RobustMQ

English

简体中文

English

简体中文

Data Source ​

RobustMQ MQTT supports multiple data sources for authentication, authorization, and blacklist data.

What Is a Data Source ​

A data source is where identity and policy data comes from:

  • users
  • ACL rules
  • blacklist entries

Broker loads these data sets and uses them in the connection auth and access-control path.

Why Multiple Data Sources ​

  • Different teams already store identity data in different systems.
  • Some scenarios prefer built-in metadata, others require external system integration.
  • It enables gradual migration: start built-in, move to external sources when needed.
  • It avoids hard-binding to a single storage model.

How to Use ​

Typical steps:

  1. Configure storage_type through the management console or CLI.
  2. Fill the matching config block (for example mysql_config, redis_config, mongodb_config).
  3. For SQL-based sources, configure queries and ensure result mappings match expected contracts.
  4. Validate auth result and cache sync behavior after startup.

Runtime Principle ​

The auth path is cache-first, with different update strategies per source:

  1. CONNECT auth checks in-memory cache first.
  2. Built-in data source (Meta Service) updates cache in real time (no delay).
  3. External data sources (MySQL/PostgreSQL/Redis/MongoDB/HTTP) update cache by periodic sync.
  4. External sources are used for sync/management, not queried on every CONNECT.

By moving the hot path to memory, broker remains stable under connection bursts.
By default, external source sync has around a 5-second delay; built-in source does not.

Currently Supported Data Sources ​

  • Built-in Data Source (Meta Service)
  • MySQL
  • PostgreSQL
  • Redis
  • MongoDB
  • HTTP

Design Goals ​

  • No forced fixed table schema.
  • SQL/query-driven adaptation for existing systems.
  • Cache-first auth hot path with manageable consistency trade-offs.

Subpages ​