Database Replication Acknowledgement Threshold for Latency Reduction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

High-availability database systems face performance delays and system instability due to the time-consuming process of propagating transaction log records from a source database to target databases, which can lead to significant latency and system read-only conditions in case of target database failures.

Innovation Solution

A database replication system with a master data node and replica data nodes, where the master node sends transaction log records to replica nodes with an acknowledgement threshold less than the total number of replica nodes, ensuring that changes are persisted before committing transactions and allowing for automatic failover in case of failures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the source database system waits for all target database systems to store transaction log records before completing commit processing, then data consistency and reliability are improved, but system latency and commit processing time increase significantly

Engineering Contradiction:

Improvedata consistencyVSAvoidcommit processing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent applies partial action by requiring acknowledgments from only a threshold number of target database systems (e.g., majority) rather than all target systems before completing the commit. This allows the source database to proceed with commit processing once sufficient replicas have confirmed storage, reducing latency while maintaining data consistency through the threshold mechanism that ensures adequate replication for recovery purposes.

Inventive Principle:

Principle #16Partial or excessive action

2Reliability

If the source database system propagates transaction log records to many target database systems, then system availability and fault tolerance are improved, but commit processing performance and speed deteriorate

Engineering Contradiction:

Improvesystem availabilityVSAvoidcommit processing speed

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system maintains high availability by replicating to multiple target database systems but applies partial action by not requiring all of them to acknowledge before commit completion. The threshold parameter allows the system to configure the balance between availability (more targets) and performance (fewer acknowledgments required), enabling scalable deployment without linear performance degradation.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent introduces an adjustable threshold parameter that controls the number of acknowledgments required for commit completion. This parameter can be tuned based on system requirements - higher thresholds provide stronger consistency guarantees while lower thresholds improve performance. The dynamic parameter adjustment allows optimization of the trade-off between availability and processing speed.

Inventive Principle:

Principle #35Parameter changes

3Manufacturing precision

If the acknowledgement threshold is set to the total number of replica data nodes, then data integrity is ensured, but system latency and response time increase

Engineering Contradiction:

Improvedata integrityVSAvoidsystem latency

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by setting the acknowledgment threshold to a value less than the total number of replica data nodes (e.g., majority or quorum). This ensures data integrity through sufficient replication while avoiding the latency of waiting for all replicas. The threshold provides a balance where data can be recovered and consistency maintained even if some replicas fail, without requiring unanimous acknowledgment.

Inventive Principle:

Principle #16Partial or excessive action

4Reliability

If the source database system requires confirmation from all target database systems before committing, then reliability is improved, but ease of operation and user experience deteriorate due to significant delays

Engineering Contradiction:

Improvetransaction commitment reliabilityVSAvoiduser response time

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system improves ease of operation by requiring acknowledgments from only a threshold number of target database systems rather than all of them. This reduces the time users wait for transaction confirmation while maintaining reliability through the threshold mechanism. Users experience faster response times because the system does not block commit completion until every single replica has acknowledged, yet data integrity is preserved through sufficient replication.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12124478B2System and method for persistence and replication of changes to a data store

Publication Date: 2024.10.22 OPEN TEXT CORP

AI summary

A database replication system comprises a plurality of replica data nodes and a master data node. The master data node is configured with an acknowledgement requirement that comprises an acknowledgement threshold that is less than a total number of data nodes in a plurality of data nodes. The master data node is configured to send, to the plurality of replica data nodes, a transaction log record that includes an update to data in the master database that has not been committed to the master database, receive acknowledgements of the transaction log records and based on a determination that the received acknowledgements of the transaction log record meet the acknowledgement requirement, set the transaction log record as a master node last acknowledged transaction log record in a master node transaction log. The last acknowledged transaction log record can be used to control recovery.