Store Monitoring System

A FastAPI application that generates uptime/downtime reports for restaurant stores based on polling data.

Requirements

• Docker and Docker Compose
• 8GB+ RAM (for processing large datasets)

Quick Start

1. Start the services:

docker-compose up -d

2. Load the data:

docker-compose exec app python scripts/load_data.py

3. Generate a report:

curl -X POST http://localhost:8000/api/v1/trigger_report

4. Check report status (use the report_id from step 3):

curl http://localhost:8000/api/v1/get_report/{report_id}

API Endpoints

• POST /api/v1/trigger_report - Start report generation
• GET /api/v1/get_report/{report_id} - Get report status or download CSV
• GET /api/v1/stats - System statistics
• GET /api/v1/health - Health check
• GET /docs - API documentation

Output Format

The system generates CSV reports with the following columns:

• store_id
• uptime_last_hour (minutes)
• uptime_last_day (hours)
• uptime_last_week (hours)
• downtime_last_hour (minutes)
• downtime_last_day (hours)
• downtime_last_week (hours)

Features

• Business Hours Filtering: Only counts uptime/downtime during store operating hours
• Timezone Support: Handles stores across different timezones with DST awareness
• Smart Interpolation: Extrapolates uptime/downtime from sparse polling data
• Background Processing: Report generation runs asynchronously via Celery
• Caching: Redis caching for improved performance on large datasets
• Batch Processing: Efficiently handles thousands of stores without memory issues

How It Works

The system:

1. Downloads store polling data, business hours, and timezone information
2. Calculates uptime/downtime using business hours filtering
3. Interpolates status between sparse polling observations
4. Generates reports with metrics for last hour, day, and week

Data Sources

The application automatically downloads the official dataset from: https://storage.googleapis.com/hiring-problem-statements/store-monitoring-data.zip

Testing

Check if everything is working:

# Health check
curl http://localhost:8000/health

# System stats
curl http://localhost:8000/api/v1/stats

# Generate and download a report
REPORT_ID=$(curl -s -X POST http://localhost:8000/api/v1/trigger_report | jq -r .report_id)
curl http://localhost:8000/api/v1/get_report/$REPORT_ID

Services

• API: FastAPI server on port 8000
• Database: PostgreSQL for data storage
• Cache: Redis for performance
• Workers: Celery for background report generation

Stopping

docker-compose down