Granular Property Grouping for Distributed Storage Change Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current distributed storage systems face high latency and compute load due to blanket notifications of changes in stored data, which result in wasteful resource usage as all scenario computing services receive notifications regardless of relevance, leading to unnecessary resource intensive operations.

Innovation Solution

Implementing granular change detection in a distributed storage system, where a change detector compares specific properties of documents and only notifies scenario computing services that have registered interest in those changes, using hash functions to efficiently determine property changes and reduce unnecessary read operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If blanket notifications are sent to all scenario computing services when data changes, then all services receive updates, but resource waste increases due to unnecessary operations on irrelevant services

Engineering Contradiction:
Improvenotification deliveryVSAvoidcompute load
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent segments the notification delivery process by dividing scenario computing services into groups based on their interest in specific data properties. Instead of sending blanket notifications to all services, the system selectively notifies only those services that have registered interest in changed properties, thereby reducing unnecessary compute operations while ensuring reliable delivery to relevant services.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements local quality by making notification delivery property-specific and service-specific. Each scenario computing service registers interest in particular properties, and notifications are tailored to match this local interest. This ensures that each service receives only the notifications relevant to its specific needs, eliminating waste from processing irrelevant updates.

Inventive Principle:
Principle #3Local quality

2Reliability

If blanket notifications are sent to all scenario computing services, then completeness of notification is maintained, but network bandwidth consumption increases

Engineering Contradiction:
Improvenotification completenessVSAvoidnetwork bandwidth
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent segments the notification stream into property-specific notifications rather than sending monolithic blanket notifications. By dividing notifications based on changed properties and matching them against service interests, the system reduces network bandwidth consumption while maintaining completeness of delivery for all relevant services.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts only the necessary notification information by identifying which properties have changed and which services are interested in those properties. This extraction process eliminates redundant notification data from the network transmission, reducing bandwidth consumption while preserving notification completeness for affected services.

Inventive Principle:
Principle #2Taking out (Extraction)

3Reliability

If all scenario computing services process all data changes, then no changes are missed, but resource waste occurs from processing irrelevant changes

Engineering Contradiction:
Improvechange detection completenessVSAvoidresource efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the change processing workload by dividing it into property-specific change events. Each scenario computing service is assigned to process only those property changes that match its registered interests, ensuring no relevant changes are missed while eliminating waste from processing irrelevant changes.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements local quality by making change processing service-specific and property-specific. Each service processes changes locally relevant to its function, rather than universally processing all changes. This improves resource efficiency while maintaining completeness of change detection for each service's domain.

Inventive Principle:
Principle #3Local quality

4Productivity

If granular change detection is implemented, then resource usage is optimized, but system complexity increases

Engineering Contradiction:
Improveresource efficiencyVSAvoidsystem architecture
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by having scenario computing services pre-register their interests in specific properties before changes occur. This advance registration creates a lookup structure that enables efficient matching of property changes to interested services, reducing the complexity of real-time decision-making while optimizing resource usage.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary component that manages the matching between property changes and service interests. This intermediary layer handles the complexity of granular change detection and notification routing, shielding individual services from complexity while enabling optimized resource usage through selective notification.

Inventive Principle:
Principle #24Intermediary (Mediator)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach reduces compute load, network bandwidth consumption, and resource waste by selectively transmitting notifications only to services interested in specific property changes, thereby optimizing resource usage and improving system performance.

Implementation Method 1

Implementing granular change detection in a distributed storage system, where a change detector compares specific properties of documents and only notifies scenario computing services that have registered interest in those changes, using hash functions to efficiently determine property changes and reduce unnecessary read operations.

Methodology Applied
Scientific EffectHash function:

Data Source

PatentEP3959643B1Property grouping for change detection in distributed storage systems
Publication Date: 2023.08.23 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Techniques of property grouping for change detection in distributed storage systems are disclosed herein. In one embodiment, a method includes grouping multiple properties of a document into multiple groups individually including a subset of the multiple properties. For each of the multiple groups, a hash value of the group is generated based on values of the subset of properties in the group. The method further includes determining whether the generated hash value of the group is different than that of a corresponding group in a previous version of the document in the distributed storage system. In response to determining that generated hash value of the group is not different than that of the corresponding group in the previous version, metadata is inserted into the new version of the document indicating that none of the subset of properties in the group has changed.