Optimization method for read performance of storage system based on charge-trapping 3d TLC flash memory

Abstract

The invention discloses a read performance optimization method of a storage system based on a charge trapping type 3D TLC flash memory, wherein the method includes the steps of obtaining all physicalpages storing all or part of requested data for any read request issued by upper layer file system, obtaining a physical page set, and adding 1 to the the number of read times of each physical page inthe physical page set; for any physical page in the physical page collection, if it is a valid page and the reading times are greater than a preset reading times threshold value, and there is no middle page corresponding to the physical page, regarding the physical page as an original physical page, and an idle physical page is determined as a target page in the flash memory, and copying the datain the original physical page into the target page; marking the original physical page as a middle page and mark the target page as a valid page; ending the optimization processing for the current read request. The invention can improve the parallelism of the read operation, thereby improving the read performance of the whole storage system.

Description

technical field [0001] The invention belongs to the field of computer storage, and more specifically relates to a method for optimizing read performance of a storage system based on a charge-trapping 3D TLC flash memory. Background technique [0002] Compared with the traditional storage system based on 2D flash memory, the storage system based on Charge Trap (CT) 3D TLC flash memory has the advantages of high storage density and strong scalability, so it has been increasingly widely used. In the charge-trapping 3D TLC, there are three pages on a word line, the low page, the middle page, and the high page. When performing a write operation, the three pages on a word line are simultaneously written. When performing a read operation, the three pages Read separately. The study found that the speed of reading operations on different types of pages is different, specifically, the read latency of the low page and high page is significantly lower than that of the middle page. [...

AI Technical Summary

Problems solved by technology

[0003] At present, in some dedicated storage systems that have high requirements on the response delay of read requests, there is a certain bottleneck in the optimization of read performance.

On the other hand, there are irregular garbage collection operations in flash memory-based storage systems. This operation takes a long time because it involves a large number of flash memory erasures and data migration. Blocking and large response delay

Images