Flexible security Shared security Security data security and control options responsibilities Scalability Less flexible, Offers a balance depends on Flexibility & Highly flexible, scaling can be of flexibility and community Scalability scales easily more complex control agreement and resources E-commerce Research Banking systems, sites using public institutions Public websites, sensitive cloud for the sharing data, Examples mobile apps, file healthcare data, storefront and government storage government private for agencies applications sensitive collaborating on customer data a project Université Constantine 2 © Belguidoum Meriem 24 Cloud service models Université Constantine 2 © Belguidoum Meriem 25 SaaS SaaS delivers software applications over the internet. Think of it like renting software instead of buying it and installing it on your computer. A SaaS provider hosts and manages everything, from the app itself to the servers and data storage. You just need an internet connection and you're good to go. Examples: Office productivity and collaboration: Google Workspace (Gmail, Docs, Sheets), Microsoft 365 (Word, Excel, Teams) Customer relationship management (CRM): Salesforce, HubSpot Advantages: Easy to use: No complicated installation or software updates to worry about. Accessible: Use it from any device that connects to the internet. Always up-to-date: The provider handles all the software maintenance and security updates automatically. Disadvantages: Limited customization: You might not be able to customize the software as much as you could with a traditional desktop application. Reliance on the provider: Your access to the software and the security of your data depend on the SaaS provider. Université Constantine 2 © Belguidoum Meriem 26 IaaS IaaS provides you with fundamental computing resources like virtual servers, storage, and networking over the internet. You have full control over the operating systems, applications, and data you run on this infrastructure, but you're responsible for managing them. Examples: Virtual servers: Amazon EC2, Microsoft Azure Virtual Machines, Google Compute Engine Storage: Amazon S3 (Simple Storage Service), Azure Blob Storage, Google Cloud Storage Networking: Amazon VPC (Virtual Private Cloud), Azure Virtual Network, Google Cloud VPC Advantages: Cost-effective: Pay only for the resources you use, often on a pay-as-you-go basis. Scalable: Easily adjust your resources up or down to meet changing demands. Flexible: Customize your infrastructure to your specific needs. Disadvantages: Requires IT expertise: You need to manage the operating systems, applications, and data. Potential for security risks: You're responsible for securing your virtual infrastructure. Université Constantine 2 © Belguidoum Meriem 27 PaaS PaaS provides a complete platform for developing, testing, deploying, and managing applications without worrying about the underlying infrastructure. It includes the operating system, middleware, databases, and other essential services. Examples: Application platforms: Heroku, Google App Engine, AWS Elastic Beanstalk Databases: AWS RDS (Relational Database Service), Azure SQL Database, Google Cloud SQL Advantages: Simplified development and deployment: Focus on building and delivering apps faster, without infrastructure management overhead. Enhanced collaboration: Tools that facilitate teamwork among developers. Scalability and reliability: Built-in scalability and high-availability features.. Disadvantages: Less control: Limited control over the underlying infrastructure. Vendor lock-in: Can be challenging to migrate applications to a different PaaS provider. Université Constantine 2 © Belguidoum Meriem 28 Cloud service models Université Constantine 2 © Belguidoum Meriem 29 Cloud service models Université Constantine 2 © Belguidoum Meriem 30  Comparison between Cloud Providers Université Constantine 2 © Belguidoum Meriem 31  Comparison between Cloud Providers Caractéristique AWS Azure GCP Market Second Place, Rapid Third Place, Strong Global Leader Position Growth Growth in AI and ML Launch Date 2006 2010 2008 Expanding rapidly, Geographic Strong presence in the strong presence in Largest global coverage Coverage US and Europe North America and Europe Pay-as-you-go, Pay-as-you-go, Usage Pay-as-you-go, Pricing Models Reserved Instances, Commitment Enterprise Contracts Scalability savings Discounts - Expertise in AI and - Wide range of services - Strong integration with Machine Learning - Maturity and the Microsoft ecosystem - Leading Kubernetes Strengths experience - Robust hybrid services platform (GKE) - Rich ecosystem and - Strong presence in the - Attractive offers for active community public sector startups Université Constantine 2 © Belguidoum Meriem 32  Comparison between Cloud Providers - Less mature than - Fewer services than - Complexity of pricing AWS AWS Weaknesses - Fewer support options - Less extensive - Reliance on the for small businesses geographic Microsoft ecosystem coverage Samsung, eBay, Snapchat, HSBC, Key Clients Netflix, Airbnb, Spotify Twitter PayPal Comparison in terms of: Compute Networking Database Data analytics ML-AI Université Constantine 2 © Belguidoum Meriem 33  Comparison between Cloud Providers : compute Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - ECS (Elastic Container - Google Kubernetes - Azure Kubernetes Service): Container Engine (GKE): Managed Service (AKS): Managed orchestration service, Kubernetes service, Kubernetes service, integrates with Docker, integrates with other integrates with Azure supports EC2 and Fargate GCP services, supports Active Directory, supports (serverless containers). multiple cluster various container EKS (Elastic Kubernetes configurations. runtimes. Service): Managed Cloud Run: Serverless PaaS Azure Container Kubernetes service for platform, container- (Containers & Instances: Run containers deploying, managing, and based, scales Serverless) without managing servers, scaling containerized automatically based on quick deployment, ideal applications. demand. for short-lived tasks. Lambda: Serverless Cloud - Azure compute platform, run Functions: Serverless Functions: Serverless code without managing compute, event-driven compute, event-driven servers, event-driven functions, integrates triggers, integrates with execution, scales with other GCP other Azure services. automatically. services. Université Constantine 2 © Belguidoum Meriem 34  Comparison between Cloud Providers : compute Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - App - Elastic Beanstalk: Easy- Engine: Platform for to-use platform for - App Service: Platform building and deploying and managing for web apps, mobile deploying scalable web applications and back ends, and APIs, web applications and services, supports supports various PaaS APIs, supports various various languages (Java, languages (.NET, Java, (Application languages (Java,.NET, PHP, Python, etc.). Node.js, PHP, Python). Platforms) Python, PHP, Go, Lightsail: Simplified - Static Web Node.js). virtual server Apps: Platform for - Cloud Functions for deployment for web hosting static websites Firebase: Serverless apps, blogs, and small and web applications. backend for mobile businesses. and web applications. Université Constantine 2 © Belguidoum Meriem 35  Comparison between Cloud Providers : compute Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - EC2 (Elastic Compute Cloud): Wide range of - Virtual - Compute instance types (general Machines: Various VM Engine: Customizable purpose, compute- sizes and types, support machine types, IaaS (Virtual optimized, memory- for multiple operating support for various Machines) optimized, etc.), systems (Windows, operating systems, customizable Linux), flexible pricing preemptible instances configurations, pay-as- options. for cost savings. you-go pricing. Université Constantine 2 © Belguidoum Meriem 36  Comparison between Cloud Providers : networking Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - VPC (Virtual Private - Virtual - Virtual Private Cloud Cloud): Create isolated Network: Create isolated (VPC): Create isolated networks, define private networks, private networks, Virtual subnets, control routing, connect to on-premises define subnets, Networks manage security with networks, define control routing, security groups and subnets, manage configure firewall network ACLs. network security. rules. - Elastic Load Balancing: Distribute - Azure Load - Cloud Load incoming traffic across Balancer: Distribute Balancing: Distribute multiple targets (EC2 traffic across VMs, traffic across multiple Load Balancing instances, containers, IP containers, or other backend instances, addresses), supports backend services, supports HTTP(S), various types of load supports various load TCP, and UDP load balancers (Application, balancing algorithms. balancing. Network, Classic). Université Constantine 2 © Belguidoum Meriem 37  Comparison between Cloud Providers : networking Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - Cloud - Route 53: Highly - Azure DNS: Host DNS DNS: Managed DNS available and scalable domains in Azure, service, global DNS DNS service, domain manage DNS records, distribution, high registration, traffic integrate with other availability, routing, health checks. Azure services. programmatic control. - Cloud CDN: Globally - CloudFront: Global distributed CDN, - Azure CDN: Global CDN Content CDN for delivering integrates with for content delivery, Delivery content (websites, Google Cloud Storage integrates with other Networks videos, software) with and other services, Azure services, supports (CDNs) low latency and high optimized for high- multiple pricing tiers. transfer speeds. performance content delivery. Université Constantine 2 © Belguidoum Meriem 38  Comparison between Cloud Providers : networking Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - Direct - Cloud Connect: Establish a Interconnect: Dedicat ExpressRoute: Dedicate dedicated private ed connection from d private connection to connection between your network to Networking Azure. - VPN your data center and Google Cloud. - Connectivity Gateway: Create secure AWS. - VPN: Create Cloud VPN: Secure site-to-site VPN secure connections connections between connections. between your network your network and and AWS. Google Cloud. Université Constantine 2 © Belguidoum Meriem 39  Comparison between Cloud Providers : database Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - Cloud SQL: Managed - RDS (Relational - Azure SQL relational databases Database Service): Database: Managed SQL (MySQL, PostgreSQL, Managed relational Server database, SQL Server), databases (MySQL, intelligent performance automated backups, PostgreSQL, Oracle, SQL tuning, built-in security high availability, easy Server), automated features, flexible integration with other Relational backups, point-in-time deployment options. GCP services. Databases recovery, easy - Azure Database for - Cloud scaling. Aurora: MySQL MySQL/PostgreSQL: Ma Spanner: Globally and PostgreSQL- naged MySQL and distributed, scalable compatible database, PostgreSQL services, database, strong high performance and high availability, consistency, ideal for scalability, self-healing automated backups. mission-critical storage. applications. Université Constantine 2 © Belguidoum Meriem 40  Comparison between Cloud Providers : database Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - Cloud Bigtable: NoSQL database, wide-column - Cosmos DB: Multi- store, high throughput for model NoSQL - DynamoDB: Fully large datasets, low latency database, supports managed NoSQL for read/write NoSQL various APIs database, key-value operations. Databases (document, key- store, low latency, high - Cloud Firestore: NoSQL value, graph), scalability. document database, globally distributed, designed for mobile, web, low latency. and server development, flexible querying. - ElastiCache: In- - Azure Cache for - Memorystore: Managed memory data store Redis: Managed in-memory data store (Redis, Memcached), Redis service, high (Redis, Memcached), low Caching improves performance performance, low latency, high throughput, of web applications by latency, scalable integrates with other GCP caching frequently caching. services. accessed data. Université Constantine 2 © Belguidoum Meriem 41  Comparison between Cloud Providers : data analytics Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - HDInsight: Managed - EMR (Elastic Hadoop, Spark, Kafka, - Dataproc: Managed MapReduce): Managed and HBase clusters, Hadoop and Spark Hadoop, Spark, Hive, integration with other clusters, integrated and Pig clusters for big Azure services. - with other GCP data processing. - Azure Stream services, customizable Kinesis: Real-time data Analytics: Real-time data cluster Data streaming service, stream processing, configurations. Processing & collect, process, and analyze data from IoT - Cloud Analytics analyze streaming devices, applications, Dataflow: Batch and data. - and other sources. - real-time data Athena: Serverless Data Factory: Cloud- processing service, query service for data in based data integration unified programming S3, pay-per-query and transformation model for both batch pricing. service, build ETL and streaming data. pipelines. Université Constantine 2 © Belguidoum Meriem 42  Comparison between Cloud Providers : data analytics Service AWS (Amazon Web GCP (Google Cloud Azure (Microsoft) Category Services) Platform) - Azure Synapse - Redshift: Data - BigQuery: Serverless Analytics: Analytics warehouse service, data warehouse, service that brings Data optimized for large-scale petabyte-scale, fast together data Warehousing data analysis, SQL-based SQL queries, warehousing, big data querying, petabyte- integrated machine analytics, and data scale storage. learning capabilities. integration. - Power BI: Business - Looker: Business - QuickSight: Business intelligence service, intelligence and data Data intelligence service, connect to various data visualization platform, Visualization & create and share sources, create explore and analyze Business interactive dashboards, interactive reports and data, create custom Intelligence data visualizations, and dashboards, embed dashboards, reports. visuals in applications. embedded analytics. Université Constantine 2 © Belguidoum Meriem 43  Comparison between Cloud Providers : ML - Azure Machine - SageMaker: Platform for Learning: Cloud- - AI Platform: Platform for building, training, and deploying based ML service, building, deploying, and Machine ML models, supports various build and deploy managing ML models, Learning frameworks (TensorFlow, models, use pre- supports custom training Platforms PyTorch, MXNet), built-in trained models, code, pre-built algorithms, algorithms. automated and Jupyter notebooks. machine learning. - Rekognition: Image and video - Cloud Vision API: Image - Cognitive analysis service, facial analysis service, object Services: Pre- recognition, object detection, detection, facial trained AI models text extraction. recognition, optical for vision, speech, - Comprehend: Natural character recognition language, and language processing service, (OCR). - Cloud Natural decision making, sentiment analysis, entity Language API: Natural AI Services APIs for easy recognition, topic language processing integration. modeling. Polly: Text-to-speech service, sentiment analysis, Examples: service, create lifelike speech entity recognition, syntax Computer Vision, from text. - Lex: Service for analysis. - Cloud Speech-to- Face API, Speech to building conversational Text & Text-to-Speech Text, Text Analytics, interfaces (chatbots), natural APIs: Convert speech to Translator. language understanding. text and vice versa. Université Constantine 2 © Belguidoum Meriem 44  Cloud computing & IoT The cloud is the engine behind IoT: IoT devices and sensors collect tons of data. The cloud provides the perfect place to: Store all that data safely and securely, no matter how much there is. Analyze the data in real-time to find useful insights. Control and manage IoT devices from anywhere. IoT is a playground for the cloud: The cloud offers tools and platforms that make it easier to build and use IoT solutions: IoT platforms: These platforms help connect, monitor, and manage IoT devices. They also help analyze the data collected from those devices. (Examples: AWS IoT, Google Cloud IoT Core, Azure IoT Hub) Machine learning services: Use these services to find patterns in IoT data, make predictions, and automate actions. (Think: predicting equipment failures before they happen or making processes more efficient). APIs (Application Programming Interfaces): These tools help connect IoT data with other business systems and applications. Université Constantine 2 © Belguidoum Meriem 45  Cloud computing & IoT Cloud computing provides the foundation, tools, and intelligence for the Internet of Things (IoT) to thrive. Here are some real-world examples: Smart Cities: Sensors collect information on traffic, pollution, and street lighting. The cloud gathers and analyzes this data to help cities run more efficiently, making adjustments in real time. Connected Healthcare: Wearable devices (like smartwatches) and medical sensors send health data to the cloud. This allows for personalized healthcare, early detection of potential issues, and proactive interventions. Industry 4.0: Sensors monitor manufacturing equipment in real-time. The cloud analyzes this data to spot problems, optimize production, and predict maintenance needs, preventing costly downtime. The combination of cloud computing and the Internet of Things is sparking a new wave of technological innovation, changing how industries operate and improving our lives. Université Constantine 2 © Belguidoum Meriem 46  Cloud computing & IoT Cloud computing provides the foundation, tools, and intelligence for the Internet of Things (IoT) to thrive. Here are some real-world examples: Smart Cities: Sensors collect information on traffic, pollution, and street lighting. The cloud gathers and analyzes this data to help cities run more efficiently, making adjustments in real time. Connected Healthcare: Wearable devices (like smartwatches) and medical sensors send health data to the cloud. This allows for personalized healthcare, early detection of potential issues, and proactive interventions. Industry 4.0: Sensors monitor manufacturing equipment in real-time. The cloud analyzes this data to spot problems, optimize production, and predict maintenance needs, preventing costly downtime. The combination of cloud computing and the Internet of Things is sparking a new wave of technological innovation, changing how industries operate and improving our lives. Université Constantine 2 © Belguidoum Meriem 47  AWS Certifications path 2024 & jobs Université Constantine 2 © Belguidoum Meriem 48  AWS Certifications path 2024 & jobs Université Constantine 2 © Belguidoum Meriem 49  Quizz Université Constantine 2 © Belguidoum Meriem 50  References AWS documentation whitepapers: https://docs.aws.amazon.com/whitepapers/latest/aws-overview/what-is- cloud-computing.html https://www.statista.com/chart/18819/worldwide-market-share-of- leading-cloud-infrastructure-service-providers/ T. Velte, A. Velte, and R. Elsenpeter, Cloud Computing: A Practical Approach. New York, NY: McGraw-Hill, 2010. Université Constantine 2 © Belguidoum Meriem 51