Object Arrangement Apparatus for Uniform Data Distribution

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for distributing and storing data across multiple storage systems face challenges in efficiently managing storage fluctuations, particularly in large-scale data centers, where they require significant memory resources and processing time to maintain uniform distribution and minimize data movement.

Innovation Solution

An object arrangement apparatus and method that uses a pseudo-random number generation process to determine storage destinations for data, employing a hash operation and a number-storage conversion table to minimize data movement and maintain uniform distribution across changing storage configurations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If a hash operation with modular arithmetic is used to determine storage destinations, then uniform distribution of data across storages is achieved, but significant data movement is required when the number of storages changes

Engineering Contradiction:
Improveuniform distribution of dataVSAvoiddata movement time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent applies dynamics by making the hash function adaptive to changes in the number of storages. Instead of using a fixed modular arithmetic approach, the system dynamically adjusts the hash function parameters based on the current number of storages, allowing data to be redistributed efficiently without requiring complete remapping when storage capacity changes.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes parameters of the hash function based on the number of storages. By modifying the hash function's internal parameters (such as the modulus value or transformation constants) according to the current storage configuration, the system maintains uniform distribution while minimizing data movement overhead during storage scaling events.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If a table holding corresponding relationships between object data and storage is used, then data access is simplified, but memory resources are significantly consumed

Engineering Contradiction:
Improvedata access simplicityVSAvoidmemory resources
Core Design Contradiction:
Ease of operationVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential information needed for data access from a complete mapping table. Instead of storing full correspondence relationships between all object data and storages, the system extracts and stores only the hash function parameters and metadata necessary to reconstruct the mapping on-demand, significantly reducing memory consumption while maintaining access efficiency.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system implements self-service by enabling each node to independently compute storage destinations using the hash function without requiring centralized table lookups. Each storage node can autonomously determine which data it should hold by applying the hash function to data identifiers, eliminating the need for large centralized mapping tables and reducing overall memory requirements.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP2608043B1Object arrangement apparatus, method therefor, and computer program
Publication Date: 2022.10.12 NEC CORP
  • EP2608043B1 patent drawingFigure 1
  • EP2608043B1 patent drawingFigure 2
  • EP2608043B1 patent drawingFigure 3

AI summary

Disclosed is an object arrangement apparatus which arranges a plurality of objects approximately uniformly to a plurality of arrangement destinations by a comparatively easy processing configuration. This object arrangement apparatus includes an arrangement destination determination means. The arrangement destination determination means refers to sequence information generated based on unique information for objects and refers to corresponding information between arrangement destination identifiers identifying arrangement destinations of the objects and second numbers associated with the arrangement destination identifiers. The arrangement destination determination means generates first numbers by using the sequence information, the first numbers not having a correlation therebetween when the first numbers are generated with different sequence information, and repeats generating a new first number until the first number matches the second number contained in the corresponding information. The arrangement destination determination means determines, when the first number matches any of the second numbers, to arrange the object to an arrangement destination corresponding to the arrangement destination identifier related to the matched second number.