AzureKnowledgeHub

AzureKnowledgeHub is a full-stack learning-resource platform for Azure certification study. It was developed as a degree thesis in the Cloud Developer Azure programme at JENSEN YH.

The project demonstrates how a React frontend, an ASP.NET Core Web API, Entity Framework Core, and Azure services can be combined into a maintainable cloud-based application.

Features

• Learner registration and login
• JWT-based authentication
• Role-based access for Learners and Administrators
• Browse learning resources
• Search and filter by keyword, certification path, difficulty, category, and tag
• View detailed learning-resource content
• View and update a user profile
• Admin-only resource creation, editing, and deletion
• Admin user-management endpoint
• Swagger/OpenAPI documentation
• Azure SQL persistence through Entity Framework Core
• Database migrations and demo data seeding
• Unit tests for password, resource, and search services

Architecture

The backend uses a layered, Clean-Architecture-inspired structure:

API/Controllers
        ↓
Application/Services, Interfaces, DTOs
        ↓
Domain/Entities
        ↓
Infrastructure/Repositories, EF Core, DatabaseSeeder
        ↓
Azure SQL Database

The frontend and backend are separate applications. The React client communicates with the ASP.NET Core API over HTTP, while only the backend API accesses the database.

Technology Stack

Backend

• C#
• ASP.NET Core Web API (.NET 9)
• Entity Framework Core
• SQL Server and Azure SQL Database
• JWT Bearer authentication
• Swagger/OpenAPI

Frontend

• React
• Vite
• JavaScript
• CSS

Testing and QA

• xUnit
• Moq
• FluentAssertions
• Swagger/OpenAPI
• Postman

Delivery and deployment

• GitHub Actions
• Azure App Service
• Azure Static Web Apps

Screenshots

Frontend

Search and filtering

Administrator interface

API documentation with Swagger

API testing with Postman

Unit tests

Azure Deployment

During the thesis project, the application was deployed using:

• Azure Static Web Apps for the React frontend
• Azure App Service for the ASP.NET Core API
• Azure SQL Database for persistent storage
• GitHub Actions for frontend CI/CD

The original school Azure environment is no longer accessible. The screenshots document the thesis deployment, while a new Azure subscription and configuration are required for a new deployment.

Local Setup

Requirements

• .NET 9 SDK
• Node.js and npm
• SQL Server LocalDB or SQL Server
• Visual Studio 2022 or another .NET-compatible IDE

Configure the backend

The development configuration uses SQL Server LocalDB by default. Configure local secrets outside source control:

cd AzureKnowledgeHub
dotnet user-secrets init
dotnet user-secrets set "JwtSettings:SecretKey" "replace-with-a-long-local-development-secret"
dotnet user-secrets set "SeedAdmin:Enabled" "false"

Demo administrator seeding is disabled by default. If it is explicitly required for local testing, configure a local-only password:

dotnet user-secrets set "SeedAdmin:Enabled" "true"
dotnet user-secrets set "SeedAdmin:Password" "replace-with-a-local-demo-password"

Never commit these values to GitHub.

Start the API

From the repository root:

dotnet run --project .\AzureKnowledgeHub\AzureKnowledgeHub.csproj --launch-profile http

The API is available at:

http://localhost:5029

Swagger is available at:

http://localhost:5029/swagger

Start the frontend

In a second terminal:

cd frontend\azureknowledgehub-frontend
npm install
npm run dev

The frontend uses VITE_API_BASE_URL when configured. See .env.example for the local development example.

API Endpoints

Authentication

POST /api/auth/register
POST /api/auth/login

Learning resources

GET    /api/learningresources
GET    /api/learningresources/{id}
POST   /api/learningresources       # Admin only
PUT    /api/learningresources/{id}  # Admin only
DELETE /api/learningresources/{id}  # Admin only

Search

GET /api/search/resources

Profile and administration

GET  /api/profile
PUT  /api/profile
GET  /api/admin/users              # Admin only
POST /api/admin/users/admin        # Admin only

Testing and QA

The project uses a combination of automated unit testing and manual API testing.

Automated unit testing

The backend includes unit tests written with:

• xUnit
• Moq
• FluentAssertions

The automated tests cover:

• Password hashing and password verification
• Learning-resource creation and validation
• Learning-resource updates and deletion
• Search and filtering behavior
• Pagination and invalid search arguments
• Service-layer and repository interactions

Run the tests from the repository root:

dotnet test

Latest local test result:

10 passed
0 failed
0 skipped

API testing

The REST API was manually tested using Swagger UI and Postman. Test activities included:

• Authentication requests
• Learning-resource retrieval
• Search and filtering
• Protected admin endpoints
• HTTP status-code validation
• JSON response validation

Role-based access testing

The application has separate Learner and Administrator access paths.

Learners can:

• Register and log in
• View their profile
• Browse learning resources
• Search and filter resources

Administrators can additionally:

• Create learning resources
• Update learning resources
• Delete learning resources
• Access protected administrator endpoints

Test evidence

Testing limitations

The current automated test suite focuses mainly on the service layer. Future QA improvements could include:

• Controller tests
• Integration tests with a test database
• End-to-end frontend tests
• Automated authorization tests
• Continuous test execution in GitHub Actions
• Performance testing for larger search datasets

Security Notes

• Secrets and connection strings must be provided through local user secrets or deployment configuration.
• Demo administrator seeding is disabled by default.
• JWT signing keys must not be committed to source control.
• Publish profiles and Azure deployment metadata should remain private.
• JWT tokens are stored in browser localStorage in this prototype; a production version could use more restrictive cookie-based session handling.

Known Limitations and Future Improvements

• Search currently uses EF Core LIKE queries.
• Search does not yet provide semantic matching, ranking, or spell correction.
• Test coverage focuses mainly on the service layer.
• Controller and end-to-end test coverage could be expanded.
• Application Insights could improve production observability.
• Azure Key Vault could provide centralized secret management.
• Azure AI Search or vector search could improve large-scale and semantic search.
• Infrastructure as Code could make Azure deployment reproducible.

Academic Context

This application was developed and evaluated as a degree thesis. The thesis investigated cloud architecture, search functionality, scalability, maintainability, authentication, testing, and Azure deployment.

The project received the grade Väl Godkänt (VG).

Author

Created by RMSSanali.

License

This project was created for educational and portfolio purposes.