Skip to content

Latest commit

 

History

History
90 lines (49 loc) · 3.25 KB

File metadata and controls

90 lines (49 loc) · 3.25 KB
graph LR
    Server["Server"]
    Session["Session"]
    Parser["Parser"]
    Planner["Planner"]
    Executor["Executor"]
    DDL_Handler["DDL Handler"]
    Information_Schema["Information Schema"]
    Server -- "Creates" --> Session
    Session -- "Forwards SQL to" --> Parser
    Parser -- "Provides AST to" --> Planner
    Planner -- "Provides Plan to" --> Executor
    Executor -- "Returns Result to" --> Session
    Session -- "Dispatches DDL Statements to" --> DDL_Handler
    Planner -- "Queries Metadata from" --> Information_Schema
Loading

CodeBoardingDemoContact

Details

One paragraph explaining the functionality which is represented by this graph. What the main flow is and what is its purpose.

Server

Acts as the network gateway, listening for client connections and speaking the MySQL wire protocol. It handles handshake, authentication, and the creation of a dedicated Session for each authenticated client.

Related Classes/Methods:

  • server

Session

Represents a single user session and manages its context, including transaction state, session variables, and security settings. It acts as the central orchestrator, driving the SQL query through the subsequent parsing, planning, and execution stages.

Related Classes/Methods:

  • session

Parser

Responsible for lexical analysis and parsing of the SQL query text. It transforms the raw SQL string into an Abstract Syntax Tree (AST), a structured, hierarchical representation of the query's logic, which is then passed to the Planner.

Related Classes/Methods:

  • parser

Planner

Receives the AST from the Parser and performs query optimization. It validates the query, resolves names, and uses a cost-based optimizer to transform the AST into an efficient physical execution plan that can be distributed across the cluster.

Related Classes/Methods:

  • planner

Executor

The execution engine that brings the physical plan to life. It interprets the plan, dispatches tasks to the underlying storage layer (TiKV) to fetch or modify data, and processes the results before returning them to the client via the Session.

Related Classes/Methods:

  • executor

DDL Handler

A specialized component for handling Data Definition Language (DDL) statements like CREATE TABLE or ADD INDEX. It manages the complex, stateful process of performing online schema changes across the distributed cluster without blocking ongoing transactions.

Related Classes/Methods:

  • ddl

Information Schema

Provides access to database metadata, adhering to the ANSI-standard INFORMATION_SCHEMA. It exposes internal system tables containing information about schemas, tables, columns, and statistics, making the database's structure queryable via standard SQL.

Related Classes/Methods:

  • infoschema