This article explains the principles, advantages, and drawbacks of various distributed transaction solutions—including XA two‑phase commit, three‑phase commit, TCC, Saga, and message‑based eventual consistency—while showing how they are implemented in Java/Spring environments and when each should be used.
This article explains the principles, workflows, advantages and drawbacks of various distributed transaction solutions—including XA two‑phase commit, Java Transaction API, chained transaction managers, three‑phase commit, TCC, reliable message‑based eventual consistency and maximum‑effort notification—while comparing ACID and BASE theories for modern backend systems.
The article presents a scalable, two‑dimensional order state machine architecture that separates vertical business isolation via annotated StateProcessor classes from horizontal logic reuse through modular checkers and plugins, employing distributed locks, optimistic DB locking, and a registry to ensure extensible, consistent multi‑dimensional order processing.
This article explains the fundamentals of distributed transactions, compares local and global transaction models, examines protocols like XA and 2PC, introduces flexible transaction patterns such as TCC and compensable operations, and evaluates reliable message‑based solutions for achieving eventual consistency.
