Uber architecture overview in Microservices - Architecture Diagram

System Overview - Uber architecture overview

Uber is a ride-hailing platform that connects riders with drivers in real time. It requires handling high volumes of location data, trip requests, and payments with low latency and high availability.

Key requirements include real-time matching, scalable microservices, fault tolerance, and data consistency across services.

Architecture Diagram

User
  |
  v
Load Balancer
  |
  v
API Gateway
  |
  +-----------------------------+
  |                             |
  v                             v
Matching Service           Trip Management Service
  |                             |
  v                             v
Driver Location Service     Payment Service
  |                             |
  v                             v
Cache (Redis)               Database (PostgreSQL)
  |
  v
Database (Cassandra)

Components

User

client

Riders and drivers using the Uber app

Load Balancer

load_balancer

Distributes incoming requests evenly to API Gateway instances

API Gateway

api_gateway

Routes requests to appropriate microservices and handles authentication

Matching Service

service

Matches riders with nearby drivers in real time

Trip Management Service

service

Manages trip lifecycle including start, progress, and completion

Driver Location Service

service

Tracks and updates driver locations continuously

Payment Service

service

Handles fare calculation, payment processing, and receipts

Cache (Redis)

cache

Stores frequently accessed data like driver locations for fast retrieval

Database (Cassandra)

database

Stores large scale, distributed location and trip data

Database (PostgreSQL)

database

Stores transactional data such as payments and user profiles

Request Flow - 9 Hops

User → Load Balancer

Load Balancer → API Gateway

API Gateway → Matching Service

Matching Service → Cache (Redis)

Cache (Redis) → Matching Service

Matching Service → API Gateway

API Gateway → User

Trip Management Service → Database (Cassandra)

Payment Service → Database (PostgreSQL)

Failure Scenario

Component Fails:Cache (Redis)

Impact:Driver location queries slow down as Matching Service must query database directly, increasing latency

Mitigation:Fallback to database queries with optimized indexes; add cache replication and failover to reduce downtime

Architecture Quiz - 3 Questions

Which component is responsible for routing user requests to the correct microservice?

A API Gateway

B Load Balancer

C Matching Service

D Cache

Design Principle

This architecture uses microservices to separate concerns like matching, trip management, and payments. It employs caching to reduce latency for frequent queries and uses different databases optimized for specific data types. Load balancers and API gateways ensure scalability and secure routing.

Practice

1. What is the main reason Uber uses microservices in its architecture?

easy

A. To reduce the number of servers needed

B. To store all data in a single database for simplicity

C. To avoid using APIs for communication

D. To separate different tasks into independent services for better scalability

Solution

1. Step 1: Understand microservices purpose

   Microservices break a large system into smaller, independent parts to handle specific tasks.

2. Step 2: Relate to Uber's needs

   Uber needs to handle many users and real-time updates, so separating tasks helps scale and manage complexity.

3. Final Answer:

   To separate different tasks into independent services for better scalability -> Option D

4. Quick Check:

   Microservices = Independent scalable services [OK]

Hint: Microservices split tasks for easy scaling and management [OK]

Common Mistakes:

• Thinking microservices mean one big database
• Assuming no APIs are used
• Believing microservices reduce servers directly

2. Which of the following is a correct way Uber's microservices communicate?

easy

A. Using APIs and message queues

B. Direct database queries between services

C. Sharing memory space directly

D. Using FTP to transfer data files

Solution

1. Step 1: Identify communication methods in microservices

   Microservices communicate via APIs (for requests) and message queues (for async events).

2. Step 2: Match with Uber's architecture

   Uber uses APIs and message queues to enable services to talk without tight coupling.

3. Final Answer:

   Using APIs and message queues -> Option A

4. Quick Check:

   Communication = APIs + message queues [OK]

Hint: Microservices talk via APIs and message queues [OK]

Common Mistakes:

• Thinking services query each other's databases
• Assuming shared memory is used
• Believing FTP is used for service communication

3. Consider Uber's ride request flow: User app sends request -> Dispatch service -> Driver service -> Notification service. Which service likely handles real-time driver location updates?

medium

A. Driver service

B. Dispatch service

C. Notification service

D. User app

Solution

1. Step 1: Understand each service role

   User app sends requests, Dispatch matches rides, Driver service manages driver data, Notification sends alerts.

2. Step 2: Identify who tracks driver location

   Driver service manages driver info including real-time location updates.

3. Final Answer:

   Driver service -> Option A

4. Quick Check:

   Driver location updates = Driver service [OK]

Hint: Driver service manages driver data and location [OK]

Common Mistakes:

• Confusing Dispatch with driver location tracking
• Thinking Notification service tracks location
• Assuming User app handles driver location

4. If Uber's Notification service fails to send ride updates, what is the best way to fix it without affecting other services?

medium

A. Restart the entire system including all microservices

B. Fix and restart only the Notification service

C. Merge Notification service with Dispatch service

D. Stop all services to prevent errors

Solution

1. Step 1: Understand microservices isolation

   Each microservice runs independently, so fixing one doesn't require restarting all.

2. Step 2: Apply best practice for failure

   Fix and restart only the failing Notification service to avoid downtime elsewhere.

3. Final Answer:

   Fix and restart only the Notification service -> Option B

4. Quick Check:

   Isolated fixes = Restart single service [OK]

Hint: Fix only the failing microservice to avoid system downtime [OK]

Common Mistakes:

• Restarting all services unnecessarily
• Merging services causing complexity
• Stopping all services causing downtime

5. Uber wants to handle a sudden surge of users during a big event. Which architectural approach best supports this scaling need?

hard

A. Limit user requests to reduce load manually

B. Combine all services into one monolithic app for faster response

C. Use microservices with auto-scaling and load balancing

D. Use a single powerful server to handle all traffic

Solution

1. Step 1: Understand scaling in microservices

   Microservices allow scaling individual parts independently using auto-scaling and load balancing.

2. Step 2: Compare options for surge handling

   Monolithic apps and single servers can't scale easily; limiting users reduces experience.

3. Final Answer:

   Use microservices with auto-scaling and load balancing -> Option C

4. Quick Check:

   Scaling surge = Microservices + auto-scaling [OK]

Hint: Auto-scale microservices to handle traffic spikes smoothly [OK]

Common Mistakes:

• Thinking monolith scales better
• Relying on single server power
• Manually limiting users instead of scaling