Master System Classifies Write Data for Distributed Storage

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current memory systems face challenges in efficiently distributing and managing write data across multiple nonvolatile memory devices, particularly in ensuring transaction integrity and handling sudden power-offs, which complicates data recovery and storage operations.

Innovation Solution

A data processing system that includes a master system and multiple nonvolatile memory devices, where the master system classifies and distributes write data into transactions, determines device roles based on capacity and overwrite capabilities, and performs specific storage and recovery operations to ensure data integrity and efficient distribution across the devices.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If write data is stored in a single memory device, then data integrity is easier to manage, but storage capacity and redundancy are limited

Engineering Contradiction:
Improvedata integrityVSAvoidstorage capacity
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent divides the storage system into multiple memory devices (first nonvolatile memory device, second nonvolatile memory device, and third nonvolatile memory device) and segments write data into different categories (transactional write data and non-transactional write data). Transactional write data is distributed across multiple devices to enable redundancy and capacity expansion, while non-transactional write data is stored in a dedicated device. This segmentation allows the system to achieve both data integrity through distribution and enhanced storage capacity.

Inventive Principle:
Principle #1Segmentation

2Quantity of substance

If transactional write data is distributed across multiple memory devices, then storage capacity and redundancy are improved, but data management complexity increases

Engineering Contradiction:
Improvestorage capacityVSAvoiddata management complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent introduces a master system as an intermediary component that receives write data from the host, classifies it into transactional and non-transactional categories, and manages the distribution to appropriate slave systems. The master system maintains a mapping relationship between transactional write data and corresponding slave systems, simplifying the data management process. This intermediary structure reduces the complexity that would otherwise arise from directly managing distributed storage across multiple devices.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If all write data is stored in nonvolatile memory devices, then data persistence is improved, but handling of sudden power-offs and data recovery becomes more complex

Engineering Contradiction:
Improvedata persistenceVSAvoidpower-off recovery complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments write data into transactional and non-transactional categories, storing non-transactional write data in a dedicated third nonvolatile memory device that is not subject to overwriting restrictions. This segmentation allows the system to maintain data persistence while simplifying recovery operations, as non-transactional data provides a stable baseline that can be used for recovery without the complexity of managing transactional states across multiple devices.

Inventive Principle:
Principle #1Segmentation

4Adaptability or versatility

If memory devices are classified based on overwrite capability, then data storage flexibility is improved, but system configuration complexity increases

Engineering Contradiction:
Improvedata storage flexibilityVSAvoidsystem configuration complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The master system serves as an intermediary that handles the classification of memory devices based on their overwrite capabilities. It receives information about device characteristics (such as whether a device can be overwritten) and automatically determines the appropriate slave system assignments. This intermediary approach abstracts the configuration complexity from the data storage operations, allowing the system to maintain flexibility in device selection while keeping the configuration process centralized and simplified.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS11586382B2Memory system and data processing system for distributedly storing input data
Publication Date: 2023.02.21 MIMIRIP LLC
  • US11586382B2 patent drawing
  • US11586382B2 patent drawing
  • US11586382B2 patent drawing

AI summary

A data processing system includes first memory system including a first nonvolatile memory device; a second memory system including a second nonvolatile memory device; and a master system including a third nonvolatile memory device. The master system classifies any one of the first memory system and the second memory system as a first slave system and the other as a second slave system depending on a predetermined reference, wherein the master system is coupled to a host, and includes a write buffer for temporarily storing a plurality of write data, and wherein the master system classifies the write data, into first write data grouped into a transaction and second write data which are not grouped into the transaction, stores the second write data in the third nonvolatile memory device, and stores the first write data in the first nonvolatile memory device or the second nonvolatile memory device.