LRU flash memory cache management method based on dynamic page weight

Abstract

The invention discloses an LRU flash memory cache management method based on dynamic page weight. The method comprises: S1, page requests in a request queue are read, and page request types and areaswhere the page requests are located are identified and classified; and S2, judging the state of inserting the buffer area, judging the obsolete page by using an LRU method based on the dynamic page weight, adjusting the state of the buffer area, and executing a page request. According to the technical scheme, the buffer area type is divided into the working area and the exchange area, the cold page, the hot page, the dirty page and the clean page are distinguished, the page request type and the buffer area are determined, and then the obsoleted page is judged to complete the page request by combining the LRU method based on the dynamic page weight. According to the technical scheme, read-write consumption and time delay in the flash memory cache read-write process can be effectively reduced, and meanwhile the hit rate in flash memory cache read-write can be greatly increased.

Description

technical field [0001] The invention relates to the field of storage systems, in particular to a dynamic page weight-based LRU flash cache management method. Background technique [0002] NAND flash storage technology has been widely used in enterprise applications due to its high performance, small size, and low energy consumption. However, with the continuous development of big data technology in recent years, the processing and analysis of massive data has put forward higher requirements for the data throughput and I / O delay of the storage system. Defects such as symmetric I / O delay and block erasure make it impossible to completely replace hard disk storage. [0003] Combining caching technology with storage devices can effectively reduce I / O latency and reduce asymmetry in different storage layers. However, most of the current cache management methods are optimized for hard disk storage devices, and there is a lack of a cache area management method for flash memory. ...

AI Technical Summary

Problems solved by technology

However, most of the current cache management methods are optimized for hard disk storage devices, and there is a lack of cache area management methods for flash memory.

When the traditional cache management method is applied to flash memory, there is a problem of low hit rate, and the above-mentioned problems of high latency and high write consumption cannot be solved.

Images