SSD Address Grouping Across Access Units to Cut Write Amplification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The use of larger access units in solid state drives (SSDs) leads to increased write amplification and higher costs due to the need for larger caches to manage address mapping tables, while existing solutions fail to effectively address these issues when dealing with smaller data blocks.
Innovation Solution
A data processing method that involves storing logical addresses as first access units in a first storage pool and combining them into logic address groups when a quantity requirement is met, transferring these groups to a second storage pool as second access units, thereby managing data using larger access units to reduce write amplification and cache requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If a Large Access Unit (Large AU) is used in the solid state drive, then the quantity of secondary mapping tables managed by FTL is reduced, but the Write Amplification (WA) increases when the Large AU is read and rewritten
Solution Approach 1:
The patent divides the storage pool into multiple segments (first storage pool and second storage pool) with different access unit sizes. The first storage pool uses small access units (4KB) while the second storage pool uses large access units (256KB). This segmentation allows the system to reduce the quantity of mapping tables by utilizing large access units in the second pool, while avoiding write amplification by handling small writes in the first pool, thus resolving the contradiction between reducing mapping table quantity and preventing write amplification.
2Loss of energy
If a cache is used to store sequential streams for Large AU, then the Write Amplification is addressed, but the hardware cost increases due to larger cache requirements
Solution Approach 1:
The patent introduces a new dimension to the storage architecture by creating a two-tier storage pool system with different access unit sizes. Instead of using a single cache to store sequential streams, the system uses the first storage pool (with small access units) as a buffer layer. This dimensional change allows the system to manage write operations without requiring large cache space, as the first pool handles small writes while the second pool handles large writes, thereby reducing hardware cost while addressing write amplification.
3Loss of energy
If small access unit size is used for data writing, then the Write Amplification is reduced, but the cache space requirement increases to store sequential streams
Solution Approach 1:
The patent segments the storage system into two pools: the first storage pool with small access units (4KB) that handles sequential writes with minimal cache requirements, and the second storage pool with large access units (256KB) that reduces mapping table quantity. By segmenting the storage pools, the system achieves low write amplification through small access units in the first pool without requiring excessive cache space, as each pool operates independently with its own access unit size optimized for its function.
Data Source
AI summary
Embodiments of the present disclosure provide a data processing method and system, a device, a storage system, and a medium. The data processing method includes: receiving a logic address corresponding to data to be processed; storing the logic address as a first access unit in a first storage pool; and in a case that the quantity of the stored first access units in the first storage pool meets a requirement, combining a plurality of first access units into a logic address group and storing the logic address group as a second access unit in a second storage pool so as to process the data based on the second access unit.