Distributed Graph Write Recovery via Geo-Aware Primary Replica Migration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed graph databases, efficiently migrating a primary replica node cluster to another secondary replica node cluster for reinitiated graph data write operations after a write failure is challenging due to varying geographical locations of secondary clusters, affecting write operation efficiency.

Innovation Solution

A method and apparatus that select a new primary replica node cluster based on geographical location information of secondary replica nodes, updating replica topology information to optimize the reinitiated graph data write operation, considering factors like geographical proximity and client device location to enhance efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a primary replica node cluster is migrated to a secondary replica node cluster after write failure, then the system can reinitiate graph data write operations, but selecting an inappropriate secondary cluster based on geographical location increases communication overheads and reduces write operation efficiency

Engineering Contradiction:
Improvewrite operation success rateVSAvoidwrite operation efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent changes the selection parameter from random or arbitrary cluster choice to geographical location-based selection. By using geographical location information as a key parameter for selecting the new primary replica node cluster, the system optimizes communication overhead and maintains write operation efficiency while ensuring reliability through appropriate cluster migration.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent performs preliminary evaluation of secondary replica node clusters based on their geographical location information before selecting the new primary cluster. This preliminary action involves assessing communication overhead and geographical proximity to the client device, ensuring that the selected cluster will maintain optimal write operation efficiency from the outset.

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If the system selects a secondary replica node cluster with poor geographical location matching, then any secondary cluster can be chosen for migration, but communication overheads increase and write operation performance deteriorates

Engineering Contradiction:
Improvecluster selection flexibilityVSAvoidcommunication overhead
Core Design Contradiction:
Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The patent applies local quality by selecting the new primary replica node cluster based on its specific geographical location characteristics. Instead of treating all secondary clusters uniformly, the system evaluates each cluster's geographical proximity to the client device and selects the locally optimal cluster, thereby minimizing communication overhead and energy loss.

Inventive Principle:
Principle #3Local quality

3Device complexity

If the system does not consider geographical location information in cluster selection, then the selection process is simpler, but write operation efficiency and performance are compromised due to increased communication overhead

Engineering Contradiction:
Improveselection process complexityVSAvoidwrite operation efficiency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent performs preliminary gathering and evaluation of geographical location information for all secondary replica node clusters before the selection decision. This preliminary action stores location data and establishes evaluation criteria, so that when cluster selection is needed, the system can quickly identify the optimal cluster without complex real-time calculations, thus maintaining efficiency while managing complexity.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4647936A1Graph data write methods and graph data write apparatuses for distributed graph database
Publication Date: 2025.11.12 ALIPAY (HANGZHOU) INFORMATION TECH CO LTD
  • EP4647936A1 patent drawingFigure 1~2
  • EP4647936A1 patent drawingFigure 3
  • EP4647936A1 patent drawingFigure 4

AI summary

Embodiments of this specification provide a graph data write method and a graph data write apparatus for a distributed graph database. In response to receiving a graph data write request, a graph data write operation is initiated to a graph data storage layer based on stored replica topology information of the distributed graph database. A new primary replica node cluster is selected, at least partially based on geographical location information of a secondary replica node cluster in response to a failure of the graph data write operation, from a first secondary replica node cluster set that completes a previous round of graph data synchronization operation, and replica topology information at an engine analysis layer is synchronously updated. Then, updated replica topology information is obtained from the engine analysis layer, and the graph data write operation is reinitiated to the graph data storage layer based on the updated replica topology information.