Replication Latency Routing for Data Synchronization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data processing systems face challenges in maintaining synchronization among independent processors, leading to inefficiencies in workload distribution when replication latency exceeds acceptable thresholds, affecting the proper operation of processors and their databases.

Innovation Solution

Defining a maximum replication latency threshold and a reset threshold, estimating replication latency, and dynamically switching between different replication latency states (normal, severe, critical, and acute) to adjust workload distribution accordingly, ensuring that query transactions are routed to the most synchronized site to maintain system stability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Power

If data is distributed to independent processors to perform calculations, then processing capability and system scalability are improved, but synchronization between processing sites deteriorates when replication latency increases

Engineering Contradiction:

Improveprocessing capabilityVSAvoidsynchronization between processing sites

Core Design Contradiction:

PowerVSStability of the object's composition

Solution Approach 1:

The system dynamically adjusts query routing based on real-time replication latency conditions. When latency is within acceptable thresholds, queries are distributed to multiple sites for parallel processing. When latency exceeds thresholds, the system switches to routing all queries to the primary site to ensure data consistency, thus adapting the system behavior to current synchronization state

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system monitors replication latency as a key parameter and uses threshold-based logic to change operational parameters. When latency exceeds defined thresholds (acute replication latency state), the routing parameter changes from distributed to centralized query processing, and vice versa when latency returns to acceptable levels

Inventive Principle:

Principle #35Parameter changes

2Reliability

If query transactions are routed to the most synchronized site to maintain stability, then data consistency is improved, but system productivity may deteriorate due to reduced parallel processing

Engineering Contradiction:

Improvedata consistencyVSAvoidsystem throughput

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system dynamically switches between distributed and centralized query routing based on replication latency conditions. In normal operation with acceptable latency, queries are distributed to multiple sites for parallel processing (high productivity). When latency exceeds thresholds, the system switches to routing all queries to the primary site (high reliability), and automatically returns to distributed mode when latency improves

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system continuously monitors replication latency and uses this feedback to adjust query routing decisions. The feedback loop ensures that routing decisions are based on real-time synchronization status, automatically optimizing the balance between data consistency and processing throughput

Inventive Principle:

Principle #23Feedback

3Reliability

If the system switches to acute replication latency state when latency exceeds threshold, then data access issues are prevented, but system complexity increases due to state management

Engineering Contradiction:

Improvedata access reliabilityVSAvoidreplication latency state management

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system uses simple threshold-based parameter changes to manage replication latency states. Two thresholds are defined: a maximum replication latency threshold to enter acute state, and a reset threshold to return to normal state. This parameter-based approach provides clear, actionable state transitions without complex decision logic

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system continuously monitors replication latency and uses feedback from latency measurements to trigger state transitions. When latency exceeds the maximum threshold, the system switches to acute replication latency state with centralized routing. When latency falls below the reset threshold, the system returns to normal distributed routing, providing automatic state management through feedback

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS9229999B2Using average replication latency to influence routing query transactions

Publication Date: 2016.01.05 KMIZRA LLC

AI summary

A data processing system and method of operation are disclosed. The data processing system includes a first site for processing data; a second site for processing data; and a transaction replay program configured to send a copy of data from the first site to the second site. A processor defines a maximum replication latency threshold and a reset threshold for the system, estimates a replication latency of a workload the system, switches the system to a first replication latency state when the estimated replication latency rises above the maximum replication latency and switches the system out of the first replication latency state to a second replication latency state when the estimated replication latency falls below the reset threshold.