High-Level Goal

Design and implement a self-contained microservice exposing an HTTP API that allows clients to enqueue tasks for asynchronous processing and retrieve real-time processing metrics. Each task must append a message to a shared log file, with proper synchronization for concurrent access. Implementation details (language, frameworks, storage, and queue type) are left to your discretion.

Context

Your job is to implement the core business logic of a new microservice with two endpoints (POST /tasks and GET /statistics) —createTask and getStatistics. A task represents a job to be performed by an external asynchronous worker. For our use case, the client sends a new task object with the format {“message”: string}. A worker should receive tasks and execute them in the following manner:

• Append a log entry to a shared file (with the timestamp, the worker ID, the task ID, and the message), handling concurrent writes safely.
• Simulate processing time by sleeping for a configured duration.
• Simulate random failures by a configured failure rate.
• Trigger retries with a configured delay on failures. Workers should be created upon need with the following caveats:
• The maximum number of workers is the number of CPU cores.
• Workers remain alive and ready to process new tasks for a configurable duration. After this idle time, they exit.

HTTP Endpoints

POST /tasks

• Accepts a JSON payload in the format {“message”: string}.
• Add a new task to the processing pipeline.
• Returns the new task ID

GET /statistics

• Returns current processing metrics in JSON:
• Number of tasks processed.
• Number of task retries.
• Number succeeded vs. number failed.
• Average processing time per attempt.
• Current queue length.
• Idle workers count.
• Hot (In use) workers count.

Environment Variables

Configure the service via environment variables as defined in the existing README:

• SERVER_PORT — HTTP server port
• TASK_SIMULATED_DURATION — Simulated duration (ms) per task attempt
• TASK_SIMULATED_ERROR_PERCENTAGE - Simulated task error percentage
• TASK_ERROR_RETRY_DELAY — delay (ms) between retries
• WORKER_TIMEOUT — the time the worker needs to be idle before it is cleaned up
• TASK_MAX_RETRIES — maximum retry attempts per task

Operate instruction

Build & Run (local):

npm install
npm run build
npm start

Dev (live reload):

npm install
npm run dev

Docker:

• Dockerfile and docker-compose.yml included.

docker compose build
docker compose run

Testing:

• Jest + Supertest tests included. Run with npm test.