sr-world-cup-scoreboard

SR code challenge - imarrero

Goals

Create a c# library to manage in memory Football World Cup matches data

Some actions must be implemented:

1. Start a game. Our data partners will send us data for the games when they start, and these should capture (Initial score is 0 – 0).
2. Finish game. It will remove a match from the scoreboard.
3. Update score. Receiving the pair score; home team score and away team score updates a game score.
4. Get a summary of games by total score.

Detailed information is provided by PDF document

How to run the project

1. Clone repository
2. Open sr-world-cup-scoreboard.sln
3. Build Solution
4. Execute Units tests: More Info about Unit Tests
5. Execute BDD tests: More Info about Behaviour tests
6. Run all tests, debug if required.

Project management

Tickets created in Trello:

• https://trello.com/invite/b/wz4z2Hrw/b0f966ada9b722016cf8d60688a60a6e/tickets

• All commits are related to each ticket.

• As this is a test, all commits are in main branch. The correct way to go in a production environment should be more like:

• • Create a development and release branch
• • Code each ticket on a feature branch (from dev)
• • After local tests passed -> create a pull request to merge in development
• • If team CR approves the PR, then will be merged into development This way, on each "release date" a new release branch can be create with all dev code. Then, depending on the company policy, (new test are done / staging etc..) finally release is merged into MASTER or MAIN and deployed to production.

Future improvements

1. Infrastructure: Separate external parameters from internal use:

• Create a Facade/API/Gateway
• Set parameters validation in a centralized class (FluentValidation / DaataAnnotations.. )
• Use a model as a entry parameter

2.- Create performance tests

• Evaluate current Concurrent dictionary vs Performance (maybe using _locks in the critical methods is enough)
• Make actions Async to really use Concurrent dictionary features
• • • No different consumers should touch the same MATCH at the same time: Evaluate

3. Persist data

4. Security

• Sanity text parameters
• Configuration for max dictionary items ( to avoid overflow of matches )

5. Implement Logging:

• System logs
• Business data logs: Each CRUD operation should be saved, in order to get a complete trace of what happened in each match.
• Send to Elastic or similar
• Create a Kibana or similar monitoring website: CONNECT TO SLACK or similar!!

5. Create a WebAPI / Microservice to consume this library

• • User Swagger for documentation / secure access
• • Improve Error handling: Error codes centralized + Friendly Descriptions + Language resolution according request
• • • (Bonus) Create an API gateway to separate public endpoints from internal endpoints
• • • (Bonus) Containerize and deploy on an orchestrator
• • • (Bonus) Monitorize Logs with Kibana, scale pods as necessary, communicate events with other microservices in the cluster/cloud using rabbit/kafka.

Sorry for the overthinking. Just brainstorming surely Im missing important things