Solid state disk fault pre-diagnosis and data security self-repair processing method and device

Abstract

The present application relates to a kind of solid state disk fault pre-diagnosis and data security self-repair processing method and device, belong to memory technical field, method includes: in the running process of solid state disk, the block erase time of each memory block is monitored, when the block erase time of one or more memory blocks is first exceeded early warning threshold, the memory block is marked as early warning state;The block erase time variation rate of memory block marked as early warning state is continuously monitored, if the block erase time variation rate exceeds acceleration degradation threshold, then determine that memory block enters high-risk state;Hot data identification operation is carried out to the memory block of high-risk state, identify the data page of the wear value in the memory block higher than set level;The identified high wear value data page is migrated to the reserved healthy memory block, and after migration is completed, write protection locking operation is carried out to the memory block of original high-risk state;Realize the effect of data security initiative protection and storage life collaborative optimization.

Description

Technical Field

[0001] This invention relates to the field of memory technology, and in particular to a method and apparatus for solid-state drive fault pre-diagnosis and data security self-repair processing. Background Technology

[0002] Solid-state drives (SSDs), as mainstream storage devices, pose a core threat to data security due to sudden failures caused by the physical degradation of their storage units. Current industry solutions suffer from fundamental flaws: firstly, they rely on static thresholds to trigger warnings, failing to recognize dynamic degradation characteristics such as accelerated oxide layer degradation of storage units, leading to severely delayed warnings (e.g., fixed erase time limits or bad block count thresholds); secondly, the use of full-block data migration for repair mechanisms bundles low-risk cold data with high-wear hot data, significantly increasing the effective load, accelerating the wear of healthy blocks, and causing secondary data corruption due to the lack of redundant verification during the migration process. These issues mean that SSDs still pose an unacceptable risk of data loss in high-reliability scenarios such as finance and healthcare. Summary of the Invention

[0003] The main objective of this invention is to provide a method and apparatus for pre-diagnosis of solid-state drive (SSD) faults and self-repair of data security. By accurately identifying and directionally migrating high-wear data pages within high-risk storage blocks, it achieves proactive data security protection and synergistic optimization of storage lifespan.

[0004] To achieve the above objectives, the present invention provides a solid-state drive fault pre-diagnosis and data security self-repair method, comprising the following steps:

[0005] During the operation of the solid-state drive, the block erase time of each storage block is monitored. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as a warning state.

[0006] The block erase time change rate of storage blocks marked as warning status is continuously monitored. If the block erase time change rate exceeds the accelerated degradation threshold, the storage block is determined to be in a high-risk state.

[0007] Perform a thermal data identification operation on the storage block in a high-risk state to identify data pages within the storage block whose wear value is higher than a set level;

[0008] The identified high-wear data pages are migrated to reserved healthy storage blocks, and after the migration is completed, write-protected locking is performed on the storage blocks that were originally in a high-risk state.

[0009] Furthermore, the warning threshold is 1.5 to 3 times the baseline erase time, which is obtained based on the statistical analysis of historical failure samples from the same batch of solid-state drives.

[0010] Further, the step of performing a hot data identification operation on the storage block in a high-risk state, and identifying data pages within the storage block whose wear value is higher than a set level, includes:

[0011] Record the read interference count and programming loop count for all data pages within the storage block;

[0012] Compare the read interference count of each data page with the preset read interference threshold, and the number of programming loops with the preset loop threshold;

[0013] Data pages that simultaneously exceed both the read interference threshold and the programming cycle threshold are marked as high-wear data pages;

[0014] Generate a hot data identifier table containing the physical addresses of all high-wear data pages.

[0015] Further, the step of recording the read interference count and programming loop count of all data pages within the storage block includes:

[0016] In response to word line voltage fluctuation signals, read operations exceeding the voltage stability threshold are converted into read interference count increments;

[0017] Capture the injection pulse signal of the charge pump and convert the programming operation to complete charge verification into an increment of programming loop count;

[0018] The read interference count is stored in the non-volatile register of the memory block controller, and the number of programmable loops is stored in the physically isolated loop marker cell.

[0019] Furthermore, the step of relocating identified high-wear data pages to reserved healthy storage blocks includes:

[0020] During the idle period of the solid-state drive, the physical address of the high-wear data page is loaded according to the hot data identification table, and a read command is sent to the storage block with high source risk, while a double redundant write is initiated to the target healthy storage block.

[0021] The data page erasure of the source high-risk storage block is triggered only when the target healthy storage block returns two sets of write verification signals that are consistent.

[0022] Update the address mapping table, bind the logical address of the migrated data page to the physical address of the healthy storage block, and generate a migration completion log.

[0023] Furthermore, the healthy storage block is selected from a reserved pool independent of the user storage area, and the capacity of the reserved pool accounts for 0.5%-1% of the total capacity of the storage unit.

[0024] This invention proposes a solid-state drive fault pre-diagnosis and data security self-repair processing device, comprising:

[0025] The first marking unit is used to monitor the block erase time of each storage block during the operation of the solid-state drive. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as a warning state.

[0026] The second marking unit is used to continuously monitor the block erase time change rate of storage blocks marked as warning status. If the block erase time change rate exceeds the accelerated degradation threshold, the storage block is determined to have entered a high-risk state.

[0027] The diagnostic unit is used to perform thermal data identification operations on the storage block in a high-risk state, and to identify data pages in the storage block whose wear value is higher than a set level;

[0028] The repair unit is used to migrate the identified high-wear data pages to a reserved healthy storage block, and after the migration is completed, to perform a write-protection locking operation on the storage block that was originally in a high-risk state.

[0029] Furthermore, the diagnostic unit includes:

[0030] The memory sub-unit is used to record the read interference count and programming loop count of all data pages within the storage block;

[0031] The comparison sub-unit is used to compare the read interference count of each data page with the preset read interference threshold, and the number of programming loops with the preset loop threshold;

[0032] The marking subunit is used to mark data pages that simultaneously exceed the read interference threshold and the programming cycle threshold as high-wear data pages;

[0033] The statistics subunit is used to generate a hot data identifier table containing the physical addresses of all high-wear data pages.

[0034] Furthermore, the repair unit includes:

[0035] The write subunit is used to load the physical address of the high-wear data page according to the hot data identification table during the idle period of the solid-state drive, send a read command to the storage block with high source risk, and simultaneously initiate double redundant write to the target healthy storage block.

[0036] The erase subunit is used to trigger the erasure of the data page corresponding to the source high-risk storage block only when the target healthy storage block returns two sets of write verification signals that are consistent.

[0037] The update sub-unit is used to update the address mapping table, bind the logical address of the migrated data page to the physical address of the healthy storage block, and generate a migration completion log.

[0038] The present invention also proposes a solid-state drive having the above-described apparatus deployed thereon to specifically execute the above-described method via the apparatus.

[0039] The solid-state drive (SSD) fault pre-diagnosis and data security self-repair processing method and apparatus provided by this invention have the following advantages: they achieve early warning of faults by dynamically monitoring the block erase time change rate, accurately locate high-wear data pages in high-risk storage blocks for targeted migration, effectively avoid the additional losses caused by full block migration, and at the same time, adopt a redundancy verification mechanism to ensure data integrity during the migration process, significantly extend the life of the SSD and improve data security. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating a solid-state drive fault pre-diagnosis and data security self-repair method according to an embodiment of the present invention;

[0041] Figure 2 This is a structural block diagram of a solid-state drive fault pre-diagnosis and data security self-repair processing device according to an embodiment of the present invention.

[0042] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] Reference Figure 1 This is a flowchart illustrating a method for pre-diagnosis of solid-state drive faults and self-repair of data security proposed in this invention. The method includes the following steps:

[0045] S1. During the operation of the solid-state drive, the block erase time of each storage block is monitored. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as a warning state.

[0046] S2, continuously monitor the block erase time change rate of storage blocks marked as warning status. If the block erase time change rate exceeds the accelerated degradation threshold, the storage block is determined to be in a high-risk state.

[0047] S3, perform a hot data identification operation on the storage block in a high-risk state to identify data pages in the storage block whose wear value is higher than a set level;

[0048] S4, the identified high-wear data pages are migrated to the reserved healthy storage blocks, and after the migration is completed, write protection locking is performed on the storage blocks that were originally in a high-risk state.

[0049] Specifically, during the power-on operation of the solid-state drive (SSD), the physical layer timer collects the block erase time of each storage block in real time. When the main controller chip detects that the block erase time of a specific storage block exceeds the warning threshold for the first time, it immediately marks it as a warning state through the status register of the storage block controller. Subsequently, the firmware starts a high-frequency monitoring mode, recording the block erase time of the warning-state storage blocks every 5 minutes. If the time increase exceeds the increase threshold in three consecutive recording cycles and the single increase is greater than 120% of the previous one, the storage block is determined to be in a high-risk state and an interrupt signal is triggered. After the high-risk state is confirmed, the storage block controller retrieves the read value from the non-volatile register. The interference count and the number of programming cycles in the cycle marking unit are compared with the preset read interference threshold and the programming cycle threshold. The physical addresses of data pages that exceed the limits simultaneously are written to the hot data identification table. During the hard disk idle period, a dual-channel verification migration is initiated based on the hot data identification table: while reading data pages from the source high-risk storage block, double redundancy writes are performed to the target healthy storage block. Only when the two sets of write verification signals are consistent after being verified by the voltage comparator are the corresponding data pages of the source block erased and the address mapping table updated. Immediately after the migration is completed, a write-prohibit command is sent to the high-risk storage block, a lock flag is written to the mapping table, and the lock time and high-risk parameters are recorded to the physical marking log area.

[0050] The aforementioned warning threshold is 1.5 to 3 times the baseline erase time, which is obtained based on the statistical analysis of historical failure samples from the same batch of solid-state drives.

[0051] In one embodiment, performing a hot data identification operation on the storage block in a high-risk state, and identifying data pages within the storage block whose wear values ​​exceed a set level, includes:

[0052] Record the read interference count and programming loop count for all data pages within the storage block;

[0053] Compare the read interference count of each data page with the preset read interference threshold, and the number of programming loops with the preset loop threshold;

[0054] Data pages that simultaneously exceed both the read interference threshold and the programming cycle threshold are marked as high-wear data pages;

[0055] Generate a hot data identifier table containing the physical addresses of all high-wear data pages.

[0056] The steps of recording the read interference count and programming loop count of all data pages within the storage block include:

[0057] In response to word line voltage fluctuation signals, read operations exceeding the voltage stability threshold are converted into read interference count increments;

[0058] Capture the injection pulse signal of the charge pump and convert the programming operation to complete charge verification into an increment of programming loop count;

[0059] The read interference count is stored in the non-volatile register of the memory block controller, and the number of programmable loops is stored in the physically isolated loop marker cell.

[0060] Specifically, when a memory block is determined to be in a high-risk state, the memory block controller immediately activates the hot data identification process: First, the word line voltage monitoring circuit captures the voltage fluctuation signal during the read operation in real time. If the fluctuation amplitude exceeds the preset voltage stability threshold, the read interference count increment is triggered and the result is written to the non-volatile register. Simultaneously, the charge pump current sensor collects the injected pulse signal. After each charge verification is completed, the pulse counter accumulates the programming cycle count, and the accumulated value is stored in the cycle marker cell manufactured by deep well isolation technology. After the full data page parameters are collected, the hardware comparison module performs two comparisons in parallel: comparing the read interference count value in the non-volatile register with the preset read interference threshold, and comparing the programming cycle count in the cycle marker cell with the preset cycle threshold. When both parameters of the same data page exceed the limit, the address latch locks its physical address. Finally, the DMA controller arranges all locked addresses consecutively by block, generates a hot data identification table, and writes it to the dedicated cache area of ​​the flash translation layer.

[0061] Furthermore, the word line voltage monitoring circuit captures the real-time voltage signal during the read operation through the probe pin directly coupled to the word line metal layer of the memory cell. This signal is then input to the high-speed comparator after common-mode noise is eliminated by a differential amplifier. The voltage stability threshold is set by a precision adjustable resistor network to ±15% of the rated operating voltage of the memory cell. When the comparator detects that the fluctuation amplitude exceeds the limit, it immediately triggers a monostable multivibrator to generate a 50ns wide counting pulse. This pulse drives a 12-bit asynchronous counter to perform an accumulation operation. The counting result is written to a partitioned non-volatile register array composed of ferroelectric memory through a tri-state buffer. The physical address of the register is strictly mapped to the row address of the memory block. At the same time, the counter overflow signal triggers the status register to be set, generating a read interference event flag for the main control chip to poll and read.

[0062] After the address latch completes the physical address locking of the high-wear data page within the high-risk memory block, channel 0 of the DMA controller receives the 32-bit address sequence output by the latch. The discrete addresses are then reorganized into blocks of continuous order by a hardware sorting circuit to generate a hot data identifier table aligned to 128-byte boundaries. This table is transmitted via a double data rate bus in an independent clock domain to a dedicated cache area in the flash memory conversion layer—this cache area consists of physically isolated SRAM modules. After receiving the DMA transfer completion interrupt signal, the memory controller writes a frame header identifier containing a parity check code to the cache area and simultaneously activates the write-protect latch to map the cache area to read-only status. Finally, the consistency between the written data and the latch source address is compared by a verification circuit, and the verification result is recorded in the bits of the status register.

[0063] In one embodiment, the step of relocating identified high-wear data pages to a reserved healthy storage block includes:

[0064] During the idle period of the solid-state drive, the physical address of the high-wear data page is loaded according to the hot data identification table, and a read command is sent to the storage block with high source risk, while a double redundant write is initiated to the target healthy storage block.

[0065] The data page erasure of the source high-risk storage block is triggered only when the target healthy storage block returns two sets of write verification signals that are consistent.

[0066] Update the address mapping table, bind the logical address of the migrated data page to the physical address of the healthy storage block, and generate a migration completion log.

[0067] As mentioned above, the healthy storage block is selected from a reserved pool independent of the user storage area, and the capacity of the reserved pool accounts for 0.5%-1% of the total capacity of the storage unit.

[0068] When the SSD enters an idle state, the controller chip loads the information of the high-wear data pages based on the physical address in the hot data identification table; simultaneously, the dual-channel migration engine is activated, specifically by sending a read enable signal to the row decoder of the source high-risk storage block, and driving two independent charge pumps to perform double-redundant writes to the target healthy storage block. The first write path completes data temporary storage via page buffer A before triggering the application of the programming voltage, while the second path performs verification readback comparison via page buffer B. Only when the verification comparators of both paths output a high level simultaneously, the AND gate circuit generates an erase trigger signal, setting the block erase voltage enable line of the source high-risk storage block. After charge discharge is completed, the mapping table update engine writes the mapping relationship between the original logical address and the physical address of the healthy storage block into the address mapping table latch, and records a migration completion log containing a timestamp and migration block fingerprint in the physical marker log area, with the log storage depth prioritizing the earliest record.

[0069] Reference Appendix Figure 2 This is a structural block diagram of a solid-state drive fault pre-diagnosis and data security self-repair processing device proposed in this invention. The device includes:

[0070] The first marking unit is used to monitor the block erase time of each storage block during the operation of the solid-state drive. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as a warning state.

[0071] The second marking unit is used to continuously monitor the block erase time change rate of storage blocks marked as warning status. If the block erase time change rate exceeds the accelerated degradation threshold, the storage block is determined to have entered a high-risk state.

[0072] The diagnostic unit is used to perform thermal data identification operations on the storage block in a high-risk state, and to identify data pages in the storage block whose wear value is higher than a set level;

[0073] The repair unit is used to migrate the identified high-wear data pages to a reserved healthy storage block, and after the migration is completed, to perform a write-protection locking operation on the storage block that was originally in a high-risk state.

[0074] During SSD power-on operation, the physical layer timer collects the block erase time of each storage block in real time. When the first marking unit of the main controller chip detects that the block erase time of a specific storage block exceeds the warning threshold for the first time, it immediately marks it as a warning state through the status register of the storage block controller. Subsequently, the second marking unit of the firmware starts a high-frequency monitoring mode, recording the block erase time of the warning-state storage blocks every 5 minutes. If the time increase exceeds the increase threshold in three consecutive recording cycles and the single increase is greater than 120% of the previous one, the storage block is determined to have entered a high-risk state and an interrupt signal is triggered. After the high-risk state is confirmed, the storage block controller retrieves the diagnostic data from the non-volatile register. The interruption count in the interruption unit is compared with the programming loop count in the loop marking unit. The preset interruption threshold and programming loop threshold are then compared, and the physical addresses of data pages that exceed the limits are written to the hot data identification table. During the hard disk idle period, the repair unit initiates a dual-channel verification migration based on the hot data identification table: while reading data pages from the source high-risk storage block, double redundancy writes are performed to the target healthy storage block. Only when the two sets of write verification signals are consistent after being verified by the voltage comparator are the corresponding data pages of the source block erased and the address mapping table updated. Immediately after the migration is completed, a write-prohibit command is sent to the high-risk storage block, a lock flag is written to the mapping table, and the lock time and high-risk parameters are recorded to the physical marking log area.

[0075] Furthermore, the diagnostic unit includes:

[0076] The memory sub-unit is used to record the read interference count and programming loop count of all data pages within the storage block;

[0077] The comparison sub-unit is used to compare the read interference count of each data page with the preset read interference threshold, and the number of programming loops with the preset loop threshold;

[0078] The marking subunit is used to mark data pages that simultaneously exceed the read interference threshold and the programming cycle threshold as high-wear data pages;

[0079] The statistics subunit is used to generate a hot data identifier table containing the physical addresses of all high-wear data pages.

[0080] Specifically, when a memory block is determined to be in a high-risk state, the memory block controller immediately activates the hot data identification process: First, the word line voltage monitoring circuit captures the voltage fluctuation signal during the read operation in real time. If the fluctuation amplitude exceeds the preset voltage stability threshold, the read interference count increment is triggered and the result is written to the memory sub-cell of the non-volatile register. Simultaneously, the charge pump current sensor collects the injected pulse signal. After each charge verification is completed, the pulse counter accumulates the number of programming cycles, and the accumulated value is stored in the cycle marker cell manufactured by deep well isolation technology. After the full data page parameters are collected, the hardware comparison module performs two comparisons in parallel: comparing the read interference count value in the non-volatile register with the preset read interference threshold, and comparing the number of programming cycles in the cycle marker cell with the preset cycle threshold. When both parameters of the same data page exceed the limit, the address latch locks its physical address. Finally, the DMA controller arranges all locked addresses consecutively by block, generates a hot data identification table, and writes it to the dedicated cache area of ​​the flash translation layer.

[0081] In addition, the repair unit includes:

[0082] The write subunit is used to load the physical address of the high-wear data page according to the hot data identification table during the idle period of the solid-state drive, send a read command to the storage block with high source risk, and simultaneously initiate double redundant write to the target healthy storage block.

[0083] The erase subunit is used to trigger the erasure of the data page corresponding to the source high-risk storage block only when the target healthy storage block returns two sets of write verification signals that are consistent.

[0084] The update sub-unit is used to update the address mapping table, bind the logical address of the migrated data page to the physical address of the healthy storage block, and generate a migration completion log.

[0085] Specifically, when the SSD enters an idle state, the controller chip loads the information of the high-wear data page based on the physical address of the hot data identification table; simultaneously, the dual-channel migration engine is activated, specifically by sending a read enable signal to the row decoder of the source high-risk storage block, and driving two independent charge pumps to perform double-redundant writing to the target healthy storage block. The first write path completes data temporary storage through page buffer A and triggers the application of programming voltage, while the second path performs verification readback comparison through page buffer B; only when the verification comparators of both paths output a high level simultaneously, the AND gate circuit generates an erase trigger signal to set the block erase voltage enable line of the source high-risk storage block; after charge discharge is completed, the mapping table update engine writes the mapping relationship between the original logical address and the physical address of the healthy storage block into the address mapping table latch, and records a migration completion log containing a timestamp and migration block fingerprint in the physical marker log area, with the log storage depth prioritizing the earliest record.

[0086] The present invention is applied to a solid-state drive, which is equipped with the aforementioned device to specifically perform the above-described method.

[0087] In summary, during the operation of the solid-state drive (SSD), the block erase time of each storage block is monitored. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as being in a warning state. The rate of change of the block erase time of the storage blocks marked as being in a warning state is continuously monitored. If the rate of change of the block erase time exceeds the accelerated degradation threshold, the storage block is determined to be in a high-risk state. A hot data identification operation is performed on the storage blocks in the high-risk state to identify data pages with wear values ​​higher than a set level. The identified high-wear data pages are then migrated to reserved healthy storage blocks. After the migration is completed, a write protection lock operation is performed on the original high-risk storage blocks. This achieves the effect of proactive data security protection and synergistic optimization of storage lifespan.

[0088] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for pre-diagnosis of solid-state drive (SSD) faults and self-repair of data security, characterized in that, Including the following steps: During the operation of the solid-state drive, the block erase time of each storage block is monitored. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as a warning state. The block erase time change rate of storage blocks marked as warning status is continuously monitored. If the block erase time change rate exceeds the accelerated degradation threshold, the storage block is determined to be in a high-risk state. Perform a thermal data identification operation on the storage block in a high-risk state to identify data pages within the storage block whose wear value is higher than a set level; The identified high-wear data pages are migrated to reserved healthy storage blocks. After the migration is completed, write protection locking is performed on the original high-risk storage blocks. The steps of migrating the identified high-wear data pages to reserved healthy storage blocks include: during the idle period of the solid-state drive, loading the physical address of the high-wear data page according to the hot data identification table, sending a read command to the source high-risk storage block, and simultaneously initiating a double-redundant write to the target healthy storage block; triggering the erasure of the corresponding data page of the source high-risk storage block only when the target healthy storage block returns two sets of write verification signals that are consistent; updating the address mapping table, binding the logical address of the migrated data page to the physical address of the healthy storage block, and generating a migration completion log.

2. The solid-state drive fault pre-diagnosis and data security self-repair method according to claim 1, characterized in that, The warning threshold is 1.5 to 3 times the baseline erase time, which is obtained based on the statistical analysis of historical failure samples of the same batch of solid-state drives.

3. The solid-state drive fault pre-diagnosis and data security self-repair method according to claim 1, characterized in that, Performing a hot data identification operation on the storage block in a high-risk state, and identifying data pages within the storage block whose wear value exceeds a set level, includes the following steps: Record the read interference count and programming loop count for all data pages within the storage block; Compare the read interference count of each data page with the preset read interference threshold, and the number of programming loops with the preset loop threshold; Data pages that simultaneously exceed both the read interference threshold and the programming cycle threshold are marked as high-wear data pages; Generate a hot data identifier table containing the physical addresses of all high-wear data pages.

4. The solid-state drive fault pre-diagnosis and data security self-repair method according to claim 3, characterized in that, The steps of recording the read interference count and programming loop count of all data pages within the storage block include: In response to word line voltage fluctuation signals, read operations exceeding the voltage stability threshold are converted into read interference count increments; Capture the injection pulse signal of the charge pump and convert the programming operation to complete charge verification into an increment of programming loop count; The read interference count is stored in the non-volatile register of the memory block controller, and the number of programmable loops is stored in the physically isolated loop marker cell.

5. The solid-state drive fault pre-diagnosis and data security self-repair method according to claim 1, characterized in that, The healthy storage blocks are selected from a reserved pool independent of the user storage area, and the capacity of the reserved pool accounts for 0.5%-1% of the total capacity of the storage units.

6. A solid-state drive fault pre-diagnosis and data security self-repair processing device, characterized in that, include: The first marking unit is used to monitor the block erase time of each storage block during the operation of the solid-state drive. When the block erase time of one or more storage blocks exceeds the warning threshold for the first time, the storage block is marked as a warning state. The second marking unit is used to continuously monitor the block erase time change rate of storage blocks marked as warning status. If the block erase time change rate exceeds the accelerated degradation threshold, the storage block is determined to have entered a high-risk state. The diagnostic unit is used to perform thermal data identification operations on the storage block in a high-risk state, and to identify data pages in the storage block whose wear value is higher than a set level; The repair unit is used to migrate identified high-wear data pages to reserved healthy storage blocks. After the migration is completed, write protection locking is performed on the original high-risk storage blocks. The step of migrating the identified high-wear data pages to the reserved healthy storage blocks includes: during the idle period of the solid-state drive, loading the physical address of the high-wear data page according to the hot data identification table, sending a read command to the source high-risk storage block, and simultaneously initiating a double-redundant write to the target healthy storage block; triggering the erasure of the corresponding data page of the source high-risk storage block only when the target healthy storage block returns two sets of write verification signals that are consistent; updating the address mapping table, binding the logical address of the migrated data page to the physical address of the healthy storage block, and generating a migration completion log.

7. The solid-state drive fault pre-diagnosis and data security self-repair processing device according to claim 6, characterized in that, The diagnostic unit includes: The memory sub-unit is used to record the read interference count and programming loop count of all data pages within the storage block; The comparison sub-unit is used to compare the read interference count of each data page with the preset read interference threshold, and the number of programming loops with the preset loop threshold; The marking subunit is used to mark data pages that simultaneously exceed the read interference threshold and the programming cycle threshold as high-wear data pages; The statistics subunit is used to generate a hot data identifier table containing the physical addresses of all high-wear data pages.

8. The solid-state drive fault pre-diagnosis and data security self-repair processing device according to claim 7, characterized in that, The repair unit includes: The write subunit is used to load the physical address of the high-wear data page according to the hot data identification table during the idle period of the solid-state drive, send a read command to the storage block with high source risk, and simultaneously initiate double redundant write to the target healthy storage block. The erase subunit is used to trigger the erasure of the data page corresponding to the source high-risk storage block only when the target healthy storage block returns two sets of write verification signals that are consistent. The update sub-unit is used to update the address mapping table, bind the logical address of the migrated data page to the physical address of the healthy storage block, and generate a migration completion log.

9. A solid-state drive, characterized in that, The device according to any one of claims 6-8 is deployed to specifically perform the method according to any one of claims 1-5.

Images