Reminder Indices for Data Coherency in Partitioned NoSQL Storage

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

NoSQL databases with 'eventually consistent' processes face challenges in maintaining immediate data coherency across distributed systems, leading to delays in propagating changes and increased computational and transactional efforts in servicing queries, especially in large-scale deployments.

Innovation Solution

The implementation of promoted properties and aggregated properties allows for efficient data handling by correlating data fields across partitioned storage spaces, reducing the need for multiple queries and improving data retrieval through pre-computed views and replication ordering, while using reminder indices and change reminders to manage data changes and replication efficiently.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If NoSQL databases use eventually consistent processes to store data in distributed systems, then adaptability and flexibility are improved, but data coherency and consistency deteriorate with delays in propagating changes

Engineering Contradiction:

ImproveflexibilityVSAvoiddata coherency

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent applies preliminary action by pre-computing views and maintaining reminder indices that track changes before they need to be propagated. Change reminders are scheduled in advance to ensure data coherency is restored across partitioned storage spaces without requiring immediate consistency, thus maintaining flexibility while improving reliability.

Inventive Principle:

Principle #10Preliminary action

2Loss of information

If multiple queries are executed to retrieve correlated data from partitioned storage spaces, then data retrieval completeness is improved, but response time and computational effort increase

Engineering Contradiction:

Improvedata retrieval completenessVSAvoidresponse time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The patent merges multiple queries into a single efficient query by pre-computing views that combine data from multiple partitioned storage spaces. The reminder index correlates data fields across partitions, allowing the system to retrieve all necessary correlated data in one operation rather than executing multiple separate queries, thus reducing response time while maintaining data completeness.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent applies preliminary action by pre-computing views that organize and correlate data from partitioned storage spaces before queries are executed. This pre-computation reduces the computational effort and response time required during actual query operations, as the data is already structured and correlated in advance.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If data changes are propagated immediately across distributed storage spaces, then data coherency is improved, but processor load and transactional efforts increase

Engineering Contradiction:

Improvedata coherencyVSAvoidprocessor load

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent applies periodic action by using change reminders that are scheduled to propagate data changes at intervals rather than immediately. The reminder index tracks changes and schedules their propagation across partitioned storage spaces, reducing instantaneous processor load while ensuring data coherency is eventually restored through periodic execution of these change reminders.

Inventive Principle:

Principle #19Periodic action

4Productivity

If pre-computed views are maintained across partitioned storage spaces, then data retrieval efficiency is improved, but device complexity and storage requirements increase

Engineering Contradiction:

Improvedata retrieval efficiencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the pre-computed views into partitioned storage spaces, with each partition managing a subset of the data. The reminder index is also segmented across partitions, allowing the system to maintain pre-computed views for improved retrieval efficiency while distributing the complexity and storage requirements across multiple manageable segments rather than a single monolithic structure.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS10235406B2Reminder processing of structured data records among partitioned data storage spaces

Publication Date: 2019.03.19 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Systems, methods, and software for management of partitioned data storage spaces is provided herein. An exemplary method includes storing sets of structured data records among partitioned data storage spaces, with data fields of the structured data records correlated among the sets by relational associations. The method includes receiving a change action related to a selected structured data record, and responsive to the change action, scheduling execution of the change action in a reminder index. The method includes executing the change action according to at least scheduling indicated by the reminder index.