Host-Assisted Sub-Page Garbage Collection for SSDs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solid-state mass storage devices face inefficiencies in garbage collection, particularly when dealing with data structures smaller than a page, as they cannot effectively identify and manage invalid sub-page data, leading to reduced storage capacity and performance.
Innovation Solution
A host-assisted garbage collection system that allows the host computer system to analyze and manage data at a sub-page level, identifying and removing invalid data structures, reorganizing valid data into new structures, and writing them back to the solid-state mass storage device, thereby enabling more efficient use of storage space and reducing write amplification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional page-based garbage collection is used, then the storage device can manage data at the page level, but it cannot effectively identify and manage invalid sub-page data structures, leading to reduced storage capacity and performance
Solution Approach 1:
The patent segments the data validity detection process into two levels: the storage device performs page-level segmentation (reading and identifying candidate pages), while the host computer performs sub-page level segmentation (analyzing data structures within pages to identify invalid portions). This segmentation allows precise identification of invalid sub-page data without requiring the storage device to handle complex sub-page analysis, thus improving detection precision while maintaining garbage collection efficiency.
Solution Approach 2:
The host computer acts as an intermediary between the storage device and the garbage collection process. The storage device provides page-level information to the host, which then performs sophisticated sub-page level validity analysis using application knowledge. This intermediary approach enables precise invalid data identification without burdening the storage device with complex processing, resolving the contradiction between detection precision and collection efficiency.
2Reliability
If the storage device performs full garbage collection on pages containing mixed valid and invalid data structures, then all invalid data can be removed, but valid data must be rewritten unnecessarily, increasing write amplification
Solution Approach 1:
The patent extracts only the invalid data structures from pages during the garbage collection process. The host computer analyzes each page to identify specific invalid data structures and extracts only those portions for removal. This extraction approach allows the system to remove invalid data while preserving valid data structures, eliminating unnecessary rewrites and reducing write amplification while maintaining data integrity.
Solution Approach 2:
The patent applies local quality by treating different portions of a page differently based on their validity status. Instead of treating entire pages as uniformly valid or invalid, the system identifies and processes only the specific sub-page regions containing invalid data structures. This localized approach ensures that valid data structures are preserved and not unnecessarily rewritten, reducing write amplification while maintaining data integrity.
3Productivity
If the storage device uses host-assisted garbage collection with sub-page level analysis, then it can improve storage efficiency and reduce write amplification, but it requires increased interaction and coordination between the host and device
Solution Approach 1:
The patent applies preliminary action by having the storage device perform initial page-level identification of candidate pages containing invalid data structures before transferring them to the host computer. This preliminary filtering reduces the amount of data the host needs to process, simplifying the coordination required. The host then focuses only on analyzing pages already identified as candidates, rather than processing all pages, thus improving storage efficiency while managing coordination complexity.
Solution Approach 2:
The patent uses partial action by implementing a two-stage approach where the storage device performs partial garbage collection (identifying and reading candidate pages) and the host performs the remaining partial analysis (sub-page level validity detection). This division of labor allows the system to achieve improved storage efficiency through sub-page level analysis without requiring either component to perform the complete, complex task alone, thus managing overall system complexity.
Data Source
AI summary
Systems and methods suitable for assisting data storage on a solid-state mass storage device by furthering interaction between a host and the solid-state mass storage device. The method includes providing data in a target block from a solid-state mass storage device to a host computer system, analyzing the data to identify valid and invalid data of the data in the target block with the host computer system, removing the invalid data, reformatting the valid data into a new data structure with the host computer system, writing the new data structure to the mass storage device, and marking the data in the target block as invalid.