Dynamic Address Mapping for SSD Failure Recovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In high-capacity SSDs with multiple sub-storage devices, simple address mapping is ineffective in addressing failures due to performance degradation, as the burden of a failing flash memory device is not efficiently distributed among other devices.

Innovation Solution

An address mapping method that uses a hash function to dynamically distribute logical addresses across multiple sub-storage devices, including over-provision areas, to prevent performance degradation by redistributing the workload and storing data in underutilized areas.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If simple address mapping is used in high-capacity SSDs with multiple sub-storage devices, then the address mapping process is simple and fast, but performance degradation occurs when flash memory devices fail because the workload is not efficiently distributed

Engineering Contradiction:

Improveaddress mapping efficiencyVSAvoidperformance stability under failure conditions

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent implements dynamic address mapping that adapts based on the operational status of sub-storage devices. When failures are detected, the system dynamically redistributes logical addresses from failed devices to healthy ones using hash functions, ensuring the mapping strategy changes in response to system conditions rather than remaining static

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the mapping parameters by switching between different hash functions based on the distribution of over-provision areas across sub-storage devices. This parameter adjustment allows optimal workload distribution to be maintained even when device capacity or status changes due to failures

Inventive Principle:

Principle #35Parameter changes

2Ease of manufacture

If logical addresses from failed flash memory devices are remapped to a single sub-storage device, then the remapping process is simple, but performance degradation concentrates on that single device

Engineering Contradiction:

Improveremapping implementation simplicityVSAvoidoverall SSD performance

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The patent segments the remapping process by dividing the set of logical addresses to be remapped across multiple target sub-storage devices. Instead of concentrating all remapped addresses in one device, the system uses hash functions to distribute them across multiple devices, with each device handling a portion of the remapping workload

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces hash functions as intermediary mechanisms that mediate between logical addresses and physical sub-storage devices. These hash functions act as intermediaries that systematically distribute addresses across multiple devices, preventing direct concentration of workload on any single device while maintaining systematic remapping

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If over-provision areas are used to store remapped data, then data redundancy and reliability are improved, but the capacity of over-provision areas may be insufficient when multiple devices fail

Engineering Contradiction:

Improvedata redundancy capabilityVSAvoidavailable over-provision capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent makes over-provision areas universal by allowing them to serve multiple functions: storing remapped data from failed devices, providing spare capacity for future failures, and distributing workload across multiple devices. The same over-provision areas can be dynamically allocated to different logical addresses depending on which devices are healthy, maximizing their utility

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system creates a composite storage structure by combining user data areas and over-provision areas into a unified addressable space. Through the hash function-based mapping, logical addresses can map to either user data areas or over-provision areas depending on device status, creating a composite structure that flexibly uses all available capacity

Inventive Principle:

Principle #40Composite materials

Data Source

PatentUS11513971B2Address mapping method and operation method of storage device

Publication Date: 2022.11.29 SAMSUNG ELECTRONICS CO LTD

AI summary

An address mapping method of a storage device which includes a plurality of sub-storage devices each including an over-provision area includes detecting mapping information of a received logical address from a mapping table, selecting a hash function corresponding to the received logical address depending on the mapping information, selecting any one, which is to be mapped onto the received logical address, of the plurality of sub-storage devices by using the selected hash function, and mapping the received logical address onto the over-provision area of the selected sub-storage device. The selected hash function is selected from a default hash function and a plurality of hash functions to provide a rule for selecting the any one of the plurality of sub-storage devices.