Heterogeneous Storage Management With Tree-Based Data Placement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current storage systems struggle to maximize I/O performance and reduce amplification in heterogeneous storage environments, often failing to exploit cumulative bandwidth and suffering from rigid data movement policies and high I/O amplification.

Innovation Solution

A heterogeneous storage manager provides a unified namespace that abstracts different storage devices, employing a tree data structure for fine-grained data placement and dynamic CPU allocation to optimize I/O requests across multiple devices, ensuring concurrent access and load balancing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional storage systems are used to manage big data, then data storage capacity is provided, but I/O performance is limited and bandwidth utilization is not maximized

Engineering Contradiction:
ImproveI/O performanceVSAvoidbandwidth utilization
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The storage system is segmented into multiple heterogeneous storage devices (e.g., SSDs, HDDs, memory devices) with different performance characteristics. The manager divides I/O requests and distributes them across these segmented storage devices based on their specific capabilities, allowing parallel access and maximizing cumulative bandwidth utilization.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system introduces a new dimension of heterogeneity by incorporating multiple types of storage devices with different performance characteristics (speed, capacity, cost) into a unified namespace. This multi-dimensional storage architecture allows the system to exploit cumulative bandwidth across diverse devices rather than being constrained to a single storage type.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If data is distributed across multiple storage devices, then bandwidth utilization improves, but system complexity increases

Engineering Contradiction:
Improvebandwidth utilizationVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

A storage manager acts as an intermediary layer between the application and the heterogeneous storage devices. This manager handles the complexity of device management, data placement, and I/O distribution transparently, allowing the application to access multiple storage devices through a unified namespace without dealing with the underlying complexity directly.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The storage manager provides a universal interface that handles multiple functions: managing heterogeneous devices, maintaining a unified namespace, distributing I/O requests, and handling data movement. This single multi-functional component manages the complexity rather than requiring separate mechanisms for each function.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Reliability

If rigid data movement policies are used in heterogeneous storage, then system control is maintained, but I/O amplification increases and performance decreases

Engineering Contradiction:
Improvesystem controlVSAvoidI/O amplification
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system employs dynamic data movement policies where the storage manager adaptively decides where to place and retrieve data based on current system state, performance metrics, and device availability. This dynamic approach replaces rigid policies, reducing unnecessary data movement and I/O amplification while maintaining system control through intelligent management.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The storage manager implements feedback mechanisms that monitor system performance, device status, and I/O patterns. Based on this feedback, the manager dynamically adjusts data placement and movement decisions to optimize performance and minimize I/O amplification, rather than following fixed predetermined policies.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4239462B1Systems and methods for heterogeneous storage systems
Publication Date: 2025.10.29 SAMSUNG ELECTRONICS CO LTD
  • EP4239462B1 patent drawingFigure 1
  • EP4239462B1 patent drawingFigure 2
  • EP4239462B1 patent drawingFigure 3

AI summary

Systems and methods for managing a storage system are disclosed. The storage system includes a first storage device and a second storage device different from the first storage device. A first storage operation is received for a first portion of a file, and a data structure associated with the file is identified. Based on the data structure, the first storage device is identified for the first portion of the file. The first storage operation is transmitted to the first storage device. In response to the first storage operation, the first storage device updates or accesses the first portion of the file.