A delay-based solid state disk data storage method

Abstract

The application discloses a solid state disk data storage method based on time delay, which comprises the following steps: S1, dividing the storage space of the solid state disk into multiple storage areas, testing each storage area, and obtaining the data writing speed of each storage area and the influence coefficient of the data reading work; S2, when it is needed to write the to-be-stored data into the solid state disk, judging whether there is an area for data reading in the solid state disk, and calculating the real data writing speed of each storage area according to the judgment result; S3, based on the time delay requirement of the to-be-stored data, the real data writing speed of each storage area and the remaining storage space, then selecting the storage area for the parallel writing of the to-be-stored data, and ensuring that the data writing is completed at the same time. The application considers the real data writing speed of different storage areas of the solid state disk under the influence of data reading, and obtains an efficient data storage scheme under the condition of guaranteeing the satisfaction of the data writing time delay.

Description

Technical Field

[0001] This invention relates to the field of data storage, and in particular to a time-delay-based solid-state drive data storage method. Background Technology

[0002] Solid-state drives (SSDs) are hard drives made using arrays of solid-state electronic storage chips. Different areas of an SSD can be written to simultaneously. This is because the SSD controller can manage multiple flash memory channels, enabling parallel operations. When high data write latency is required, parallel data writing is often necessary. However, the write speed of an SSD is often affected by data read speeds, making it difficult to plan parallel data writing effectively. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a latency-based solid-state drive data storage method. This method comprehensively considers the actual writing speed of different storage areas of the solid-state drive under the influence of data reading. Under the condition of ensuring that the data writing latency is met, the method reasonably divides the data to be stored into blocks and selects storage areas to obtain an efficient data storage solution.

[0004] The objective of this invention is achieved through the following technical solution: a time-delay-based solid-state drive data storage method, comprising the following steps:

[0005] S1. Divide the solid-state drive's storage space into multiple storage areas, test each storage area, and obtain the data write speed of each storage area and the impact coefficient of data read operations;

[0006] S2. When it is necessary to write data to be stored into the solid-state drive, determine whether there is a region in the solid-state drive where data is being read, and calculate the actual data writing speed of each storage region based on the determination result;

[0007] S3. Based on the latency requirements of the data to be stored, the actual writing speed of data in each storage area, and the remaining storage space, the selected storage areas are used for parallel writing of the data to be stored, ensuring that the data writing is completed simultaneously.

[0008] The beneficial effects of the present invention are: (1) After dividing the solid-state drive storage area, the present invention comprehensively considers the impact of data reading on each storage area and obtains a more accurate real data writing speed, providing a basis for the selection of storage areas;

[0009] (2) Under the condition of meeting the storage delay requirements of the data to be stored, the present invention determines the number of initially selected storage areas, and selects reasonable storage areas based on the actual data writing speed and the remaining storage space of each storage area, and divides the data blocks so that the divided data blocks can complete the data writing simultaneously in the selected storage areas, thereby improving the efficiency of data writing and ensuring the high efficiency of data storage of the data to be stored. Attached Figure Description

[0010] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0011] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0012] like Figure 1 As shown, a time-delay-based solid-state drive data storage method is characterized by the following steps:

[0013] S1. Divide the solid-state drive's storage space into multiple storage areas, test each storage area, and obtain the data write speed of each storage area and the impact coefficient of data read operations;

[0014] Step S1 includes:

[0015] S101. Divide the storage space of the solid-state drive into multiple storage areas;

[0016] S102. For any given storage area, the impact coefficients on data write speed and data read operation are tested:

[0017] When no data is read from any of the storage areas of the solid-state drive, the data write speed of the current storage area is tested: a fixed amount of data is written to the current storage area, and the first delay from the start of writing data to the completion of writing data is recorded. Then, the fixed amount of data is divided by the first delay to obtain the first data write speed.

[0018] When no data is being read in the current storage area, but there are storage areas on the SSD where data is being read, a data write speed test is performed on the current storage area: a fixed amount of data is written to the current storage area, and the second delay from the start of writing data to the completion of writing data is recorded. Then, the fixed amount of data is divided by the second delay to obtain the second data write speed.

[0019] While data is being read from the current storage area, a data write speed test is performed on the current storage area: a fixed amount of data is written to the current storage area, and the third delay from the start of writing data to the completion of writing data is recorded. Then, the fixed amount of data is divided by the third delay to obtain the third data write speed.

[0020] S103. Divide the second data write speed by the first data write speed to obtain the first influence coefficient, which is used to characterize the influence of data reading in other storage areas on the data write speed of the current storage area;

[0021] The second influence coefficient is obtained by dividing the third data write speed by the first data write speed, which is used to characterize the impact of data reading in the current storage area on the data write speed in the current storage area.

[0022] S104. For each storage region, repeat steps S102 to S103.

[0023] S2. When it is necessary to write data to be stored into the solid-state drive, determine whether there is a region in the solid-state drive where data is being read, and calculate the actual data writing speed of each storage region based on the determination result;

[0024] Step S2 includes:

[0025] S201. When it is necessary to write data to the solid-state drive (SSD), determine whether there is a region on the SSD where data is currently being read:

[0026] If it does not exist, the first data write speed of each storage area is the actual data write speed;

[0027] If it exists, proceed to step S202;

[0028] S202. For any storage area, determine whether the area where data is being read is the current storage area or another storage area of ​​the solid-state drive;

[0029] If the area where data is being read is another storage area of ​​the solid-state drive, the actual data write speed of the current storage area is obtained by multiplying the first data write speed of the current storage area by the first influence coefficient.

[0030] If the area where data is being read is the current storage area, then the actual data write speed of the current storage area is obtained by multiplying the first data write speed of the current storage area by the second influence coefficient.

[0031] S3. Based on the latency requirements of the data to be stored, the actual writing speed of data in each storage area, and the remaining storage space, the selected storage areas are used for parallel writing of the data to be stored, ensuring that the data writing is completed simultaneously.

[0032] Step S3 includes:

[0033] S301. Calculate the remaining storage space of each storage area, and remove storage areas with remaining storage space less than a preset threshold. The remaining storage areas are used as candidate storage areas. The remaining candidate storage areas are sorted from largest to smallest according to the actual data writing speed.

[0034] S302. Determine the number M of storage regions selected during initialization based on the storage latency requirements of the data to be stored;

[0035] Assume the size of the data to be stored is... The storage latency of the data to be stored must not exceed [a certain value]. Then the number M of the initial selected storage regions is:

[0036] ;

[0037] S303. From the storage areas, select M storage areas in descending order of their actual data write speeds. The actual data write speeds of these M storage areas are denoted as... ,in This represents the actual speed at which data is written to the i-th storage region out of M storage regions, where i = 1, 2, ..., M;

[0038] Calculate the time required to write data to M storage areas in parallel. ;

[0039] S304. Divide the data to be stored into M data blocks, with the sizes of the data blocks being respectively... , where the size is The data blocks are allocated to the i-th storage region among M storage regions, i=1,2,…,M;

[0040] S305. Determine if any of the M storage regions has remaining storage space smaller than the allocated data block size:

[0041] If it exists, increment the value of M by 1 and return to step S303;

[0042] If it does not exist, according to the division in step S304, each data block is written in parallel to the corresponding storage area.

[0043] The foregoing description illustrates and describes a preferred embodiment of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A time-delay-based solid-state drive data storage method, characterized in that: Includes the following steps: S1. Divide the solid-state drive's storage space into multiple storage areas, test each storage area, and obtain the data write speed of each storage area and the impact coefficient of data read operations; S101. Divide the storage space of the solid-state drive into multiple storage areas; S102. For any given storage area, the impact coefficients on data write speed and data read operation are tested: When no data is read from any of the storage areas of the solid-state drive, the data write speed of the current storage area is tested: a fixed amount of data is written to the current storage area, and the first delay from the start of writing data to the completion of writing data is recorded. Then, the fixed amount of data is divided by the first delay to obtain the first data write speed. When no data is being read in the current storage area, but there are storage areas on the SSD where data is being read, a data write speed test is performed on the current storage area: a fixed amount of data is written to the current storage area, and the second delay from the start of writing data to the completion of writing data is recorded. Then, the fixed amount of data is divided by the second delay to obtain the second data write speed. While data is being read from the current storage area, a data write speed test is performed on the current storage area: a fixed amount of data is written to the current storage area, and the third delay from the start of writing data to the completion of writing data is recorded. Then, the fixed amount of data is divided by the third delay to obtain the third data write speed. S103. Divide the second data write speed by the first data write speed to obtain the first influence coefficient, which is used to characterize the influence of data reading in other storage areas on the data write speed of the current storage area; The second influence coefficient is obtained by dividing the third data write speed by the first data write speed, which is used to characterize the impact of data reading in the current storage area on the data write speed in the current storage area. S104. For each storage region, repeat steps S102 to S103; S2. When it is necessary to write data to be stored into the solid-state drive, determine whether there is a region in the solid-state drive where data is being read, and calculate the actual data writing speed of each storage region based on the determination result; S3. Based on the latency requirements of the data to be stored, the actual writing speed of data in each storage area, and the remaining storage space, the selected storage areas are used for parallel writing of the data to be stored, ensuring that the data writing is completed simultaneously.

2. The time-delay-based solid-state drive data storage method according to claim 1, characterized in that: Step S2 includes: S201. When it is necessary to write data to the solid-state drive (SSD), determine whether there is a region on the SSD where data is currently being read: If it does not exist, the first data write speed of each storage area is the actual data write speed; If it exists, proceed to step S202; S202. For any storage area, determine whether the area where data is being read is the current storage area or another storage area of ​​the solid-state drive; If the area where data is being read is another storage area of ​​the solid-state drive, the actual data write speed of the current storage area is obtained by multiplying the first data write speed of the current storage area by the first influence coefficient. If the area where data is being read is the current storage area, then the actual data write speed of the current storage area is obtained by multiplying the first data write speed of the current storage area by the second influence coefficient.

3. The time-delay-based solid-state drive data storage method according to claim 1, characterized in that: Step S3 includes: S301. Calculate the remaining storage space of each storage area, and remove storage areas with remaining storage space less than a preset threshold. The remaining storage areas are used as candidate storage areas. The remaining candidate storage areas are sorted from largest to smallest according to the actual data writing speed. S302. Determine the number M of storage regions selected during initialization based on the storage latency requirements of the data to be stored; Assume the size of the data to be stored is... The storage latency of the data to be stored must not exceed [a certain value]. Then the number M of the initial selected storage regions is: ； S303. From the storage areas, select M storage areas in descending order of their actual data write speeds. The actual data write speeds of these M storage areas are denoted as... ,in This represents the actual speed at which data is written to the i-th storage region out of M storage regions, where i = 1, 2, ..., M; Calculate the time required to write data to M storage areas in parallel. ; S304. Divide the data to be stored into M data blocks, with the sizes of the data blocks being respectively... , where the size is The data blocks are allocated to the i-th storage region among M storage regions, i=1,2,…,M; S305. Determine if any of the M storage regions has remaining storage space smaller than the allocated data block size: If it exists, increment the value of M by 1 and return to step S303; If it does not exist, according to the division in step S304, each data block is written in parallel to the corresponding storage area.

Images