Architecture Microservices et Scalabilité

Microservices & Cloud Computing - Course Notes

• Course: Microservices & Cloud Computing
• Dates: 20/11/2024 - 21/11/2024
• Institution: ESIEA - 4a
• Presenters: Kevin Navarri, Thomas Desrut, Julien Trouillet
• Topic Overview: The course covers various cloud concepts, including microservices architecture, essential patterns, security, deployment, maintenance, logging, monitoring, and cloud computing introductions.

Agenda

• Presentation 1: Introduction to Information Systems (SI)
• Presentation 2: Microservices & Cloud Computing (this session)
• Presentation 3: Cloud Architectures
• Presentation 4: Cloud Scalability

Agenda - Microservices & Cloud Computing

• About Us: Introduction to the presenters and Capgemini
• Introduction to Microservices: (Details of this section)
• Essential Microservices Patterns: (Details of this section)
• Securing Microservices: (Details of this section)
• Deploying and Maintaining Microservices: (Details of this section)
• Logging: (Details of this section)
• Monitoring: (Details of this section)
• Introduction to Cloud Computing: (Details of this section)

Monoliths and Their Limitations

• Definition: A monolithic application combines all modules into one program.
• Benefits: Easy deployment, managed complexity, centralized testing.
• Limitations: Costly scalability, difficulty in updating specific components, inability to scale individual parts of the application, and more.

Definition and Properties of a Microservice

• Definition: A microservice implements distinct features or functionality as a small, self-contained application.
• Key Properties:
  • Modularity: Each microservice represents a specific business function.
  • Autonomy: Independent development and deployment of each service.
  • Data Decentralization: Each service can have its own database for data isolation.

Benefits of Microservices

• Scalability: Only the necessary services are scaled, reducing resource waste.
• Development Speed: Independent services enable faster deployment cycles.
• Optimization/Compatibility: Allows use of different languages/technologies.
• Fault Isolation: Failures in one service do not affect the entire application.

"Decentralized" Architecture: Governance & Catalog

• Shift: From centralized to decentralized governance in microservices.
• Monoliths: Centralized governance for a single application.
• Micro Services: Governance per service, allowing each team to have its own schedule.
• Service Catalog: A repository of available services.
• Pitfalls: Increased complexity and issues with distributed transactions for rollback.

Notable Example: Netflix with AWS

• Challenges addressed: Scalability, data management in a massive distributed system, and resolving memory leaks due to the complexity of the codebase.
• Components: Explanation of (Zuul - Proxy Layer, Platform zone, API).

Business-Driven Orientation: What is DDD

• Domain-Driven Development (DDD): Aligning software design with the core business domain, ensuring that business needs are incorporated during development.
• Ubiquitous Language: Shared, precise language between developers and domain experts.
• Bounded Contexts: Dividing complex systems into sub-domains to avoid complexity issues.

Circuit Breaker

• Definition: A design pattern used to prevent cascading failures in a system.
• Mechanism: When a service fails repeatedly, the circuit "opens" and stops requests to that service.
• Implementation: Commonly implemented with tools such as Hystrix, Resilience4j.

Retry Pattern

• Strategies:
  • Fixed Delay: Using a constant delay between retry attempts.
  • Incremental Delay: Increasing the delay with each retry attempt.
  • Exponential Backoff: Exponentially increasing the delay for each attempt.

Event Broker Pattern

• Intermediary: Event Broker manages events between producers and consumers.
• Key Features:
  • Event Routing: Delivering events to appropriate consumers.
  • Scalability: Handling events in parallel efficiently.
  • Loose Coupling: Independent components that can be added, removed or changed without harming the system.

Authentication vs. Authorization

• Authentication: Verifying user identity (e.g., password, token).
• Authorization: Determining access permissions for an authenticated user.

SSO and JWT Patterns

• Single Sign-On (SSO): Access multiple applications with a single login.
• OAuth 2.0: Authorization protocol for granting limited access.
• OpenID Connect (OIDC): Identity layer on top of OAuth, adds user identity and profile information and security.

OAuth Pattern

• Flow: Explanation for how third-party applications can access user data on other platforms, without requiring users to provide credentials.

OIDC Pattern

• Flow: Detailed steps and components involved in the OIDC flow for authentication and authorization.

Important Elements in CI/CD and DevOps

• CI/CD (Continuous Integration/Continuous Delivery): Automates software integration, building, testing, and deployment processes.
• DevOps: A set of practices that aims at integrating software development and operations teams to improve efficiency and collaboration.
• Importance of tooling: The critical role of using DevOps tools during this process.

Simple CI/CD for Java Projects

• Tools used: Illustration of common tools used in a simple CI/CD pipeline setup, explaining their roles and purposes during development.

Complex CI/CD for Bank Projects

• Challenges: Explanation of the complexities in implementing a CI/CD pipeline for a bank project, including the different environments (development, testing, production).
• Tools: Illustration of a more complex implementation of a CI/CD pipeline, including many of the services required.

Centralized Logging Tools (ELK Stack)

• Component Breakdown: Explanation of Elasticsearch, Logstash, and Kibana, and their roles in centralized logging.
• Use Case: How these tools help aggregate logs from multiple services.

Importance of Logs for Diagnostics and Issue Resolution

• Benefits: Importance of logs as diagnostic tool for investigating issues and tracing events.
• Example Logs: Demonstration examples of logs and their uses to troubleshoot problems in a software application.

Introduction to Monitoring Tools

• Tools: Explanation of Grafana, Prometheus, and Splunk, their specializations in system monitoring and visualization.
• Use Case: How these tools help visualize data and identify potential issues or problems.

Importance of Observability: Metrics, Alerts, Performance Tracking

• Metrics: Collecting data (like CPU usage) in order to understand how a system behaves in production.
• Alerts: Defining thresholds or parameters that need to be sent to a corresponding team when a problem is detected.
• Performance Tracking: Monitoring to make sure the system is functioning and performing as expected.

I Heard About Public and Private Cloud, IaaS, PaaS & SaaS

• Different Cloud Types: Public, Private, Hybrid explanation of these types.
• IaaS: Infrastructure-as-a-service.
• PaaS: Platform-as-a-service.
• SaaS: Software-as-a-service.
• Cloud Service Provider Examples: Explanation and illustration of major cloud service provider examples like AWS, Azure and Google Cloud.

Where is the Cloud?

• Global Infrastructure: Explanation of how cloud providers spread their services around the world in different Regions.
• Availability Zones: Explanation of how multiple geographically separated data centers give redundancy and low latency.

A Cloud Service Is...

• Key characteristics:
  • On-demand IT services
  • Automated deployment
  • Elasticity of resources
  • Pay-per-use

Differences Between Non-Cloud and Cloud Projects

• Project Management: Explanations on handling "pet" projects versus "cattle" projects in cloud environments.

Liabilities (Clouder vs. Client)

• Shared Responsibility: Distinction of responsibilities from the service provider and the customer when using the cloud.

Additional Information

• Contact Information: List of contact info for the instructors.