SSD garbage collection method and device, computer equipment and storage medium

Abstract

The invention relates to an SSD garbage collection method and device, computer equipment and a storage medium. The method comprises the following steps: selecting a sacrificial block, and reading to-be-moved data; judging whether the to-be-migrated data is cold data or hot data; if the data is the cold data, sorting the blocks according to the erasure times, forming a maximum moving group by all blocks of which the erasure time difference with the current maximum erasure time block is less than K, selecting the block with the longest operation time in the maximum moving group as a target block, erasing the target block, and moving the cold data to be moved to the target block; and if the hot data is the hot data, sorting the blocks according to the erasure times, forming a moving minimum group by all the blocks of which the erasure time difference with the current minimum erasure time block is less than M, selecting the block with the shortest operation time in the moving minimum groupas a target block, erasing the target block, and moving the hot data to be moved to the target block. The overall service life of the SSD can be effectively prolonged, and the performance of the SSDcan be effectively improved.

Description

technical field [0001] The present invention relates to the technical field of solid-state disk garbage collection, and more specifically refers to a method, device, computer equipment and storage medium for SSD garbage collection. Background technique [0002] There is an existing target block selection strategy for SSD (solid state drive) garbage collection (GC). The usual method is to manage the target block according to the number of erasures (EC) of the free block. When the GC operation applies for the target block, the current EC is given priority. The smallest block is used as the target block; the disadvantage of this method is that the reference factors for target block selection are too single, and different types of data are not specially processed, such as "cold data" for GC operations, and the applied target block is the current EC The smallest block, and "cold data" may not be accessed for a long period of time in the future, which causes a waste of resources a...

AI Technical Summary

Problems solved by technology

[0002] There is an existing target block selection strategy for SSD (solid state drive) garbage collection (GC). The usual method is to manage the target block according to the number of erasures (EC) of the free block. When the GC operation applies for the target block, the current EC is given priority. The smallest block is used as the target block; the disadvantage of this method is that the reference factors for target block selection are too single, and different types of data are not specially processed, such as "cold data" for GC operations, and the applied target block is the current EC The smallest block, and "cold data" may not be accessed for a long period of time in the future, which causes a waste of resources and affects the overall service life of the SSD. In addition, this method does not fully consider the physical properties of NAND Features, the operation time (reading, writing and erasing) of each block (Block) is different. Compared with the allocation of "hot data" to the target block with a long operation time and the target block with a short operation time, in the long run, the allocation of "hot data" A target block with a short operation time will inevitably improve the overall performance of the SSD; therefore, it cannot meet the demand

Images