Assisting Node for Database Replication During Communication Failure

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In high-availability database clusters, communication failures between primary and backup clusters can lead to data inconsistency and corruption, especially during hardware or network failures, as existing solutions often require extensive reintegration of transactions and may reduce replication system performance.

Innovation Solution

A method where an assisting node within the primary cluster takes over to transmit logs to the backup cluster, maintaining data integrity by running an assisting process that forwards untransmitted logs and new logs to the replay master, even during communication failures, ensuring continuous replication and data consistency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a member node in the primary cluster experiences communication failure when sending logs to the replay master, then data consistency is compromised, but using existing solutions requires extensive reintegration of transactions and reduces replication system performance

Engineering Contradiction:
Improvedata consistencyVSAvoidreplication system performance
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent introduces an intermediary component called the 'log shipping assistant' that acts as a mediator between the failed member node and the replay master. When communication failure is detected, the assistant takes over the log shipping function, forwarding logs from the failed node to the replay master without requiring transaction reintegration. This intermediary approach maintains data consistency while avoiding the performance penalty associated with reprocessing transactions.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If communication failure occurs between primary and backup clusters, then data integrity is at risk, but implementing robust failure handling increases system complexity

Engineering Contradiction:
Improvedata integrityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The log shipping assistant operates autonomously to detect communication failures and self-correct by taking over log shipping duties. The system monitors its own communication status and automatically activates the assistant when needed, without requiring complex external intervention or manual reconfiguration. This self-service approach maintains data integrity while minimizing the complexity overhead through automated, localized response to failures.

Inventive Principle:
Principle #25Self-service

3Loss of information

If logs are not transmitted during communication failure, then data loss occurs, but retransmitting logs after failure increases transmission time and reduces replication speed

Engineering Contradiction:
Improvetransaction lossVSAvoidtransmission time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The log shipping assistant continuously monitors communication status and proactively takes over log shipping before data loss can occur. By detecting communication failures early and immediately assuming the log shipping function, the system prevents transaction loss without requiring post-failure retransmission. This preliminary action approach eliminates the time penalty associated with retransmitting logs after failure recovery.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS9940206B2Handling failed cluster members when replicating a database between clusters
Publication Date: 2018.04.10 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US9940206B2 patent drawing
  • US9940206B2 patent drawing
  • US9940206B2 patent drawing

AI summary

Data integrity is maintained during failed communications between a member node of a primary cluster and a backup cluster by assigning an assisting member node to run an assisting process that transmits data entered into the member node to the backup cluster. In this way, a replicated database is maintained during a partial communication failure between the primary cluster and the backup cluster.