Mapping method for decreasing solid state disk metadata refreshing frequency

Abstract

The invention provides a mapping method for decreasing the solid state disk metadata refreshing frequency so as to decrease the metadata refreshing frequency. The method comprises the steps that a caching mapping table is divided into an increment caching mapping table and a continuous caching mapping table, all updated maps which are not refreshed to Nand Flash are saved in the increment cachingmapping table; if the mapping amount in the increment caching mapping table in the read-write process reaches a preset threshold value, the operation for refreshing the continuous caching mapping table is triggered, if the phenomenon that only a small number of increment mapping tables are saved due to abnormal power failure happens, the overall mapping table is updated after power is on, and rebuilding is completed. The table refreshing frequency in the read-write process is lowered, the metadata refreshing quantity during abnormal failure is decreased, and meanwhile the requirement for the rebuilding time of a solid state disk is met.

Description

technical field [0001] The invention relates to a data processing method of a hard disk, in particular to a mapping method for reducing the brushback frequency of metadata of a solid-state hard disk. Background technique [0002] SSDs need FTL (Flash Translation Layer) algorithm for address translation when reading and writing data. Usually, FTL algorithms can use three mapping methods: block level, page level, and mixed according to the minimum mapping unit. Among them, the random data read and write performance of page-level mapping is the best, so most of the SSDs currently use this mapping method. However, the disadvantage of page-level mapping is that the generated address mapping information table, i.e. metadata, is far more than other methods of mapping, and in order to ensure that the data is not lost when the SSD is hot-swapped, the FTL algorithm uses a flashback mechanism to update the metadata A global backup is made in the flash memory, but the flashback frequen...

AI Technical Summary

Problems solved by technology

However, the disadvantage of page-level mapping is that the generated address mapping information table, i.e. metadata, is far more than other methods of mapping, and in order to ensure that the data is not lost when the SSD is hot-swapped, the FTL algorithm uses a flashback mechanism to update the metadata A global backup is made in the flash memory, but the flashback frequency will directly affect the overall read and write performance of the SSD and the power-on and rebuild time

Images