CN107678694A - RAID stripe method for reconstructing and solid-state disk - Google Patents

Abstract

The invention discloses a kind of RAID stripe method for reconstructing applied to solid-state disk, forms new RAID stripe for carrying out restructuring to the physical block group of RAID memory arrays, this method includes：Step S1, detects the raw Bit-Error-Rate of physical block and is updated and be recorded in bulk state table；Step S2, the raw Bit-Error-Rate in bulk state table obtain weak block message；Step S3, weak block number mesh is obtained by bulk state table, and band tissue is carried out to physical block；Step S4, the LSB page of the physical block in same RAID stripe and MSB pages are staggeredly subjected to band tissue in block.The RAID stripe method for reconstructing of the present invention, the dynamic adjustment of solid-state disk RAID stripe is realized, avoids loss of data, while avoid waste of storage space, improved the reliability of solid-state disk, extend the life-span of solid-state disk.Meanwhile the invention also discloses a kind of solid-state disk, it includes RAID memory arrays and the controller for the RAID memory arrays adaptively adjust and be managed the data that it is stored.

Description

RAID Stripe Reconstruction Method and Solid State Disk

technical field

The invention relates to the technical field of electronic storage, in particular to a RAID stripe rebuilding method and a solid state disk.

Background technique

Because of its high performance, low energy consumption, and good shock resistance, solid-state disks have been widely used in server systems as a viable substitute for disks. With the progress of flash memory technology, especially the development from SLC (Single-LevelCell) technology to MLC (Multi-LevelCell) technology and flash memory process technology, the capacity of flash memory is increasing day by day. However, the endurance of the physical blocks of flash memory is greatly reduced, and the reliability is significantly reduced. In order to ensure data consistency, manufacturers of solid-state disks use a Redundant Array of Independent Hard Disks (RAID) mechanism to enhance the reliability of solid-state disks.

One of the functions to be realized by RAID is to strip multiple storages controlled by the controller, and striping multiple storages can combine independent and unrelated storages into an overall storage array. Striping is a method in which the controller divides continuous data into data blocks of the same size, and writes each segment of data to different storage devices in the array. A stripe is a volume formed by combining multiple storage devices.

Solid-state disk storage is different from traditional disks. NAND flash memory is a common particle that makes up solid-state disks. The writing operation of the flash memory needs to go through two operation processes of "erase-write". If the storage unit is to be written, the physical block of the storage unit must be erased. After the erase is completed, the data content of the entire physical block is cleared. The erasing operation process is called recycling.

Among them, NAND flash memory has some special error characteristics. The raw bit error rate (RBER) of NAND flash memory will increase exponentially with the increase of program/erase cycles (P/Ecycle). Moreover, due to differences in manufacturing processes, there is an uneven distribution of errors in solid state disks in physical blocks, flash memory pages, and the like. For example, under the same programming/erasing times, different physical blocks exhibit different error characteristics; in the same physical block, the page composed of the most significant bits (MSB page) is more error-prone than the page composed of the least significant bits (LSB page). Raw BER is higher and MSB pages have less wear resistance than LSB pages.

However, traditional RAID does not consider the error characteristics of NAND flash memory, and adopts a fixed-strength RAID strategy throughout the life of the solid-state disk, resulting in a large number of programming/erasing times in the later use of the solid-state disk, and the original bit error rate of the flash medium. Increasingly, the reliability provided by fixed-strength RAID is not enough to meet later needs.

In the prior art, SehwanLee et al. proposed a lifetime-aware RAID solution, which uses RAID strategies of different strengths under different programming/erasing times to ensure the reliability of the solid-state disk throughout its lifetime with the minimum storage space overhead. However, life-aware RAID ignores the difference in media wear resistance (physical block level and flash page level) due to the manufacturing process, and uses the number of programming/erasing as the only indicator of the degree of wear to adjust the RAID strength. The space capacity is sacrificed, but the RAID strength adjustment still cannot meet the reliability requirements.

On the one hand, the difference between wear resistance at the physical block level is ignored: when two or more physical blocks with higher raw BER are in the same stripe, the increase in stripe reliability cannot offset the physical On the other hand, ignoring the difference between the wear resistance of the flash memory page level: the life-aware RAID divides the LSB page and the MSB page into RAID stripes, so that the entire MSB stripe The degree of wear is much higher than that of LSB stripes, and because the maximum wear of the physical block flash page determines the wear of the physical block, when a physical block is unavailable, the LSB page in the physical block can still be used, thus bringing space waste.

In view of this, it is necessary to provide a RAID stripe rebuilding method and a solid-state disk that can solve the above-mentioned defects, can prevent RAID stripe data loss and have high space utilization, and can realize adaptive reorganization of RAID memory arrays.

Contents of the invention

A technical problem to be solved by the present invention is to provide a RAID stripe rebuilding method which can realize the self-adaptive reorganization of the RAID memory array and improve the reliability and practicability of the solid state disk.

Another technical problem to be solved by the present invention is to provide a solid state disk that can prevent RAID stripe data loss and has high space utilization.

In order to solve the above-mentioned technical problems, the present invention adopts the following technical scheme: a method for rebuilding a RAID stripe, which is applied to a solid-state disk to rebuild a physical block group of a RAID memory array to form a new RAID stripe, and the method includes the following steps: step:

Step S1, detecting the original bit error rate of the physical block and updating and recording it in the block status table;

Step S2, obtaining weak block information according to the original bit error rate in the block state table;

Step S3, obtain the number of weak blocks through the block state table, and organize the physical blocks in stripes;

Step S4, interleaving the LSB pages and MSB pages of the physical blocks in the same RAID stripe to organize stripes within the block.

Its further technical solution is: said step S1 specifically includes the following steps:

Step S1.1, setting the programming/erasing times threshold, the weak block original bit error rate threshold and the total number of physical blocks in the physical block group;

Step S1.2, judging whether the number of times of programming/erasing reaches the threshold of times of programming/erasing for which the original bit error rate of the physical block needs to be detected;

Step S1.3, according to the judgment result of step S1.2, when the number of programming/erasing times reaches the programming/erasing times threshold value of the original bit error rate detection of the physical block, the original error rate of the physical block is performed on the physical block group. Code rate detection is recorded in the block status table at the same time.

Its further technical solution is: said step S1.3 is specifically: according to the judgment result of step S1.2, when the number of programming/erasing times reaches the threshold value of the number of programming/erasing times that needs to be detected for the original bit error rate of the physical block, read Get all the flash memory pages in each physical block of the physical block group, obtain the original bit error rate of the flash memory page, and use the largest original bit error rate of the flash memory page in each physical block as the original bit error rate of the physical block, and record it in in the block state table.

Its further technical solution is: the step S1 also includes: according to the judgment result of the step S1.2, when the programming/erasing frequency does not reach the programming/erasing frequency threshold for the detection of the original bit error rate, directly execute the step S3 .

Its further technical solution is: said step S2 specifically includes the following steps:

Step S2.1, compare the original bit error rate value recorded in the block state table with the set weak block original bit error rate threshold according to step S1.3, and judge whether the original bit error rate value of the physical block exceeds the set weak block original bit error rate value. BER threshold;

Step S2.2, according to the judgment result of step S2.1, if the original bit error rate value of the physical block exceeds the weak block original bit error rate threshold, mark the physical block as a weak block and record it in the block status table.

Its further technical solution is: said step S2 further includes: according to the judgment result of step S2.1, when the original bit error rate value of the physical block does not exceed the set weak block original bit error rate threshold, then directly execute step S3.

Its further technical solution is: said step S2.2 also includes:

Step S2.2.1, bad block prediction and judgment, obtain the number of weak blocks according to the weak block information in the block state table, and judge whether the number of weak blocks exceeds 1/2 of the number of physical blocks in the physical block group;

Step S2.2.2, according to the judgment result of step S2.2.1, when the number of weak blocks exceeds 1/2 of the total number of physical blocks, predict the maximum original bit error rate in the physical block group according to the original bit error rate information in the block state table If the physical block is a bad block, replace the bad block, and update and record the original BER of the new physical block after replacement in the block status table; otherwise, directly execute step S3.

Its further technical solution is: the step S3 is specifically: obtain the number of weak blocks through the weak block information in the block state table, and perform stripe organization according to the number of weak blocks, and determine the number of RAID stripes of the physical block group according to the number of weak blocks, Place each weak block in a different RAID stripe to ensure that there is only one weak block in each RAID stripe, and at the same time distribute the remaining non-weak block physical blocks in each RAID stripe evenly according to the number.

Its further technical solution is: said step S3 also includes:

Step S3.1, storing the verification data in the weak block of the physical block group of the RAID stripe.

In order to solve the above-mentioned technical problems, the present invention also provides a solid-state disk, including a RAID memory array and a controller, and the RAID memory array is used to store data in segments; when the programming/erasing times of the solid-state disk exceed the preset When the threshold is , the controller is configured to execute the above RAID stripe rebuilding method to perform adaptive reorganization on the RAID storage array, and manage the storage data of the RAID storage array.

The beneficial technical effects of the present invention are: the RAID stripe reconstruction method provided by the present invention senses the actual wear and tear of the physical block by detecting the original bit error rate of the physical block, and divides the physical block into weak blocks and non-weak blocks according to the original bit error rate value. Blocks are separated from physical blocks and recorded in the block status table, and the physical blocks are organized in stripes according to the number of weak blocks, which improves the reliability of the solid-state disk system and prolongs the life of the solid-state disk, which can effectively protect data and prevent data loss; at the same time, The LSB page and MSB page of the physical block in the same RAID stripe are interleaved for intra-block stripe organization to overcome the uneven distribution of data page errors in the physical block and avoid the waste of storage space caused by large reliability differences. problems and improve storage space utilization. At the same time, the invention also provides a solid state disk including a RAID memory array and a controller, and the controller can perform self-adaptive reorganization on the RAID memory array to improve the reliability of the solid state disk system and prolong the service life of the solid state disk system.

Description of drawings

FIG. 1 is a flow chart of the RAID stripe rebuilding method of the present invention.

Fig. 2 is a flow chart of a specific embodiment of the RAID stripe rebuilding method of the present invention.

FIG. 3 is a schematic diagram of RAID stripe organization among physical blocks in the RAID stripe rebuilding method of the present invention.

detailed description

In order to make those skilled in the art more clearly understand the purpose, technical solutions and advantages of the present invention, the present invention will be further described below in conjunction with the accompanying drawings and embodiments. Apparently, the embodiments described below are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

The RAID stripe rebuilding method of the present invention is aimed at the RAID stripe organization with the stripe depth as the block or the RAID stripe organization whose physical position is static, and is applied to the solid state disk to rebuild the physical block group of the RAID memory array to form a new RAID stripes can be used to store data, and the RAID stripes can be dynamically adjusted to improve the reliability of the SSD system.

With reference to Fig. 1, described RAID stripe rebuilding method comprises the following steps:

Step S1, detecting the original bit error rate of the physical block and updating and recording it in the block state table.

In this embodiment, the original bit error rate of the physical block is detected by the read-after-write method, that is, the written data is read immediately after the programming of the physical block to detect the error of the data, thereby detecting the original bit error rate of the physical block For specific read-after-write methods, please refer to "New ERA: new efficient reliability-aware wear leveling for endurance enhancement of flash storage devices" (Proceedings of the 50th Annual Design Automation Conference, Article NO.163, 2013) published by MingchangYang et al. Journal literature.

Referring to Figure 2, in some embodiments, such as this embodiment, the step S1 specifically includes the following steps:

Step S1.1, setting the programming/erasing times threshold, the weak block raw bit error rate threshold and the total number of physical blocks in the physical block group.

Among them, the weak block is a physical block with high original bit error rate and poor wear resistance. Due to the characteristic that the original bit error rate in the physical block increases exponentially with the number of programming/erasing times, the number of settings for the programming/erasing times threshold is multiple, and the intervals between the set programming/erasing times thresholds are also different ; then when the number of programming/erasing times is small, the degree of wear of the physical block changes slowly, and within a range of programming/erasing times, the wear of the physical block does not change much. The interval of the threshold value of the erasing times is relatively large; with the use of the solid state disk, the wear of the physical block changes rapidly with the increase of the number of programming/erasing times, so the interval of the threshold value of the programming/erasing times is set to be small. The programming/erasing times threshold is used to judge whether to perform the original bit error rate detection of the physical block, and when the programming/erasing times exceed the programming/erasing times threshold, the original bit error rate detection is performed on the physical block; The weak block original bit error rate threshold is used to judge whether the physical block is a weak block, and when the detected physical block original bit error rate exceeds the weak block original bit error rate threshold, the physical block is a weak block.

Step S1.2, judging whether the number of times of programming/erasing reaches the threshold value of the number of times of programming/erasing required to detect the original bit error rate of the physical block.

By setting the programming/erasing times threshold, it can be judged whether the original bit error rate of the physical block needs to be detected according to the programming/erasing times, so as to control the frequency of the original bit error rate detection of the physical block, reduce the reading of the physical block data, and reduce the The overhead of raw bit error rate detection for physical blocks in SSDs. During the use of the solid state disk, its programming/erasing times are compared with the set programming/erasing times threshold, and when the programming/erasing times exceed the programming/erasing times threshold, the original bit error rate of the physical block is calculated. detection.

Step S1.3, according to the judgment result of step S1.2, when the number of programming/erasing times reaches the programming/erasing times threshold value of the original bit error rate detection of the physical block, the original error rate of the physical block is performed on the physical block group. The bit rate detection simultaneously records the corresponding original bit error rate value in the block state table.

Further, step S1.3 is specifically: according to the judgment result of step S1.2, when the number of programming/erasing times reaches the programming/erasing times threshold set in step S1.1 that needs to detect the original bit error rate of the physical block, Read all the flash memory pages in each physical block of the physical block group, obtain the original bit error rate of the flash memory page, and use the largest flash memory page original bit error rate value in each physical block as the original bit error rate of the physical block , recorded in the block state table.

Step S2, obtaining weak block information according to the original bit error rate in the block state table.

Among them, physical blocks with high original bit error rate and poor wear resistance are marked as weak blocks. The basis for judging whether a physical block is a weak block is the original bit error rate of the physical block, and detecting the original bit error rate of the physical block can provide a basis for comparing and judging whether the physical block is a weak block. The weak block information represents information of physical blocks marked as weak blocks in the block status table, and is used to know which physical blocks are weak blocks.

Continuing to refer to FIG. 2, in some embodiments, such as this embodiment, the step S2 specifically includes the following steps:

Step S2.1, compare the original bit error rate value recorded in the block state table with the set weak block original bit error rate threshold according to step S1.3, and judge whether the original bit error rate value of the physical block exceeds the set weak block original bit error rate value. Bit error rate threshold.

Step S2.2, according to the judgment result of step S2.1, if the original bit error rate value of the physical block exceeds the weak block original bit error rate threshold, mark the physical block as a weak block and record it in the block status table.

In some embodiments, such as this embodiment, the step S2.2 further includes the following steps:

Step S2.2.1, bad block prediction judgment, obtain the number of weak blocks according to the weak block information in the block state table, and judge whether the number of weak blocks exceeds 1/2 of the total number of physical blocks in the physical block group.

Step S2.2.2, according to the judgment result of step S2.2.1, when the number of weak blocks exceeds 1/2 of the total number of physical blocks, predict the maximum original bit error rate in the physical block group according to the original bit error rate information in the block state table If the physical block is a bad block, replace the bad block, and update and record the original BER of the new physical block after replacement in the block status table; otherwise, directly execute step S3.

Among them, the bad block is the physical block with the highest original bit error rate in the physical block group when the number of weak blocks exceeds 1/2 of the total number of physical blocks, and the bad block is replaced by migrating the valid data in the bad block to the original low bit error rate Erase bad blocks, and record the original bit error rate of the physical block with low original bit error rate or the original bit error rate of the free physical block in the block status table. Complete the migration of valid data before erasing bad blocks, which can avoid the performance overhead caused by RAID data recovery when bad blocks are replaced; the original bit error rate of the physical block with a low original bit error rate or the original bit error rate of the free physical block All are lower than the original bit error rate threshold of the weak block, that is, the new physical block after replacing the bad block is a non-weak block physical block, so as to ensure that the number of weak blocks is less than 1/2 of the total number of physical blocks, so that there are no bad blocks in the solid state disk , to prevent data loss during physical block writing. When the number of weak blocks exceeds 1/2 of the total number of physical blocks, the bad block prediction judgment is triggered, and the physical block with the highest original bit error rate is marked as a bad block, which is beneficial to avoid the use of bad blocks in the later use of the solid-state disk. Reduce the error rate.

In step S3, the number of weak blocks is acquired through the block status table, and the physical blocks are organized into stripes.

In combination with Figures 2 and 3, further, the step S3 specifically includes: according to the weak block information in the block state table, obtain the number of weak blocks, perform stripe organization according to the number of weak blocks, and determine the physical block group RAID according to the number of weak blocks The number of stripes, put each weak block in a different RAID stripe, ensure that there is only one weak block in each RAID stripe, and distribute the remaining non-weak physical blocks in each RAID evenly according to the number in the strip.

In some embodiments, such as this embodiment, the step S1 further includes: according to the judgment result of the step S1.2, when the number of times of programming/erasing does not reach the programming/erasing required for the detection of the original bit error rate of the physical block When the times threshold is reached, step S3 is directly executed.

In some embodiments, such as this embodiment, the step S2 further includes: according to the judgment result of step S2.1, when the original bit error rate value of the physical block does not exceed the set weak block original bit error rate threshold, then Execute step S3 directly.

Referring to Figure 3, Figure 3 shows the combination of RAID stripes between physical blocks in the solid-state disk, each row in Figure 3 represents the RAID stripes between the physical blocks in the solid-state disk when the usage of the solid-state disk is different, where the square represents the physical block, Dark physical blocks represent weak blocks. In the initial stage of solid state disk use, the number of programming/erasing is small. When the number of weak blocks is 0, all physical blocks of the entire solid state disk form a RAID stripe, as shown in the first row in Figure 3; when the number of weak blocks is 1 , the non-weak physical blocks of the solid-state disk and the weak block form a RAID stripe, as shown in the second row in Figure 3; when the number of weak blocks is 2, the number of RAID stripes is two, each There is only one weak block in a RAID stripe, and the remaining non-weak block physical blocks in the solid-state disk are evenly distributed to each RAID stripe according to the number, as shown in the third line in Figure 3; with the use of the solid-state disk, the number of weak blocks Gradually increase, when the number of weak blocks reaches half of the number of physical blocks of the solid state disk, then a weak block and a non-weak physical block form a RAID stripe, and at this time the maximum redundancy is reached.

Determine the number of RAID stripes according to the number of weak blocks, and dynamically adjust the size and composition of RAID stripes to ensure that there is only one weak block in each RAID stripe. By increasing the strength of RAID to offset the wear of weak blocks, data loss can be avoided and improved SSD reliability.

Continuing to refer to FIG. 3, in some embodiments, such as this embodiment, the step S3 further includes:

Step S3.1, storing the verification data in the weak block of the physical block group of the RAID stripe.

Wherein, the verification data is used to detect the integrity and accuracy of the data. In conjunction with Fig. 3, P represents the verification data in the figure, and the verification data is placed in the weak block of each band, because the verification data will only be in the Reading is performed only after data recovery, while weak blocks only contain verification data, which can avoid the influence of read interference, minimize interference to weak blocks, and reduce users' data reading of weak blocks with higher error rates, thereby reducing read ECC (ErrorCorrectingCode) decoding time in the fetching process improves the read performance of the SSD. When the number of programming/erasing is small, the weak block data in the physical block group is small. In the early stage of solid state disk use, when there is no weak block in the physical group, the check data is placed in the last physical block in the RAID stripe .

Based on the above steps, with the use of solid-state disks, the number of weak blocks gradually increases. When the number of weak blocks reaches 1/2 of the total number of physical blocks, each RAID stripe is a RAID1 composed of a weak block and a non-weak block. , and the verification data is placed in the weak block, and the maximum verification redundancy is reached at this time. Each RAID stripe has only one weak block, which can avoid the redundancy of parity data and reduce the waste of storage space.

Step S4, interleaving the LSB pages and MSB pages of the physical blocks in the same RAID stripe to organize stripes within the block.

Interleaving the LSB page and MSB page for intra-block stripe organization can overcome the uneven distribution of data page errors in the physical block. When the MSB page is worn out to be unusable, the LSB page can be used for data recovery to avoid MSB pages. data loss, so as to avoid the waste of storage space caused by the large reliability difference of the physical page group, and improve the utilization rate of storage space.

The present invention also provides a solid-state disk, including a RAID memory array and a controller, wherein the memory array is used to store data in segments; when the programming/erasing times of the solid-state disk exceed a preset threshold, the The controller is used to execute the above-mentioned RAID stripe rebuilding method to perform adaptive reorganization on the RAID memory array, and manage the storage data of the RAID memory array, so as to improve system reliability and prolong system life.

In summary, the RAID stripe reconstruction method of the present invention senses the actual wear and tear of the physical block by detecting the original bit error rate of the physical block, and distinguishes the physical block into weak blocks and non-weak blocks according to the original bit error rate value, Determine the number of RAID stripes according to the number of weak blocks in the stripe organization operation, ensure that each RAID stripe contains only one weak block, and distribute the non-weak block physical blocks in the physical block group of the solid state disk evenly according to the number In the RAID stripe, avoid data loss caused by two or more weak blocks in the same RAID stripe, improve the reliability of the solid-state disk system, and extend the life of the solid-state disk; perform bad block detection, and replace bad blocks Migrating valid data in advance can effectively protect data and prevent data loss during RAID stripe recovery and bad block replacement; set verification data in weak blocks to reduce reading of weak block data with higher error rates and improve SSD performance. At the same time, the LSB page and MSB page of the physical block in the same RAID stripe are interleaved for intra-block stripe organization to overcome the uneven distribution of data page errors in the physical block and avoid large reliability differences. The resulting waste of storage space can improve space utilization; set up a bad block prediction mechanism to predict physical block errors, so as to prevent data loss when the solid-state disk is used later. At the same time, the solid state disk of the present invention can adaptively reorganize the RAID memory array through the controller to dynamically adjust the composition and size of the RAID stripes, improve system reliability and prevent data loss.

The above descriptions are only preferred embodiments of the present invention, and do not limit the present invention in any form. Those skilled in the art can make various equivalent changes and improvements on the basis of the above-mentioned embodiments, and all equivalent changes or modifications made within the scope of the claims shall fall within the protection scope of the present invention.

Claims (10)

1. A method for rebuilding a RAID stripe is applied to a solid-state disk to reorganize the physical blocks of the RAID memory array to form a new RAID stripe, and it is characterized in that the method may further comprise the steps: Step S1, detecting the original bit error rate of the physical block and updating and recording it in the block status table; Step S2, obtaining weak block information according to the original bit error rate in the block state table; Step S3, obtain the number of weak blocks through the block state table, and organize the physical blocks in stripes; Step S4, interleaving the LSB pages and MSB pages of the physical blocks in the same RAID stripe to organize stripes within the block. 2. The RAID stripe reconstruction method according to claim 1, wherein said step S1 specifically comprises the following steps: Step S1.1, setting the programming/erasing times threshold, the weak block original bit error rate threshold and the total number of physical blocks in the physical block group; Step S1.2, judging whether the number of times of programming/erasing reaches the threshold of times of programming/erasing for which the original bit error rate of the physical block needs to be detected; Step S1.3, according to the judgment result of step S1.2, when the number of programming/erasing times reaches the programming/erasing times threshold value of the original bit error rate detection of the physical block, the original error rate of the physical block is performed on the physical block group. Code rate detection is recorded in the block status table at the same time. 3. The RAID stripe reconstruction method according to claim 2, wherein the step S1.3 is specifically: according to the judgment result of the step S1.2, when the number of times of programming/erasing reaches the original error of the physical block, When the programming/erasing times threshold of bit rate detection is reached, read all the flash pages in each physical block of the physical block group, obtain the original bit error rate of the flash page, and calculate the largest original bit error rate of the flash page in each physical block The rate is recorded in the block status table as the original bit error rate of the physical block. 4. The RAID stripe reconstruction method according to claim 2, wherein said step S1 further comprises: according to the judgment result of step S1.2, when the number of times of programming/erasing has not reached, the original code error of the physical block needs to be performed When the programming/erasing times threshold of rate detection is exceeded, step S3 is directly performed. 5. The RAID stripe reconstruction method according to claim 2, wherein said step S2 specifically comprises the following steps: Step S2.1, compare the original bit error rate value recorded in the block state table with the set weak block original bit error rate threshold according to step S1.3, and judge whether the original bit error rate value of the physical block exceeds the set weak block original bit error rate value. BER threshold; Step S2.2, according to the judgment result of step S2.1, if the original bit error rate value of the physical block exceeds the weak block original bit error rate threshold, mark the physical block as a weak block and record it in the block status table. 6. The RAID stripe reconstruction method according to claim 5, wherein said step S2 further comprises: according to the judgment result of step S2.1, when the original bit error rate value of the physical block does not exceed the set weak block The original bit error rate threshold, then directly execute step S3. 7. The RAID stripe rebuilding method as claimed in claim 5, is characterized in that, described step S2.2 also comprises: Step S2.2.1, bad block prediction and judgment, obtain the number of weak blocks according to the weak block information in the block state table, and judge whether the number of weak blocks exceeds 1/2 of the total number of physical blocks in the physical block group; Step S2.2.2, according to the judgment result of step S2.2.1, when the number of weak blocks exceeds 1/2 of the total number of physical blocks, predict the maximum original bit error rate in the physical block group according to the original bit error rate information in the block state table If the physical block is a bad block, replace the bad block, and update and record the original BER of the new physical block after replacement in the block status table; otherwise, directly execute step S3. 8. The RAID stripe reconstruction method according to claim 1, wherein said step S3 is specifically: obtaining the number of weak blocks through the weak block information of the block state table, and performing stripe organization according to the number of weak blocks, according to The number of weak blocks determines the number of RAID stripes in the physical block group. Place each weak block in a different RAID stripe to ensure that there is only one weak block in each RAID stripe. The numbers are evenly distributed in each RAID stripe. 9. The RAID stripe reconstruction method according to claim 1, wherein said step S3 further comprises: Step S3.1, storing the verification data in the weak block of the physical block group of the RAID stripe. 10. A solid state disk, comprising a RAID memory array and a controller, characterized in that: the RAID memory array is used for segmented storage of data; when the programming/erasing times of the solid state disk exceed a preset threshold, the The controller is configured to execute the RAID stripe rebuilding method according to any one of claims 1 to 9 to perform adaptive reorganization on the RAID memory array, and manage storage data of the RAID memory array.