Multi-Shot Distributed Transaction Commit (Extended Version)

TL;DR

This paper introduces Transaction Certification Service (TCS), a new formal framework for multi-shot distributed transaction commit protocols that better models modern data stores, along with a more efficient crash-resilient implementation.

Contribution

It proposes TCS as a novel formal problem capturing multi-shot commit protocols with concurrency control, and provides a more efficient crash-resilient protocol for implementing TCS.

Findings

TCS generalizes commit protocols with various isolation levels.

The proposed protocol outperforms traditional layered approaches in time complexity.

The framework better models modern transactional data store behaviors.

Abstract

Atomic Commit Problem (ACP) is a single-shot agreement problem similar to consensus, meant to model the properties of transaction commit protocols in fault-prone distributed systems. We argue that ACP is too restrictive to capture the complexities of modern transactional data stores, where commit protocols are integrated with concurrency control, and their executions for different transactions are interdependent. As an alternative, we introduce Transaction Certification Service (TCS), a new formal problem that captures safety guarantees of multi-shot transaction commit protocols with integrated concurrency control. TCS is parameterized by a certification function that can be instantiated to support common isolation levels, such as serializability and snapshot isolation. We then derive a provably correct crash-resilient protocol for implementing TCS through successive refinement. Our protocol achieves a better time complexity than mainstream approaches that layer two-phase commit on top of Paxos-style replication.