A method, device, device and medium for processing large block data of solid state hard disk

Abstract

The invention discloses a method, device, equipment and medium for writing large block data of a solid-state hard disk. The method includes: screening large-block data write requests from write requests sent by the host; using a DM module to split the large-block data to be written corresponding to the large-block data write requests into multiple small-block data and mount them on a linked list; The DM module sends a write request to the LKM module, and the content of the write request includes the LBA first address of the large block of data to be written and the total number of small blocks of data; the LKM module receives the write request and performs cache allocation to generate several write request results Then return to the DM module; the DM module retrieves the context content of each small piece of data from the mounted linked list in units of small pieces of data according to the received several request results and initiates a data transfer request to the DMA module; in response to the DMA module completing the data If it is moved, the DM module resources and the LKM module resources are released. The scheme of the invention reduces the number of communication between modules, optimizes the efficiency of large block writing, reduces the writing delay, and increases the writing bandwidth.

Description

technical field [0001] The present invention relates to the technical field of storage, and in particular, to a method, device, device and medium for processing large block data of a solid-state hard disk. Background technique [0002] ZNS (Zoned Namespace for short) is a solution based on the NVME protocol. Compared with the traditional SSD (Solid State Drive, solid state drive or solid state drive), it has the characteristics of small write amplification and low cost. Generally, ZNSSSD only supports sequential writing. Therefore, before configuring and using ZNS SSD, you need to obtain the configuration information of all zones in the SSD, and then accurately determine the write pointer and status of each zone. For example, ZNS 128KB sequential writing generally needs to go through the process of unpacking, applying for write buffer (buff), DMA (direct memory read unit) data movement, writing completion, etc., and IPC (Inter Process Communication) is required between each ...

AI Technical Summary

Problems solved by technology

[0004] At present, the process of ZNS 128KB sequential writing is as follows: when the ZNS SSD firmware receives a 128KM write request, it first needs the DM module to unpack into 32 4KB pieces, and each 4KB sends a write request to the LKM module, and the LKM module returns the write cache (buff) address, the DM module sends a data move request to the DMA module, the DMA module completes the move, and then the DM module sends the 4KB writing completion to the LKM module, and finally the LKM module releases the buff, and this is repeated 32 times; it can be seen that the existing 128KB sequential write In the process, multiple write applications are required and repeated communication between modules makes writing inefficient and takes up a large bandwidth. Therefore, the existing large-block data writing process needs to be improved urgently.

Images