SSD Storage Management via Sub-Space Release to Reduce WAF

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In all-flash deduplication systems, the high write amplification factor (WAF) leads to slowed input/output operations and reduced SSD service life due to complicated chunk data reference relationships, causing invalid pages and increased WAF, which traditional garbage collection methods fail to effectively address.

Innovation Solution

The method involves determining the data volume of failed data in sub-storage spaces and selectively releasing target sub-storage spaces based on this data volume, delaying the release process to minimize unnecessary data movement and reduce WAF, thereby prolonging SSD life.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If traditional garbage collection methods are used to release storage space, then storage space can be recovered, but the write amplification factor (WAF) increases due to invalid pages caused by complicated chunk data reference relationships

Engineering Contradiction:
Improvestorage spaceVSAvoidwrite amplification factor
Core Design Contradiction:
Quantity of substanceVSObject-generated harmful factors

Solution Approach 1:

The patent divides the storage space into multiple sub-storage spaces based on SSD block structures. By segmenting the garbage collection process at the sub-storage space level rather than treating the entire storage space uniformly, the system can more precisely identify and release only those sub-storage spaces where all data is invalid, thereby avoiding creating invalid pages in blocks that still contain valid data references.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different treatment to different sub-storage spaces based on their local characteristics. Each sub-storage space is evaluated independently for its data validity status, and release decisions are made locally for each sub-storage space rather than globally. This local quality approach ensures that only sub-storage spaces with completely invalid data are released, minimizing the creation of invalid pages while still recovering necessary storage space.

Inventive Principle:
Principle #3Local quality

2Productivity

If storage space is released frequently to maintain available capacity, then storage availability improves, but SSD service life decreases due to increased WAF

Engineering Contradiction:
Improvestorage availabilityVSAvoidSSD service life
Core Design Contradiction:
ProductivityVSDuration of action of stationary object

Solution Approach 1:

The patent performs preliminary evaluation of data validity in each sub-storage space before initiating the release process. By determining in advance whether all data in a sub-storage space is invalid through checking chunk data reference relationships, the system can make informed release decisions that avoid unnecessary data movement and minimize WAF, thus preserving SSD service life while maintaining storage availability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements a feedback mechanism that monitors the WAF and storage space status. The garbage collection process continuously evaluates the impact of release operations on WAF and adjusts its behavior accordingly. This feedback loop ensures that storage space is released only when necessary and when it will not significantly increase WAF, thereby balancing storage availability with SSD service life extension.

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS11567671B2Method, electronic device, and computer program product for storage management
Publication Date: 2023.01.31 EMC IP HLDG CO LLC
  • US11567671B2 patent drawing
  • US11567671B2 patent drawing
  • US11567671B2 patent drawing

AI summary

Embodiments of the present disclosure relate to a method, an electronic device, and a computer program product for storage management. The method for storage management includes determining the data volume of corresponding failed data in a plurality of candidate sub-storage spaces of a disk, the plurality of candidate sub-storage spaces being formed by dividing the original storage space of the disk and having consecutive physical addresses, and the size of one candidate sub-storage space being associated with the size of the storage space of at least one block that constitutes the disk; and selecting a target sub-storage space to be released from the plurality of candidate sub-storage spaces based at least on the data volume of the corresponding failed data. As such, since the operation of moving valid data in an SSD block to a new SSD block is reduced, the WAF of a storage system, especially an SSD, can be effectively reduced, and the service life of the SSD can be prolonged.