Uber-microservices Application

You open Uber. You request a ride. In 10 seconds a driver is assigned. But how? 500 drivers are moving around your city RIGHT NOW. How does Uber find the nearest one instantly? Why does normal SQL fail at this scale?

This project demonstrates the solution using a microservices architecture with Redis Geospatial for efficient location queries, Kafka for event-driven communication, and Spring Boot for building scalable, distributed services.

---

Services Overview

Service	Port	Responsibility
location-service	8082	Tracks real-time driver locations via Redis Geospatial
ride-service	8083	Manages ride lifecycle, publishes events to Kafka
matching-service	8084	Consumes ride events, finds & assigns best driver

---

Architecture Flow

Driver Phone → Location Service → Redis (GEOADD)

Rider App → Ride Service → Kafka (ride.requested)
                                      ↓
                           Matching Service (consumer)
                                      ↓
                           Location Service (find nearby drivers)
                                      ↓
                           Matching Algorithm (score drivers)
                                      ↓
                           Kafka (ride.matched)
                                      ↓
                           Ride Service (update ride with driver)

---

How To Run

Step 1: Start Infrastructure

docker-compose up -d

This starts Redis, MySQL, Zookeeper, and Kafka.

Wait 30 seconds for Kafka to fully start before running services.

Step 2: Start Location Service

cd location-service
mvn spring-boot:run

Step 3: Start Ride Service

cd ride-service
mvn spring-boot:run

Step 4: Start Matching Service

cd matching-service
mvn spring-boot:run

---

Testing End-to-End Flow

Step 1: Add Driver Locations (Location Service)

POST http://localhost:8082/api/v1/locations/drivers/update
{
    "driverId": "driver:1",
    "latitude": 12.9716,
    "longitude": 77.5946
}

POST http://localhost:8082/api/v1/locations/drivers/update
{
    "driverId": "driver:2",
    "latitude": 12.9800,
    "longitude": 77.5800
}

POST http://localhost:8082/api/v1/locations/drivers/update
{
    "driverId": "driver:3",
    "latitude": 12.9600,
    "longitude": 77.6100
}

Step 2: Request a Ride (Ride Service)

POST http://localhost:8083/api/v1/rides/request
{
    "riderId": "rider:1",
    "pickupLatitude": 12.9716,
    "pickupLongitude": 77.5946,
    "pickupAddress": "MG Road, Bangalore",
    "dropLatitude": 12.9352,
    "dropLongitude": 77.6245,
    "dropAddress": "Koramangala, Bangalore"
}

Step 3: Check Ride Status

GET http://localhost:8083/api/v1/rides/{rideId}

You will see driverId assigned and status = ACCEPTED

Step 4: Start the Ride

PUT http://localhost:8083/api/v1/rides/{rideId}/start

Step 5: Complete the Ride

PUT http://localhost:8083/api/v1/rides/{rideId}/complete

Step 6: Get Rider History

GET http://localhost:8083/api/v1/rides/rider/rider:1

---

Verify in Redis CLI

docker exec -it redis-geo redis-cli

# See all stored drivers
ZRANGE drivers:locations 0 -1

# Check specific driver position
GEOPOS drivers:locations "driver:1"

# Distance between two drivers
GEODIST drivers:locations "driver:1" "driver:2" km

---

Key Concepts Covered

• Redis Geospatial (GEOADD, GEORADIUS)
• Kafka event-driven architecture (Producer/Consumer)
• Ride state machine (REQUESTED → MATCHING → ACCEPTED → STARTED → COMPLETED)
• Driver scoring algorithm (distance + rating weighted score)
• Service-to-service REST communication (Matching → Location Service)
• Docker Compose for infrastructure setup

Made with Love