LeetCode Backend System

A production-ready, backend coding platform that enables users to solve algorithmic problems with real-time code evaluation across multiple programming languages.

ExpressJs, TypeScript, NodeJs, MongoDB, Redis, Docker | Sep 2025

Problem Submission Evaluation

Service	Architecture	Scale
Microservices	3 Core Services	Python & C++ Supported

Overview

Leetcode is a scalable, microservices-based coding platform backend. It features secure code execution in isolated Docker containers, real-time evaluation, and comprehensive problem management capabilities.

Key Highlights

Secure Execution: Code runs in isolated Docker containers with strict resource limits (10 PIDs, 50% CPU quota, NetworkMode: none).
Real-time Processing: Asynchronous evaluation pipeline using Redis and BullMQ to decouple submissions from execution.
Multi-language Support: Python (2s timeout) and C++ (1s timeout).
Comprehensive Results: Detailed test case evaluation (AC, WA, TLE).

Interactive System Flow

System Architecture

The system consists of three core microservices working in harmony: 1. Problem Service: Handles Problem CRUD operations, search, and filtering. 2. Submission Service: Manages code submissions and the Redis evaluation queue. 3. Evaluation Service: Acts as a background worker, creating Docker containers to execute code and process results.

Evaluation Pipeline

sequenceDiagram
    participant U as User
    participant S as Submission Service
    participant R as Redis Queue
    participant E as Evaluation Service
    participant D as Docker Container

    U->>S: Submit Code
    S->>R: Queue Submission
    S->>U: Return Submission ID

    E->>R: Pull Submission
    E->>D: Create Docker Container
    E->>D: Execute Code
    D->>E: Return Results
    E->>S: Update Status

    U->>S: Check Results
    S->>U: Return Output

Features

Problem Management

CRUD Operations: Complete problem lifecycle management
Difficulty Levels: Easy, Medium, Hard categorization
Rich Descriptions: Full Markdown support for problem statements
Custom Test Cases: Flexible input/output validation

Code Execution Engine

Multi-language Support:
Python: 2-second timeout
C++: 1-second timeout
Secure Isolation: Docker containerization with resource limits
Real-time Processing: Asynchronous evaluation pipeline

Infrastructure

Microservices Architecture: Independently scalable services
Message Queuing: Redis with BullMQ for reliable job processing
Container Management: Automated Docker lifecycle management
Observability: Winston logging with correlation tracking