/commute-planner

Primary LanguageC#

Class project

To run with Docker:

  1. docker-compose up -d
  2. visit http://localhost:5015

To run using Aspire, also requires docker, and .net 9.0 (all platforms):

  1. % dotnet workload restore (only once)
  2. % dotnet build
  3. % dotnet test
  4. % dotnet run --project Applications/commute-planner.AppHost

CI/CD deployed to:

https://colorado.heyjosh.io

(https://github.com/joshwyant/commute-planner/actions/workflows/azure-dev.yml)

Apps:

  • Orchestrated through Docker, Aspire (AppHost), or Azure
  • Web App: Home page - select a popular route, see a description and whether transit or driving is faster; Routes - see the featured routes
  • DataFetcher app collects data from https://511.org and Google Maps REST APIs, upon a freshness request message.
  • DataProcessor app periodically sends message to DataFetcher app, which then sends message with the fresh data recieved. DataProcessor analyzes, combines, and persists the transformed data to be efficiently retrieved.
  • ApiService provides a REST API for the Web front-end to consume, acting as an interface to the database.
  • AppHost - handles Aspire orchestration and deployment, as well as a dashboard, logs, and graphs. Provides interfaces to Azure Insights and OpenTelemetry for production monitoring and instrumenting used by logging throughout.
  • RabbitMQ is deployed and connected by the Aspire library as a container for messaging.
  • PostgreSQL is deployed and connected by the Aspire library as a container for the data store, which is an ACID database, since the data is small, structured, and transactional.

Components:

  • ApiClient - used by the Web app as a model and interface to the REST API.
  • DataCollection - the logic for how to fetch new third party transportation data, along with communication with the data processor app.
  • DataProcessing - the logic for how to analyze, combine, and store the transformed data, as well as for signaling to the DataFetcher app.
  • EventCollaboration - Tools and extensions used to handle the bulk of the message queueing work and interfaces with RabbitMQ.
  • MapsApiClient - An implementation of the interface to the Routes REST API that Google Maps exposes.
  • ServiceDefaults - Configures the orchestration of the distributed app using Aspire.
  • TransitApiClient - An implementation of the interface to the part of the 511 Transit REST API that is exposed and used by the app.
  • TransitApiModels - Models shared by the data fetcher and transit API.

Tests:

  • Integrated into the build and deploy process. Uses Nunit for Unit tests, Moq for mocks and spies, and Aspire to handle the orchestration for integration testing
  • Fake Google Maps server
  • Fake Transit API server
  • Separate folders

Database

  • Uses Entity Framework for the Object Relational Mapping, Aspire for orchestration, and PostgreSQL for persistence.

Continuous integration/Continuous Delivery

  • On each commit to main, triggers GitHub Actions (.github/) to prepare .NET environment, install build tools, build, test, and deploy to Azure as a cohesive Web App composed of containerized apps and services. Uses Azure Secrets to store real API keys, production passwords, etc.