toyDB Architecture

toyDB is a simple distributed SQL database, intended to illustrate how such systems are built. The overall structure is similar to real-world distributed databases, but the design and implementation has been kept as simple as possible for understandability. Performance and scalability are explicit non-goals, as these are major sources of complexity in real-world systems.

This guide will walk through toyDB's architecture and code from the bottom up, with plenty of links to the actual source code.

ℹ️ View on GitHub with a desktop browser for inline code listings.

• Overview
  • Properties
  • Components
• Storage Engine
  • Memory Storage Engine
  • BitCask Storage Engine
• Key/Value Encoding
  • Bincode Value Encoding
  • Keycode Key Encoding
• MVCC Transactions
• Raft Consensus
  • Log Storage
  • State Machine Interface
  • Node Roles
  • Node Interface and Communication
  • Leader Election and Terms
  • Client Requests and Forwarding
  • Write Replication and Application
  • Read Processing
• SQL Engine
  • Data Model
    • Data Types
    • Schemas
    • Expressions
  • Storage
    • Key/Value Representation
    • Schema Catalog
    • Row Storage and Transactions
  • Raft Replication
  • Parsing
    • Lexer
    • Abstract Syntax Tree
    • Parser
  • Planning
    • Execution Plan
    • Scope and Name Resolution
    • Planner
  • Optimization
    • Constant Folding
    • Filter Pushdown
    • Index Lookups
    • Hash Join
    • Short Circuiting
  • Execution
    • Plan Executor
    • Session Management
• Server
  • Raft Routing
  • SQL Service
  • toydb Binary
• Client
  • Client Library
  • toysql Binary

---

<p align="center"> <a href="overview.md">Overview</a> → </p>