Peer-to-Peer Commit Protocol for Distributed Transaction Reliability

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing distributed transaction processing protocols, such as two-phase and three-phase commit protocols, are not resilient to all failure configurations, particularly when the coordinator and a processing node fail simultaneously, leading to blocking issues and increased network usage, complexity, and lack of formal safety proof.

Innovation Solution

A peer-to-peer commit protocol where processing nodes communicate commit statuses directly, eliminating the need for a coordinator to decide on commit or abort, reducing network traffic, and introducing a strongly connected network to minimize failure points, with mechanisms for fail-stop and crash recovery.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a coordinator-based commit protocol (two-phase or three-phase) is used, then transaction coordination is achieved, but network traffic increases and failure points increase

Engineering Contradiction:

Improvetransaction coordination reliabilityVSAvoidnetwork traffic and failure points

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent removes the coordinator entity from the commit protocol, extracting the centralized coordination function and distributing it peer-to-peer among all processing nodes. Each node independently determines commit status based on receiving commit messages from all other nodes, eliminating the single point of failure and reducing network overhead associated with coordinator-based protocols.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the centralized coordination function into distributed autonomous decisions at each peer node. Instead of one coordinator managing all commits, each node independently evaluates its own commit status based on peer messages, dividing the coordination task across multiple independent units that can fail without bringing down the entire system.

Inventive Principle:

Principle #1Segmentation

2Device complexity

If peer-to-peer commit messages are sent to all processing nodes, then network traffic is reduced, but ensuring all nodes receive commit status becomes more complex

Engineering Contradiction:

Improvenetwork trafficVSAvoidcommit status propagation verification

Core Design Contradiction:

Device complexityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent merges the commit decision-making process across all peer nodes, where each node's commit status depends on receiving commit messages from all other nodes. This mutual dependency creates a self-verifying system where successful commit propagation is detected when every node has received commits from every other node, eliminating the need for separate verification mechanisms.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent implements feedback through commit messages that flow peer-to-peer between all nodes. Each node sends its commit status to all other nodes, creating a feedback loop where the completion of commit propagation is detected when each node receives acknowledgments from all peers, providing automatic verification without centralized control.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS9465648B2Distributed transaction processing through commit messages sent to a downstream neighbor

Publication Date: 2016.10.11 HEWLETT PACKARD ENTERPRISE DEV LP

AI summary

A system includes an initiator and processing nodes. The initiator distributes portions of a transaction among the processing nodes. Each processing node has at least one downstream neighbor to which the processing node sends commit messages. The commit messages include a commit status of the processing node. The downstream neighbor is also a processing node.