Log-structured File System for Hybrid SSD HDD Storage

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional storage systems face inefficiencies in random access performance due to the manual identification and allocation of frequently accessed data between solid-state drives (SSDs) and hard disk drives (HDDs, requiring time-consuming administrative efforts and leading to suboptimal performance.

Innovation Solution

A hybrid media storage architecture with a log-structured file system that automatically performs initial placement and migration of data among different storage media, using heuristics and policies to optimize data placement and allocation between SSDs and HDDs, improving performance characteristics without manual enforcement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If manual identification and isolation of hot data is performed to store frequently accessed data in SSDs, then random access performance is improved, but administrative time and operational complexity increase

Engineering Contradiction:
Improverandom access performanceVSAvoidadministrative time
Core Design Contradiction:
SpeedVSLoss of time

Solution Approach 1:

The storage system automatically identifies hot data and manages data placement between SSDs and HDDs without requiring administrator intervention. The system self-monitors access patterns and autonomously migrates data to optimize performance, eliminating the need for manual hot data identification and isolation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system implements continuous monitoring of data access patterns and uses this feedback to dynamically adjust data placement. By tracking read/write operations and access frequency, the system automatically identifies hot data and redistributes it between storage media to maintain optimal random access performance without manual intervention.

Inventive Principle:
Principle #23Feedback

2Device complexity

If fixed allocations of data are implemented with static mapping between volume block numbers and storage media, then data organization is simplified, but adaptability to changing access patterns deteriorates

Engineering Contradiction:
Improvedata organization complexityVSAvoidadaptability to access patterns
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The system employs dynamic data placement that automatically adapts to changing access patterns. Instead of static mappings, the system continuously monitors data access behavior and dynamically redistributes data between SSDs and HDDs. This allows the storage architecture to respond to evolving workloads and access patterns while maintaining simple volume management interfaces.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system segments data into hot and cold portions based on access patterns and places them in appropriate storage media. This segmentation is dynamically adjusted over time, allowing frequently accessed data segments to reside on SSDs while less active data moves to HDDs, optimizing both performance and adaptability without increasing organizational complexity.

Inventive Principle:
Principle #1Segmentation

3Speed

If changes are made to fixed data allocations between SSDs and HDDs, then performance can be optimized, but implementation time and cost increase

Engineering Contradiction:
Improvestorage performanceVSAvoidimplementation time
Core Design Contradiction:
SpeedVSLoss of time

Solution Approach 1:

The storage system automatically optimizes data placement by monitoring access patterns and autonomously migrating data between SSDs and HDDs. This eliminates the need for administrators to manually reconfigure data allocations when performance optimization is needed, significantly reducing implementation time and operational costs while maintaining optimal storage performance.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system implements periodic analysis of data access patterns and automatically triggers data migration operations when optimization opportunities are detected. This periodic self-optimization allows the system to continuously improve performance without requiring manual intervention or costly reconfiguration projects.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentEP2263145B1Hybrid media storage system architecture
Publication Date: 2020.02.05 NETAPP INC
  • EP2263145B1 patent drawingFigure 1
  • EP2263145B1 patent drawingFigure 2
  • EP2263145B1 patent drawingFigure 3

AI summary

A hybrid media storage architecture has a log-structured file system configured to control a plurality of different storage media organized as hybrid storage media that cooperate to provide a total storage space of a storage system. The log-structured file system is configured to perform initial placement and migration of data, as well as fine-grain write allocation of the data, among storage space locations of the hybrid storage media to thereby improve the performance characteristics of the media. By defining and implementing heuristics and policies directed to, e.g., types of data, the file system may initially place data on any of the different media and thereafter migrate data between the media at fine granularity and without the need for manual enforcement.