Intelligent Core Assignment for Data Storage Systems

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional data storage systems face inefficiencies due to inflexible pre-defined core allocation profiles, which may not be optimal for performance, requiring expert administration and lacking adaptability based on actual usage data.

Innovation Solution

A method that utilizes historical workload data to simulate operations, generate utilization scores, and select or update core assignments for optimal profile selection, allowing for automatic modification of core allocations based on usage patterns.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If pre-defined core assignment profiles are used, then system configuration is simplified and ease of operation is improved, but adaptability to actual usage patterns deteriorates and optimal performance cannot be guaranteed

Engineering Contradiction:
Improveease of profile selectionVSAvoidadaptability to usage patterns
Core Design Contradiction:
Ease of operationVSAdaptability or versatility

Solution Approach 1:

The system implements feedback by analyzing historical workload data to generate utilization scores that indicate how well each pre-defined profile performs under actual usage conditions. This feedback loop allows the system to recommend optimal profiles based on real usage patterns while maintaining the simplicity of pre-defined configurations.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary analysis by simulating workload execution on each profile before deployment, generating utilization scores in advance. This preliminary action enables the system to pre-determine optimal profile selections based on historical data, combining advance preparation with operational simplicity.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If expert administration is required to select optimal profiles, then performance optimization capability is improved, but device complexity and difficulty of operation increase

Engineering Contradiction:
Improveperformance optimization capabilityVSAvoidcomplexity of profile selection process
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system performs self-service by automatically analyzing workload data and generating utilization scores to identify optimal profiles without requiring expert administration. The system autonomously evaluates each profile's performance characteristics and provides recommendations, eliminating the need for specialized knowledge while maintaining optimization capability.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system replaces the mechanical process of expert analysis and manual profile selection with an automated computational system that uses workload data simulation and utilization score calculation. This substitution transforms a complex expert-dependent process into an automated, data-driven decision-making system.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If fixed pre-defined profiles are used, then system stability and reliability are improved, but adaptability and ability to optimize based on actual usage deteriorate

Engineering Contradiction:
Improvesystem stabilityVSAvoidadaptability based on usage data
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system introduces dynamics by enabling profiles to be re-evaluated and re-selected based on changing workload patterns. While the profiles themselves remain stable pre-defined configurations, the system dynamically adapts profile selection by analyzing historical data and generating utilization scores that reflect current usage conditions, combining stability with adaptability.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS10038744B1Intelligent core assignment
Publication Date: 2018.07.31 EMC IP HLDG CO LLC
  • US10038744B1 patent drawing
  • US10038744B1 patent drawing
  • US10038744B1 patent drawing

AI summary

A method includes (a) receiving workload data for a data storage system, the data storage system having a plurality of processing cores, the data storage system being configured to operate according to any of a plurality of profiles, each of the plurality of profiles having a respective assignment of respective cores of the plurality of cores to particular storage roles of the data storage system, (b) for each profile of the plurality of profiles, simulating operation of the data storage system running a workload based on the received workload data using the assignment of respective cores of the plurality of cores to particular storage roles for that profile, thereby producing a set of utilization scores, (c) selecting a profile of the plurality of profiles based on the set of utilization scores, and (d) outputting an indication of the selected profile for use by the data storage system.