BiteSizedChunks.comLearn one small thing at a time.

Course contentsShow

566What is a Distributed Transaction?
567The ACID Problem in Distributed Systems
568Two-Phase Commit (2PC) Overview
569The Coordinator Role in 2PC
570Phase 1: Prepare Phase
571Phase 2: Commit Phase
572Participant State Transitions
573The Blocking Problem in 2PC
574Recovery Protocols and Logs
5752PC Performance Characteristics
576When 2PC is Used in Practice
5772PC vs Single-Node Transactions
578Blocking Nature of Two-Phase Commit
579Performance Overhead of Distributed Transactions
580CAP Theorem Impact on Distributed Transactions
581The Coordinator as Single Point of Failure
582Transaction Isolation Across Systems
583Alternative: Best Effort with Eventual Consistency
584Alternative: Idempotent Operations
585Alternative: Saga Pattern Introduction
586Alternative: Event Sourcing for Consistency
587When Distributed Transactions Are Unavoidable
588The Saga Pattern: Motivation and Definition
589Saga Fundamentals: Local Transactions and Compensations
590Choreography-Based Sagas
591Orchestration-Based Sagas
592Choreography vs Orchestration Tradeoffs
593Compensating Transactions: Design Principles
594Saga Isolation and Anomalies
595Semantic Lock Pattern
596Forward Recovery vs Backward Recovery
597Saga State Management and Persistence
598Saga Frameworks and Real-World Adoption

566What is a Distributed Transaction?
567The ACID Problem in Distributed Systems
568Two-Phase Commit (2PC) Overview
569The Coordinator Role in 2PC
570Phase 1: Prepare Phase
571Phase 2: Commit Phase
572Participant State Transitions
573The Blocking Problem in 2PC
574Recovery Protocols and Logs
5752PC Performance Characteristics
576When 2PC is Used in Practice
5772PC vs Single-Node Transactions
578Blocking Nature of Two-Phase Commit
579Performance Overhead of Distributed Transactions
580CAP Theorem Impact on Distributed Transactions
581The Coordinator as Single Point of Failure
582Transaction Isolation Across Systems
583Alternative: Best Effort with Eventual Consistency
584Alternative: Idempotent Operations
585Alternative: Saga Pattern Introduction
586Alternative: Event Sourcing for Consistency
587When Distributed Transactions Are Unavoidable
588The Saga Pattern: Motivation and Definition
589Saga Fundamentals: Local Transactions and Compensations
590Choreography-Based Sagas
591Orchestration-Based Sagas
592Choreography vs Orchestration Tradeoffs
593Compensating Transactions: Design Principles
594Saga Isolation and Anomalies
595Semantic Lock Pattern
596Forward Recovery vs Backward Recovery
597Saga State Management and Persistence
598Saga Frameworks and Real-World Adoption

← System Design

Lesson 596 of 1,919·16. Distributed TransactionsPro lesson

Forward Recovery vs Backward Recovery

Choosing between compensating backward to undo work or retrying forward to completion when saga steps fail.

This lesson is for subscribers

You've completed the free preview. Subscribe to unlock every lesson in every course.

See pricing Back to course