yedf/dtm
Dtm
dtm 是一个分布式事务框架和工作流编排器，旨在管理微服务之间的数据一致性。它作为一个中心化协调器，同步分布式事务并管理复杂分布式任务序列的状态和执行流。
该框架实现了多种协调模式，包括 Saga、TCC 和 XA，以防止同步错误。它特别提供了一个多语言事务协调器，用于同步用不同编程语言编写的服务之间的操作，并利用可靠的消息发件箱（outbox）将数据库更新与消息分发耦合。
其功能涵盖了通过补偿事务回滚、两阶段提交协调和精确一次（exactly-once）消息传递实现的分布式系统恢复。该基础设施支持水平容量扩展和高可用性配置，以在请求量增加时保持稳定性。
Features
- Business Process Orchestration - Functions as a distributed workflow orchestrator that manages the execution flow of complex business processes.
- Distributed Workflow Orchestrators - Coordinates complex sequences of distributed tasks while managing their execution flow and state.
- Reliable Message Delivery - Implements exactly-once message delivery using a reliable messaging outbox to maintain data integrity across microservices.
- Automatic Compensation Mechanisms - Automatically invokes compensating operations for sub-transactions to restore system consistency after a failure.
- Distributed State Recovery - Automatically invokes compensating operations to roll back failed sub-transactions and restore system consistency.
- Distributed Transaction Coordinators - Provides a centralized coordinator to manage global transaction lifecycles and maintain data synchronization across microservices.
- Polyglot Coordinators - Provides a centralized coordinator that synchronizes distributed transactions across services written in different languages.
- Saga Coordinators - Coordinates saga-based distributed transactions by executing compensating transactions in reverse order during failures.
- Two-Phase Commit Protocols - Manages a two-phase commit protocol to ensure atomic updates across distributed data stores.
- Distributed Transaction Management - Ensures data consistency across microservices using distributed transaction patterns like Saga, TCC, and XA.
- Try-Confirm-Cancel Implementations - Coordinates the Try, Confirm, and Cancel phases to manage resource reservation in distributed business operations.
- Consistency Frameworks - Provides a comprehensive framework for maintaining data integrity and synchronization across distributed services.
- State Machine Orchestrators - Uses a persistent state machine to track the lifecycle and ensure consistent progress of distributed transactions.
- Transactional Outbox Patterns - Implements the transactional outbox pattern to ensure atomicity between database updates and event publishing.
- Domain Workflow Orchestration - Orchestrates complex sequences of distributed tasks to fulfill specific business use cases.
- Horizontal Scaling Deployments - Supports horizontal capacity expansion and high availability to maintain stability as request volumes increase.
- Event Driven Messaging - Provides an event-driven messaging layer to trigger asynchronous task execution across remote microservices.
- Outbox Pattern Implementations - Couples database updates with message dispatching to ensure exactly-once delivery and prevent data loss.
- Database Tools - Distributed transaction management service.
