How patrykslomka/ada2025 works
Project Type and Purpose
• The repository is not a single monolithic system but rather a series of labs (lab1, lab2, lab3, lab4, lab6, lab7) exploring different aspects of modern cloud-native and data-driven architectures.
• The labs span topics like deploying ML prediction and training APIs (lab1 and lab3), experimenting with Kubernetes manifests for containerized deployments (lab1, lab2, lab4), microservices design with choreography (lab6), and API gateway setups (lab7).
Analyzing the File Structure
• Notice that many subdirectories are organized per lab. Each lab contains its own components – for example, lab1 has both “prediction-api” and “training-api” (with variants that use file system persistence, as seen in directories ending in “-fs”), along with a UI component.
• Labs include deployment scripts (Dockerfiles, docker-compose.yml files, Kubernetes YAML manifests, and cloud build configurations) that indicate use of containerization, orchestration, and CI/CD pipelines.
• Several labs make use of API gateway examples and microservice communication patterns—especially lab7 (API gateway examples using Krakend) and lab6 (choreography-based microservices).
Recommended Components
a. Core Components and Layers:
• Lab1: Prediction API, Training API, and Prediction UI.
• Lab3: Cloud functions (gcfun1, gcfun2, gcfun3) and ML pipeline components (data ingestor, mlp_trainer) along with an embedded prediction UI.
• Lab4: Multiple delivery services (standard, with openapi, FaaS variants), gRPC examples, and a placeservice.
• Lab6: Choreography-based microservices such as inventory, notification, order record, and user services.
• Lab7: API gateway examples and load test clients.
• Deployment/Orchestration: Include a layer that features Docker, docker-compose, Kubernetes orchestrators (as seen in the manifests directories and docker-compose files), and cloud build pipelines for CI/CD.
• Data and Integration: Identify databases, model storage (e.g., NFS volumes in lab1), message brokers/pubsub components (given the “pub_sub_util.py” script in several labs), and external integrations indicated by openapi specs.
b. Relationships & Interactions:
• For labs exploring choreography (lab6), illustrate message passing or event-driven sequences between services (inventory, order, notification, etc.).
Architectural Patterns & Design Principles
• Containerization via Docker: show that almost every microservice includes its own Dockerfile.
• Demonstrate orchestration: group related services under Kubernetes clusters or docker-compose constructs when applicable.
• Separation of concerns: differentiate between frontend components (e.g., prediction-ui) and backend processing/API components.
• CI/CD and cloud integration: include cloud build configurations for automated deployments.