Cross-Node Memory Pools for Hot and Cold Data Migration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing storage systems do not fully utilize the performance advantages of different types of memory, leading to inefficiencies in data access and storage utilization.

Innovation Solution

A storage system with a management node that creates a memory pool integrating storage devices with varying performance levels across nodes, allowing data migration based on access frequency to optimize usage and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data is stored only in high-performance storage devices, then data reading efficiency is improved, but storage space utilization deteriorates

Engineering Contradiction:
Improvedata reading efficiencyVSAvoidstorage space utilization
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

The storage system is segmented into multiple storage nodes, each containing both high-performance first storage devices and low-performance second storage devices. This segmentation allows different types of storage devices to be distributed across multiple nodes, enabling the system to utilize both high-performance devices for frequently accessed data and low-performance devices for less frequently accessed data, thus resolving the contradiction between data reading efficiency and storage space utilization.

Inventive Principle:
Principle #1Segmentation

2Quantity of substance

If data is stored only in low-performance storage devices, then storage space is saved, but data access speed deteriorates

Engineering Contradiction:
Improvestorage spaceVSAvoiddata access speed
Core Design Contradiction:
Quantity of substanceVSSpeed

Solution Approach 1:

The system applies local quality by storing frequently accessed data (hot data) in high-performance first storage devices and less frequently accessed data (cold data) in low-performance second storage devices. This differentiated storage strategy ensures that data requiring fast access is located on high-performance devices while maximizing storage capacity utilization through low-performance devices, thereby resolving the contradiction between storage space and data access speed.

Inventive Principle:
Principle #3Local quality

3Device complexity

If a single-node storage system is used, then system complexity is reduced, but performance utilization deteriorates

Engineering Contradiction:
Improvesystem complexityVSAvoidperformance utilization
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent merges multiple storage nodes into a unified storage system, where each node contributes both first and second storage devices to a shared memory pool. This merging allows the system to aggregate storage resources from multiple nodes, enabling better performance utilization through parallel access and load distribution while maintaining manageable system complexity through standardized node architecture and centralized management.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3958107B1Storage system, memory management method, and management node
Publication Date: 2025.09.10 HUAWEI TECH CO LTD
  • EP3958107B1 patent drawingFigure 1
  • EP3958107B1 patent drawingFigure 2
  • EP3958107B1 patent drawingFigure 3

AI summary

A storage system, a memory management method, and a management node are provided. The storage system includes a management node, one or more first storage devices, and one or more second storage devices. The management node is configured to create a memory pool to provide a service for storing data. Performance of the first storage device is higher than that of the second storage device, at least one of the one or more first storage devices is located in a first storage node, and at least one of the one or more second storage devices is located in a second storage node. The management node is further configured to control the data to be migrated between the first storage device and the second storage device in the memory pool. The memory pool is created based on storage devices with different performance, and these storage devices with the different performance are located in different storage nodes, thereby implementing a cross-node memory pool that integrates storage devices with different performance.