Data Placement Arbitrator for Multi-Tiered Storage Conflict Resolution

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Multi-tiered storage systems face inefficiencies and increased costs due to concurrent processes optimizing data re-allocation, leading to conflicts and unnecessary data movement, as existing systems either allow all processes to run simultaneously or employ algorithms that do not complement each other, requiring system redesign for changes.

Innovation Solution

A system comprising a data collector, data placement optimizers, a data placement arbitrator, and a data mover is implemented to collect system configuration and performance data, generate data movement plans, arbitrate conflicts, and execute optimized data placement plans, reducing overhead and improving efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple data placement optimizers run concurrently to optimize data re-allocation, then data placement optimization capability is improved, but conflicts arise and efficiency decreases

Engineering Contradiction:

Improvedata placement optimization capabilityVSAvoiddata placement efficiency

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

A data placement arbitrator is introduced as an intermediary component that receives data movement plans from multiple concurrent data placement optimizers, resolves conflicts between them, and coordinates their execution. This mediator enables multiple optimizers to run simultaneously without causing conflicts, thus maintaining high optimization capability while improving overall efficiency through coordinated execution.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If all processes are allowed to run simultaneously despite data conflicts, then system flexibility is improved, but unnecessary data movement increases and efficiency decreases

Engineering Contradiction:

Improvesystem flexibilityVSAvoidunnecessary data movement

Core Design Contradiction:

Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The data placement arbitrator implements a feedback mechanism where it receives data movement plans from multiple optimizers, identifies conflicts through analysis, and provides coordinated execution decisions back to the optimizers. This feedback loop enables the system to maintain flexibility by allowing all processes to be submitted, while simultaneously preventing unnecessary data movement through intelligent conflict detection and resolution.

Inventive Principle:

Principle #23Feedback

3Reliability

If multiple algorithms are added to mediate conflicts, then conflict resolution capability is improved, but system complexity increases and algorithms may not complement each other

Engineering Contradiction:

Improveconflict resolution capabilityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

Multiple conflict resolution algorithms are merged into a single unified data placement arbitrator component. This consolidation allows the arbitrator to implement a comprehensive set of conflict resolution strategies (such as priority-based resolution, fairness-based resolution, and conflict avoidance) within one coordinated system, improving conflict resolution capability while avoiding the complexity of multiple separate algorithms that do not complement each other.

Inventive Principle:

Principle #5Merging (Combining)

4Adaptability or versatility

If algorithms are added or removed from the system, then adaptability is improved, but system redesign is required

Engineering Contradiction:

Improvealgorithm configurabilityVSAvoidsystem implementation ease

Core Design Contradiction:

Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

The system is segmented into independent, modular components: data placement optimizers, data placement arbitrator, and data mover. Each optimizer can be independently configured with different algorithms, and the arbitrator can be configured with different conflict resolution strategies. This modular segmentation allows algorithms to be added or removed by simply configuring the appropriate module without requiring redesign of the entire system, thus improving adaptability while maintaining ease of implementation.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS8856476B2Systems, methods, and physical computer storage media to optimize data placement in multi-tiered storage systems

Publication Date: 2014.10.07 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

For optimizing data placement in a multi-tiered storage system, system configuration data and system performance data is collected. A plurality of data movement plans are generated, based in part on the system configuration data and the system performance data. A conflict between the plurality of data movement plans are arbitrated to form an execution plan. The data movement plans are performed according to the execution plan.