Fault prediction method, apparatus and device for storage device, and storage medium

Abstract

The embodiment of the invention discloses a fault prediction method and device for storage equipment, equipment and a storage medium, which are used for improving the fault prediction precision and reliability of the storage equipment. The method provided by the embodiment of the invention comprises the following steps: acquiring working state data of each electronic device in the storage equipment at each moment, wherein different data sampling types under the same electronic device correspond to different working state data; and calculating a state evaluation result of the storage device according to the working state data of each electronic device at each moment, and determining a fault degree of the storage device based on the state evaluation result. According to the embodiment of the invention, whether the storage equipment is in a good or bad state or not is judged, and the health condition of the storage equipment is deeply analyzed through the working state data of each electronic device, so that the fault prediction precision is remarkably improved, and meanwhile, the performance reduction trend of the storage equipment can be identified in advance; and the read-write failure risk caused by performance degradation is prevented.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of storage devices, and in particular to a method, apparatus, device and storage medium for predicting a fault of a storage device. Background Art

[0002] With the rapid development of the information age, storage devices have become a vital component of information systems. The performance and reliability of storage devices directly affect the stability of the system and user experience, so it is of great significance to predict their failures and manage their lifespan, so as to identify potential failure risks in advance and avoid data loss and equipment damage.

[0003] In the related art, fault prediction usually monitors the status of physical blocks in storage devices (such as the number of bad blocks, read errors, write times, etc.) to determine whether the storage device has a risk of failure. However, these prediction methods only focus on whether the storage device is in a normal ("good") or faulty ("bad") state, ignoring the problem of reduced life of components inside the storage device, making it difficult to meet the read and write speed or stability requirements of the application scenario. Therefore, the storage device fault prediction methods in the prior art lack comprehensiveness and accuracy. Summary of the invention

[0004] Based on the above problems, the embodiments of the present application provide a storage device fault prediction method, apparatus, device and storage medium, the purpose of which is to timely identify performance degradation trends and improve the comprehensiveness and accuracy of storage device fault prediction.

[0005] In a first aspect, an embodiment of the present application provides a method for predicting a failure of a storage device, comprising:

[0006] Acquire the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data;

[0007] Calculating a status evaluation result of the storage device according to the working status data of each electronic device at each moment;

[0008] The failure degree of the storage device is determined according to the status evaluation result of the storage device.

[0009] In one embodiment, the electronic device includes at least one of a memory storage device, a serial interface, a flash memory device, a bus interface, a system on chip, and a capacitor; the data sampling type of the memory storage device includes at least one of an error count and a continuous working time; the data sampling type of the serial interface includes at least one of the number of erase and write times, a temperature, and a continuous working time; the data sampling type of the flash memory device includes at least one of the number of erase and write times, a temperature, and a continuous working time; the data sampling type of the bus interface includes the number of signal transmission errors; the data sampling type of the system on chip (SoC) includes at least one of the number of restarts and a temperature; the data sampling type of the capacitor includes a capacitance and a capacitance instantaneous state value.

[0010] In one embodiment, the calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes:

[0011] Based on each of the data sampling types of each of the electronic devices, determining the target working state data with the largest value among the working state data at each moment;

[0012] Based on the target working state data of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a state evaluation sample of each electronic device;

[0013] The status evaluation result of the storage device is obtained by performing calculation according to the status evaluation samples of each of the electronic components and the weight of each of the electronic components.

[0014] In one embodiment, the calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes:

[0015] Based on each of the data sampling types of each of the electronic devices, calculating the data change rate of the working status data at every two adjacent moments to obtain a data change rate set;

[0016] Determining a target data change rate with the largest value in the data change rate set;

[0017] Based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a state evaluation sample of each electronic device;

[0018] The status evaluation result of the storage device is obtained by performing calculation according to the status evaluation samples of each of the electronic components and the weight of each of the electronic components.

[0019] In one embodiment, the calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes:

[0020] The calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes:

[0021] Based on each of the data sampling types of each of the electronic devices, determining the target working state data with the largest value among the working state data at each moment;

[0022] Based on the target working state data of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a first state evaluation sample of each electronic device;

[0023] Calculating according to the first state evaluation samples of each of the electronic devices and the weight of each of the electronic devices to obtain a first evaluation result of the storage device;

[0024] Based on each of the data sampling types of each of the electronic devices, calculating the data change rate of the working status data at every two adjacent moments to obtain a data change rate set;

[0025] Determining a target data change rate with the largest value in the data change rate set;

[0026] Based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a second state evaluation sample of each electronic device;

[0027] Calculating according to the second state evaluation samples of each of the electronic devices and the weight of each of the electronic devices to obtain a second evaluation result of the storage device;

[0028] A status evaluation result of the storage device is determined according to the first evaluation result and the second evaluation result.

[0029] In one embodiment, determining the failure degree of the storage device according to the status evaluation result of the storage device includes:

[0030] According to a plurality of interval ranges divided by a plurality of preset state evaluation result thresholds, selecting the interval range to which the state evaluation result belongs; wherein each of the interval ranges corresponds to a preset fault degree;

[0031] The preset fault degree corresponding to the interval range to which the status evaluation result belongs is used as the fault degree of the storage device.

[0032] In one embodiment, the method further comprises:

[0033] Determining whether to perform fault location according to the fault degree of the storage device;

[0034] If the fault degree exceeds a preset fault degree threshold, determining the electronic device to be maintained according to a weighted result corresponding to the maximum data change rate of each electronic device;

[0035] According to the electronic component to be maintained, a maintenance strategy corresponding to the electronic component to be maintained is executed.

[0036] In a second aspect, an embodiment of the present application further provides a fault prediction device for a storage device, comprising:

[0037] An acquisition unit, used to acquire the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data;

[0038] A calculation unit, used for calculating the status evaluation result of the storage device according to the working status data of each electronic device at each moment;

[0039] A determination unit is used to determine the failure degree of the storage device according to the status evaluation result of the storage device.

[0040] In a third aspect, an embodiment of the present application further provides a storage device, including:

[0041] CPU, memory, input and output interfaces;

[0042] The memory is a short-term storage memory or a persistent storage memory;

[0043] The central processing unit is configured to communicate with the memory and execute instruction operations in the memory to perform any one of the above-mentioned methods for predicting failures of storage devices.

[0044] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the fault prediction method of any one of the storage devices described above is executed.

[0045] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0046] The embodiment of the present application obtains the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data; the state evaluation result of the storage device is calculated according to the working status data of each electronic device at each moment, and the degree of failure of the storage device is determined based on the state evaluation result. The embodiment of the present application not only determines whether the storage device is in a "good" or "bad" state, but also more comprehensively and accurately evaluates the health status of each electronic device by analyzing the working status data of different data sampling types under different electronic devices, thereby significantly improving the accuracy of fault prediction, and at the same time, it can also identify the trend of storage device performance degradation in advance, and prevent the risk of read and write failures caused by performance degradation. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0048] Figure 1 A schematic diagram of a method for predicting a storage device failure provided in an embodiment of the present application;

[0049] Figure 2 A schematic diagram of the level of failure of a storage device provided in an embodiment of the present application;

[0050] Figure 3 A schematic diagram of the structure of a fault prediction device for a storage device provided in an embodiment of the present application;

[0051] Figure 4 A schematic diagram of the structure of a storage device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0053] With the rapid development of the information age, storage devices have become a vital component of information systems. The performance and reliability of storage devices directly affect the stability of the system and user experience, so it is of great significance to predict their failures and manage their lifespan, so as to identify potential failure risks in advance and avoid data loss and equipment damage.

[0054] In the related art, fault prediction usually monitors the status of physical blocks in storage devices (such as the number of bad blocks, read errors, write times, etc.) to determine whether the storage device has a risk of failure. However, these prediction methods only focus on whether the storage device is in a normal ("good") or faulty ("bad") state, ignoring the problem of reduced life of components inside the storage device, making it difficult to meet the read and write speed or stability requirements of the application scenario. Therefore, the storage device fault prediction methods in the prior art lack comprehensiveness and accuracy.

[0055] Based on this, in each embodiment of the present application, the working status data of each electronic device in the storage device is collected, the status evaluation result of the storage device is calculated, and then the degree of fault is determined. When the degree of fault exceeds the preset threshold, the electronic device that needs maintenance is located and the corresponding maintenance strategy is executed, so as to timely identify the risk of storage device performance degradation, and improve the accuracy and comprehensiveness of fault prediction.

[0056] The various embodiments of the present application are described in further detail below in conjunction with the accompanying drawings.

[0057] The fault prediction method provided in the embodiment of the present application can be applied to the scenario of predicting whether a storage device is faulty, and the storage device can be a mechanical hard disk (HDD, Hard Disk Drive), a solid state drive (SSD, Solid State Drive), a high-speed serial computer expansion bus standard solid state drive (Peripheral component Interconnect Express Solid State Drive, PCIe SSD), an embedded multimedia card (eMMC, Embedded MultiMedia Card), a universal flash storage (UFS, Universal Flash Storage), a flash array (Flash Array), an on-chip memory (OCM, On-Chip Memory), an embedded storage device (ESD, Embedded Storage Device), a memory card (Memory Card, such as a secure digital (SD, Secure Digital) card, a compact flash (CF, Compact Flash) card), etc., and the specifics are not limited here. For the convenience of description, the following is only a schematic illustration taking SSD as an example.

[0058] See also Figure 1 , an embodiment of the present application provides a method for predicting a fault of a storage device, the method comprising steps S101-S103.

[0059] S101: Acquire the working status data of each electronic device in the storage device at each moment.

[0060] It should be noted that different data sampling types under the same electronic device correspond to different working status data; specifically, the current working status data of each electronic device in the storage device under different data collection types can be collected at each time point (i.e., "each moment"), and the working status data reflects the health status or performance of each electronic device, such as error count, number of erases and writes, capacitance value, etc. Through this real-time or periodic collection method, the working status reflected by each electronic device at different time points and different angles can be obtained, providing a data basis for subsequent storage device status evaluation and fault prediction.

[0061] It should be noted that the "moment" mentioned in the embodiments of the present application can be understood as the moment of data sampling, which can be at certain specific time points. For example, it can be a specific time unit, such as seconds, minutes, hours, etc., depending on the time interval of data sampling, which is not specifically limited in the embodiments of the present application.

[0062] S102: Calculate the status evaluation result of the storage device according to the working status data of each electronic device at each moment.

[0063] The status evaluation result of the storage device reflects the overall health status or performance level of the storage device. By summarizing and analyzing the working status data of each electronic device under different data collection types, and combining the weights of each electronic device and each data collection type, the working status data is multiplied by the corresponding weight and then accumulated. Therefore, the current operating status and health of the storage device can be evaluated according to the obtained status evaluation results, providing a basis for subsequent fault prediction and maintenance decisions. Alternatively, a large amount of historical data can be collected to establish a regression model or a classification model based on machine learning methods, and the working status data of each electronic device under different data collection types can be used as input to predict the health score or status evaluation level of the device. In this way, more complex relationships between the working status data of different electronic devices can be mined with the help of machine learning to improve the prediction accuracy. In addition, when analyzing the working status data of each electronic device, the non-numerical variables in the working status data of each electronic device at each moment can be digitized, standardized, and regularized, and non-abnormal information can also be eliminated.

[0064] S103: Determine the failure degree of the storage device according to the status evaluation result of the storage device.

[0065] The specific implementation method of judging the degree of device failure based on the status evaluation result of the storage device calculated above can be to compare the status evaluation result with multiple preset fault degree thresholds, and classify the fault degree into different fault levels (for example: normal, warning, serious, failure). For example: a status score of 90 points or more is "normal", 70-90 is "warning", 50-70 is "serious", and below 50 is "failure". Each level corresponds to different processing measures, such as no operation is required for normal status, monitoring for warning status, warning for serious status, and repair or replacement is recommended for failure status.

[0066] In another feasible embodiment, the degree of fault can be determined in combination with the change trend of the status evaluation result (such as the speed of increase or decrease). If the evaluation result continues to decrease or the rate of change increases, even if the current status evaluation result is not low, it may indicate that a fault is about to occur, and the system can issue an early warning.

[0067] In another feasible embodiment, the classification model may be trained using historical data, with the status evaluation result as input and the fault level of the equipment as output.

[0068] In another feasible embodiment, the status evaluation result may be directly converted into a corresponding fault degree through a preset mapping table.

[0069] The embodiment of the present application obtains the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data; the state evaluation result of the storage device is calculated according to the working status data of each electronic device at each moment, and the degree of failure of the storage device is determined based on the state evaluation result. The embodiment of the present application not only determines whether the storage device is in a "good" or "bad" state, but also more comprehensively and accurately evaluates the health status of each electronic device by analyzing the working status data of different data sampling types under different electronic devices, thereby significantly improving the accuracy of fault prediction, and at the same time, it can also identify the trend of storage device performance degradation in advance, and prevent the risk of read and write failures caused by performance degradation.

[0070] In one embodiment, the electronic device includes at least one of a memory storage, a serial interface, a flash memory, a bus interface, a system on chip, and a capacitor; the data sampling type of the memory storage includes at least one of an error count and a continuous working time; the data sampling type of the serial interface includes at least one of the number of erase and write times, a temperature, and a continuous working time; the data sampling type of the flash memory includes at least one of the number of erase and write times, a temperature, and a continuous working time; the data sampling type of the bus interface includes the number of signal transmission errors; the data sampling type of the system on chip includes at least one of the number of restarts and a temperature; the data sampling type of the capacitor includes the capacitance and the instantaneous state value of the capacitance.

[0071] In the fault prediction of the prior art, the "good or bad" of the device is mainly judged based on the physical block status of the storage device, or the fault prediction of the non-volatile memory status of the storage device is performed based on the machine learning algorithm, thus failing to comprehensively evaluate the health status of each component in the storage device. The fault prediction method of the prior art tends to ignore the operating status of other key components, while in actual applications, the storage device is usually composed of multiple key electronic components, including capacitors, system-on-chip SoC, bus interface, flash memory, serial interface, memory, etc. Among them, the bus interface may include high-speed serial computer expansion bus standard (PCIe, Peripheral Component Interconnect Express), serial advanced technology attachment (SATA, Serial Advanced Technology Attachment), universal serial bus (USB, Universal Serial Bus) and non-volatile memory express interface (NVMe, Non-Volatile Memory Express); flash memory may include NAND Flash (flash memory structure with NOT-AND logic design), NOR Flash (flash memory structure with NOT-OR logic design), eMMC, UFS. Serial interfaces may include Serial Peripheral Interface (SPI), Universal Asynchronous Receiver / Transmitter (UART), Recommended Standard 232 (RS-232), and Controller Area Network (CAN). Memory storage includes Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Read-Only Memory (ROM), and Electrically Erasable Programmable Read-Only Memory (EEPROM). Problems with any electronic device may cause the storage device to malfunction or become unusable.

[0072] In response to this problem, the embodiment of the present application comprehensively monitors all key electronic devices in the storage device, and based on the working status data of these electronic devices under different data collection types and preset algorithms, comprehensively evaluates the health status of the equipment to predict the risk of failure. For example, by monitoring the continuous working time and error count of the memory storage (such as uncorrectable error count (UNC, Uncorrectable Error Count), error count that cannot be corrected by the error correction code mechanism (UECC, Uncorrectable Error Correction Code Error Count)); by monitoring the serial interface, the number of erase and write times, temperature and continuous working time of the memory connected to the serial interface can be obtained; the number of erase and write times, temperature and continuous working time of the flash memory are monitored, the number of restarts and temperature of the system on chip are monitored, and the capacitance value and instantaneous state value of the capacitor are monitored to judge the health status and failure risk of these electronic devices.

[0073] It is important to understand that regarding the health monitoring of memory storage, the UNC count indicates that an uncorrectable error has occurred in the storage, which may result in data loss or read anomalies. Therefore, monitoring the UNC count helps to take measures to avoid data loss before errors accumulate to a certain level. The UECC count records errors that cannot be corrected by the ECC mechanism, reflecting the deterioration of the health of the storage. If the UECC count is too high, it means that the risk of storage failure has increased. By monitoring the UECC, the storage can be replaced or the data can be migrated before the problem becomes serious.

[0074] The continuous working time of memory storage, serial interface connected storage and flash storage indicates that they will age after long-term operation. Especially for storage devices that are frequently read and written, monitoring the continuous working time helps to assess the degree of hardware aging and arrange maintenance or replacement in advance according to the service life to avoid premature damage to the equipment.

[0075] It is understandable that the number of times the memory and flash memory connected to the serial interface can be erased and written is limited. The more times the memory is erased and written, the higher the degree of wear and aging of the memory. Monitoring the number of erases and writes helps to determine the remaining life of the flash memory and prevent data loss due to excessive erases and writes. In addition, the high temperature of the memory and flash memory connected to the serial interface will accelerate the aging of the memory, resulting in a decrease in data retention and an increase in read and write errors. Therefore, by monitoring the temperature of the memory and flash memory connected to the serial interface, it can help to take heat dissipation measures at the appropriate time, thereby extending the service life of the device.

[0076] Regarding monitoring the number of SoC restarts and temperature, frequent SoC restarts usually indicate system instability or hardware problems, such as power failure, memory failure, etc. Monitoring the number of restarts can help identify abnormal behavior of the device, facilitating timely repair or replacement. Similarly, SoCs will experience performance degradation or failure in high temperature environments, so monitoring the temperature helps manage system heat dissipation, prevent overheating damage, and ensure stable operation of the SoC.

[0077] Regarding the capacitance value and instantaneous state value of capacitors, the capacitance value indicates the ability of capacitors to store charge. When the capacitance value drops significantly, it means that the capacitor has leakage or aging problems, so the health of the capacitor can be predicted and replaced when necessary. The instantaneous state value of the capacitor (such as 0 for problems and 1 for normal) is a simplified health status indication. By monitoring the instantaneous state value in real time, when an abnormality is detected, the system can take timely measures to avoid equipment shutdown or damage caused by capacitor failure.

[0078] Monitoring the number of signal transmission errors on the bus interface can effectively predict whether there is a signal quality problem. A high number of errors may indicate a degradation of signal quality or an interface failure, so timely monitoring is required to help maintain the reliability of the communication channel.

[0079] It should be noted that the data sampling type of each of the above-mentioned electronic devices is not limited to the listed items, but any one or more of them can be selected for sampling, which can be adjusted according to different storage devices, environmental conditions, actual needs or application scenarios.

[0080] The embodiments of the present application can comprehensively identify potential problems of each key electronic device in the storage device, not only focusing on the overall "good or bad" of the device, but also providing a refined status monitoring and fault prediction mechanism. By evaluating the working status data and the preset algorithm, the performance degradation trend can be identified before the device fails according to the working status data corresponding to each electronic device, and an effective maintenance strategy can be formulated, thereby significantly improving the stability and reliability of the device, reducing the occurrence of data loss and system failures, and reducing operation and maintenance costs.

[0081] In some specific embodiments, the specific implementation process of step S102 may include any of the following methods:

[0082] Method 1: Based on each data sampling type of each electronic device, determine the target working status data with the largest value in the working status data at each moment; based on the target working status data of each data sampling type in each electronic device and the weight of each data sampling type, obtain a status evaluation sample of each electronic device; calculate according to the status evaluation samples of each electronic device and the weight of each electronic device to obtain the status evaluation result of the storage device.

[0083] In the process of health assessment of storage devices, in order to obtain the overall status assessment results, the system monitors each data sampling type of each electronic component in the storage device (such as capacitors, SoC, DRAM, etc.) to obtain its working status data at different time points. By analyzing this data, it is possible to identify whether each electronic component has a fault or potential problem, and evaluate the overall status of the storage device.

[0084] In order to perform health assessment and fault prediction on storage devices, the state data with the largest value can be selected as the target working state data from the working state data of each data sampling type at different times for each electronic device. This target working state data with the largest value usually represents the most severe state of the electronic device during the sampling period, indicating the maximum load or the most serious health condition of the electronic device.

[0085] Specifically, in each sampling period, the working status data of different data sampling types of each electronic device in the storage device are collected and the status evaluation result of the device is obtained. The status evaluation result can be the status score (such as a health score from 0 to 100) or the graded status (such as good, sub-healthy, unhealthy, faulty, etc.) of the device in the sampling period. Non-numerical variables such as graded status can be digitized, standardized, and regularized, and used as the target variable (dependent variable) of the linear regression model.

[0086] The working state data of different data sampling types of each electronic device and the state evaluation results of the storage device collected in each sampling period are combined into a data sample set. The working state data of each data sampling type in each sample can be used as a feature variable (independent variable), and the corresponding state evaluation result can be used as a target variable.

[0087] Then, a linear regression model is established using the collected historical data sample set to analyze the impact of each data sampling type on the health status of the equipment.

[0088] For example, the linear regression model corresponding to the historical data samples and the storage device status evaluation results within the sampling period T1 can be expressed as:

[0089]

[0090] Among them, β0 is a constant term (intercept), which represents the basic value of state evaluation when all characteristic variables are zero; β11, β11, β31, β12, β22, β13...βm iRepresents the regression coefficient of each data sampling type under each electronic device in the sampling period T1 (that is, the weight of each data sampling type). For example, β11, β11, β31 are the data sampling types of the first electronic device; ε is the error term; the size of each regression coefficient reflects the influence of the data sampling type on the health status of the electronic device and even the storage device.

[0091] Finally, for each data sampling type, the regression coefficient obtained by the linear regression equation at different sampling periods is calculated, and the average value, median, etc. of the regression coefficient of different sampling periods are calculated to determine the weight corresponding to each data sampling type in the actual application; or the regression coefficient of each data sampling type can be solved by the least squares method or gradient descent method.

[0092] Next, based on the target working state data of each data sampling type of each electronic device and combined with the weight of each data sampling type, the state evaluation sample of the electronic device is calculated. The weight of each data sampling type reflects the degree of influence of the working state data of this type on the health of the electronic device. The greater the weight, the greater the influence of the data sampling type on the health of the device. Then, based on the state evaluation samples of each electronic device and the corresponding weights, the system performs weighted calculations on these data and finally obtains the state evaluation result of the storage device.

[0093] Specifically, a weight may be assigned to each electronic device, and the weight of the electronic device may be determined according to the degree of influence of the electronic device on the storage device.

[0094] For example, the weight W of each electronic device is i It can be obtained by following the steps below:

[0095] 1. Set the total weight and:

[0096] Set a weight W for each electronic component i , so that the sum of the weights of all electronic devices is:

[0097]

[0098] 2. Determine the multiple parameter C i :

[0099] In order to reasonably distribute the weights of different electronic components, the minimum weight W1 is selected as the benchmark weight, and the multiple parameter C is used. i To express the weight of other electronic components relative to W1, that is: W1 = C1*W1, W2 = C2*W1, W3 = C3*W1, W4 = C4*W1, W5 = C5*W1, W6 = C6*W1. Among them, C1 = 1, C i >1(when i>1). To ensure that other weights are greater than W1.

[0100] Furthermore, the weight W corresponding to each electronic device can be set i The order of size is W6>W5>W4>W3>W2>W1. For example, the capacitor corresponds to W6, the SoC corresponds to W5, the memory storage corresponds to W4, the bus interface corresponds to W3, the serial interface corresponds to W2, and the flash memory corresponds to W1.

[0101] In order to reasonably determine the weight ratio of each component, it is necessary to allocate a multiple parameter C according to the degree of influence of each electronic device on the overall state of the storage device. i , please refer to the following examples for details:

[0102] Capacitor: Once the capacitor is damaged, there is no intermediate value. It is scrapped once damaged, which has a great impact on the storage device, so C6=3.5 is set.

[0103] SoC: SoC is the main controller of the storage device. Damage to it will affect all functions of the entire storage device, so C5=3 is set.

[0104] Memory storage: Memory storage such as DRAM contains cached data. Damage to this electronic device may affect the data in the flash memory device, causing downtime; in some cases, it may not affect the data, so let C4=2.5.

[0105] Bus interface: Bus interface communication abnormalities will affect the communication between the host and the storage device, so set C3=2.

[0106] Serial interface: The memory connected to the serial interface mainly stores metadata and may only affect power on and off or SPI operations, so C2=1.5 is set.

[0107] Flash memory: Flash memory, such as Nand Flash, is the main storage medium. It has more fault tolerance processing and has less impact on the health status of the storage device, so C1=1.

[0108] 3. Calculate the actual weight of each electronic device:

[0109] Determine the multiple parameter C of each electronic device i Then, calculate C i Total: 3.5+3+2.5+2+1.5+1=13.5.

[0110] Then calculate the weight W of each electronic device according to the following formula i :

[0111]

[0112] For example, for memory storage, the weight is

[0113] Through the above steps, the weight corresponding to each electronic device can be obtained. Finally, the state data sample value of each electronic device is multiplied by the weight of the electronic device to obtain the state evaluation sample of each electronic device:

[0114]

[0115] Among them, R i result represents the state evaluation sample of the i-th electronic device; represents the i-th electronic device at each time point (t1, t2, ..., t n-1 ) Select the maximum value of the collected working status data; represents the weighted state evaluation value of the u-th electronic device in data acquisition type D1; It represents the i-th electronic device in the data acquisition class Dm at each time point (t1, t2, ..., t n-1 ) Select the maximum value of the collected working status data; It represents the weighted state evaluation value of the i-th electronic device under the data acquisition type Dm.

[0116] By multiplying the status evaluation sample of each electronic device by the weight of each electronic device and then adding them together, the status evaluation result of the storage device can be obtained. The specific calculation formula is as follows:

[0117]

[0118] Among them, Result represents the status evaluation result of the storage device; R i result represents the state evaluation sample of the i-th electronic device; W i represents the weight of the i-th electronic device.

[0119] Method 2: Based on each data sampling type of each electronic device, calculate the data change rate of the working status data at every two adjacent moments to obtain a data change rate set; determine the target data change rate with the largest value in the data change rate set; based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type, obtain a state evaluation sample of each electronic device; calculate according to the state evaluation samples of each electronic device and the weight of each electronic device to obtain the state evaluation result of the storage device.

[0120] The embodiment of the present application mainly evaluates the health status of the storage device by the data change rate of the working status data of each data sampling type of each electronic device at adjacent moments. Specifically, for each electronic component, the system calculates the data change rate of the working status data at every two adjacent moments to quantify the speed of the working status data change of each data sampling type under the i-th electronic device in a short period of time. The specific calculation formula can be referred to:

[0121]

[0122] in, represents the i-th electronic device at different time points (t1, t2, ..., t n-1 ) collected working status data; T represents the time interval between two adjacent moments; represents the square of the data change rate of the working state data of the i-th electronic device between time t1 and t2 under the Dm data sampling type, represents the square of the data change rate of the i-th electronic device in the Dm data sampling type working state data between time t2 and t3, ..., is the working state data of the ith electronic device in the Dm data sampling type at time t n-2 and t n-1 The square of the rate of change between .

[0123] It can be understood that taking the square of the data change rate of the working status data between two moments can widen the change gap, thereby better reflecting the difference in the change speed within a unit time, and facilitating the subsequent determination of the degree of failure of the storage device.

[0124] During the entire sampling period of each electronic device, the maximum value of the data change rate represents the most drastic change of the component during the sampling period. A drastic change means that the electronic device may be in an abnormal state or face potential failure risks. By detecting the peak value of the change rate, possible problems in the equipment can be discovered in a timely manner, which is convenient for early warning and preventive maintenance. Therefore, the embodiment of the present application selects the maximum data change rate as the status evaluation sample of the data sampling type of the electronic device, which is used to subsequently determine the status evaluation result of the storage device.

[0125] The formula for the data change rate of different data sampling types for each electronic device can be expressed as:

[0126]

[0127] The above formula is used to obtain the data change rate with the maximum value for each data sampling type D1, D2, ..., Di under the i-th electronic device. in, represents the i-th electronic device at different times (t1, t2, ..., t n-1 ) data change rate; represents the i-th electronic device at different times (t1, t2, ..., t n-1 ) data change rate; represents the i-th electronic device at different times (t1, t2, ..., t n-1 ) data change rate; data change rate of different data sampling types of the i-th electronic device Combined into a set of data change rates of the electronic device.

[0128] To find the target data change rate with the largest value in the above data change rate set, the formula can be expressed as:

[0129]

[0130] Among them, ΔR i result represents the state evaluation sample of the i-th electronic device, It represents the product of the target data change rate with the maximum value under each data sampling type under the i-th electronic device and the corresponding weight of each data sampling type, The state evaluation sample representing the i-th electronic device is determined by the data sampling type with the largest data change rate.

[0131] The status evaluation result of the storage device is obtained by calculating according to the status evaluation samples of each electronic device and the weight of each electronic device. The formula can be expressed as:

[0132]

[0133] Wherein, ΔResult represents the status evaluation result of the storage device; ΔR i result represents the state evaluation sample of the i-th electronic device; W i represents the weight of the i-th electronic device.

[0134] The embodiments of the present application can make the fault prediction of storage devices more focused on electronic devices with the most drastic changes and the greatest impact, which helps to timely identify potential fault risks in storage devices, thereby improving the accuracy and real-time performance of fault prediction.

[0135] Method three: Based on each data sampling type of each electronic device, determine the target working status data with the largest value in the working status data at each moment; based on the target working status data of each data sampling type in each electronic device and the weight of each data sampling type, obtain the first state evaluation sample of each electronic device; calculate according to the first state evaluation sample of each electronic device and the weight of each electronic device to obtain the first evaluation result of the storage device; for each data sampling type of each electronic device, calculate the data change rate of the working status data at every two adjacent moments to obtain a data change rate set; determine the target data change rate with the largest value in the data change rate set; based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type, obtain the second state evaluation sample of each electronic device; calculate according to the second state evaluation sample of each electronic device and the weight of each electronic device to obtain the second evaluation result of the storage device; determine the state evaluation result of the storage device based on the first evaluation result and the second evaluation result.

[0136] The embodiment of the present application mainly evaluates the failure degree of the storage device by combining the maximum value of the working status data of each electronic device and the maximum value of the data change rate.

[0137] It should be noted that, for the specific step details of the process of obtaining the first evaluation result of the storage device in method three, reference can be made to the relevant description of the implementation method of method one in the aforementioned embodiment; and, for the specific step details of the process of obtaining the second evaluation result of the storage device in this embodiment, reference can be made to the relevant description of the implementation method of method two in the aforementioned embodiment, which will not be repeated here.

[0138] The data obtained through the above steps are:

[0139] Flash memory: data sampling types are D11, D21, D31, state evaluation samples are R1result and ΔR1result; weight is W1.

[0140] Serial interface: data sampling types are D12, D22, D33, state evaluation samples are R2result and ΔR2result; weight is W2.

[0141] Bus interface: The data sampling type is D13, the status evaluation samples are R3result and ΔR3result; the weight is W3.

[0142] Memory storage: data sampling types are D14, D24, D34, state evaluation samples are R4result and ΔR4result; weight is W4.

[0143] SoC: data sampling types include D15, D25, and D35, state evaluation samples are R5result and ΔR5result; the weight is W5.

[0144] Capacitor: Data sampling types are D16 and D26, status evaluation samples are R6result and ΔR6result; weight is W6.

[0145] It should be noted that the above process of determining the status evaluation result of the storage device through the memory storage, serial interface, flash memory, bus interface, system on chip, and capacitor is only an example and does not have any limiting effect. The embodiment of the present application can also determine the specific implementation method of the status evaluation result of the storage device according to the electronic devices specifically related to the storage device. For example, the eMMC is not configured with a capacitor, so when calculating the status evaluation result of the eMMC, it can be calculated based on other electronic devices configured with the eMMC, such as the number of shoe shines of the flash memory, the number of restarts of the SoC connected to the eMMC, etc. Therefore, without departing from the concept of the present technical solution, any obvious replacement is within the scope of protection of the present invention.

[0146] After obtaining the first evaluation result Result and the second evaluation result ΔResult of the storage device, the status evaluation result of the storage device can be determined in the following ways:

[0147] (1) The final status evaluation result is obtained by performing a weighted average calculation on the first evaluation result and the second evaluation result. For example, the status evaluation result of the storage device = p1*first evaluation result Result+p2*second evaluation result ΔResult. Among them, p1 and p2 are the weights of the first evaluation result and the second evaluation result, respectively, which can be specifically set according to the degree of influence of the two on the health status of the storage device (for example, p1+p2=1), and the embodiments of the present application do not limit this.

[0148] (2) The maximum value between the first evaluation result and the second evaluation result is taken as the status evaluation result of the storage device, that is, Max(Result, ΔResult). The method of taking the maximum value is suitable for scenarios that are highly sensitive to failure risks and is conducive to ensuring that problems are discovered in a timely manner.

[0149] (3) The minimum value between the first evaluation result and the second evaluation result is taken as the final status evaluation result, that is, Max(Result, ΔResult). The method of taking the minimum value is suitable for evaluating the most stable status of the equipment and hoping to obtain a more optimistic health assessment result, especially when the system is relatively stable and failures do not occur frequently.

[0150] (4) By setting a threshold, it is determined whether the first evaluation result and the second evaluation result exceed the set threshold respectively, and the level of the status evaluation result is determined according to the condition. For example: if both are lower than threshold 1, the status evaluation result is "healthy"; if one of them is higher than threshold 1 but lower than threshold 2, the status evaluation result is "sub-healthy"; if both are higher than threshold 2, the status evaluation result is "faulty".

[0151] In one embodiment, the failure degree of the storage device is determined based on the status evaluation result of the storage device, including: selecting the interval range to which the status evaluation result belongs according to multiple interval ranges divided according to multiple preset status evaluation result thresholds; wherein each interval range corresponds to a preset failure degree; and using the preset failure degree corresponding to the interval range to which the status evaluation result belongs as the failure degree of the storage device.

[0152] The embodiment of the present application sets multiple preset status evaluation result thresholds Thresh1, Thresh2, Thresh3, etc. for the status evaluation result of the storage device. The definition of the Thresh threshold is determined by the test conditions and importance of the above-mentioned data sampling type and the degree of loss, and the embodiment of the present application is not limited. Each preset status evaluation result threshold represents an interval range, and each interval range corresponds to a preset fault degree (such as Level 1, Level 2, Level 3, Level 4, etc., or healthy, sub-healthy, unhealthy, faulty, etc.). The fault level of the device is determined based on which threshold interval the value of the status evaluation result falls into.

[0153] The specific classification is as follows:

[0154] Level 1 (Health): 0 to Thresh1, indicating that the device is in good condition.

[0155] Level 2 (Sub-health): Between Thresh1 and Thresh2, it means the device is in a sub-healthy state and the fault type needs to be further determined.

[0156] Level 3 (Unhealthy): Between Thresh2 and Thresh3, it indicates that the device is unhealthy. You need to confirm the cause of the fault and evaluate whether the device needs to be replaced.

[0157] Level 4 (fault): greater than Thresh3, indicating a device fault. It is recommended to replace or scrap the storage device.

[0158] For further information, please refer to Figure 2 , Figure 2It is a schematic diagram of the fault degree level of a storage device; it can be seen that the left side of the figure is the fault degree level corresponding to the status evaluation result Result based on the storage device, and the right side of the figure is the fault degree level corresponding to the status evaluation result ΔResult based on the storage device. It can be understood that corresponding to the method one or method two of determining the storage device status evaluation result in the embodiment of the present application, the obtained storage device status evaluation result can be directly brought into the above-mentioned interval range to determine the storage device failure degree corresponding to the storage device status evaluation result. As for method three, the fault degree can be further determined based on the first evaluation result Result and the second evaluation result ΔResult obtained, that is, the maximum value from the fault degrees determined according to Result and ΔResult is taken as the fault degree of the storage device, which can be expressed by the formula: Level = Max{Level Result , Level ΔResult}, where Level Result Indicates the fault severity level corresponding to the status evaluation result Result, Level ΔResult Indicates the fault severity level corresponding to the status evaluation result ΔResult.

[0159] The embodiment of the present application calculates Result and ΔResult respectively, classifies them in combination with thresholds, and then determines the final fault level by taking the maximum value, which can effectively capture the most serious state, ensure that the fault risk is discovered and handled in time, avoid the accumulation of potential problems leading to major faults, and reduce maintenance costs.

[0160] When the storage device reaches a certain level of failure, it is necessary to promptly determine the cause of the failure and locate the specific faulty electronic device in order to execute the corresponding maintenance strategy to ensure the normal operation of the system. Therefore, in one embodiment, it also includes: determining whether to perform fault location based on the failure level of the storage device. If the failure level is lower than the preset threshold, it is considered that the device is in normal condition or only has minor problems, and no further processing is required. If the failure level exceeds the preset failure level threshold, it means that the device has a more serious health problem, and fault location needs to be performed at this time to promptly determine the source of the failure.

[0161] During the fault location process, the system will analyze the maximum data change rate of each electronic device, and combine it with the weight of each electronic device to calculate the status evaluation sample of each electronic device. For details, please refer to the relevant description of obtaining the status evaluation sample of each electronic device in method two in the aforementioned embodiment, which will not be repeated here.

[0162] By analyzing these status evaluation samples, electronic devices that have a greater impact on the health of the equipment can be identified as "electronic devices to be maintained", and maintenance strategies corresponding to the electronic devices to be maintained can be executed. The maintenance strategy can be determined according to the degree of failure. For example, if the storage device is in a sub-healthy state, the system will locate the specific faulty electronic device and analyze the cause of the failure. More specifically, when a problem occurs in the SoC and causes the storage device to enter a sub-healthy state, the system will further analyze the working status data of the SoC under different data sampling types to identify the specific source of the fault. The data sampling types of the SoC may include temperature, number of restarts, etc. If the temperature data is found to be abnormal (such as the temperature value is too high or the temperature change rate is large), it can be determined that the maintenance strategy is cooling processing. Furthermore, the maintenance strategy may also include operations such as replacing faulty components, cleaning equipment, lowering the temperature, or optimizing load distribution to restore or improve the normal operating state of the equipment.

[0163] It should be pointed out that when the storage device is in an unhealthy state, automatic error handling operations may be performed inside the storage device, which may include data migration within the disk, reconstruction of the flash translation layer (FTL), re-downloading of backup firmware, etc., to minimize the impact of the fault on the data. If the error handling inside the storage device cannot solve the problem, the device can also actively initiate a data migration request to the host to transfer the data to other storage media to ensure data integrity and normal operation of the system. After detecting that the storage device is in an unhealthy state, when automatic error handling can be performed inside the storage device, fault location and fault cause analysis can be performed based on the status evaluation samples of each electronic device obtained by the above-mentioned fault prediction method.

[0164] Through the above method, the embodiment of the present application realizes the timely location and effective maintenance of storage device failures, which can not only identify key electronic components that may affect the operation of the device at an early stage, but also implement targeted maintenance strategies according to different fault types and severity, thereby improving the stability and reliability of the system. In addition, the internal error handling mechanism and external data migration request function of the device further enhance the data protection capability and reduce the risk of data loss caused by failures.

[0165] In order to implement the storage device fault prediction method of the embodiment of the present application, the embodiment of the present application also provides a storage device fault prediction device, such as Figure 3 As shown, the device comprises:

[0166] An acquisition unit 301 is used to acquire the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data;

[0167] A calculation unit 302, configured to calculate a status evaluation result of the storage device according to the working status data of each electronic device at each moment;

[0168] The determination unit 303 is configured to determine the failure degree of the storage device according to the status evaluation result of the storage device.

[0169] In one embodiment, the electronic device includes at least one of a memory storage, a serial interface, a flash memory, a bus interface, a system on chip, and a capacitor; the data sampling type of the memory storage includes at least one of an error count and a continuous working time; the data sampling type of the serial interface includes at least one of the number of erase and write times, a temperature, and a continuous working time; the data sampling type of the flash memory includes at least one of the number of erase and write times, a temperature, and a continuous working time; the data sampling type of the bus interface includes the number of signal transmission errors; the data sampling type of the system on chip includes at least one of the number of restarts and a temperature; the data sampling type of the capacitor includes capacitance and a capacitance instantaneous state value.

[0170] In one embodiment, the device further comprises: a processing unit; the processing unit is used to determine the target working state data with the largest value in the working state data at each moment based on each data sampling type of each electronic device.

[0171] The calculation unit 302 is further configured to obtain a state evaluation sample of each electronic device based on the target working state data of each data sampling type in each electronic device and the weight of each data sampling type;

[0172] The calculation unit 302 is further configured to perform calculations based on the status evaluation samples of each of the electronic components and the weights of each of the electronic components to obtain a status evaluation result of the storage device.

[0173] In one embodiment, the calculation unit 302 is further configured to calculate the data change rate of the working state data at every two adjacent moments based on each data sampling type of each electronic device to obtain a data change rate set;

[0174] The determining unit 303 is further configured to determine a target data change rate having a maximum value in the data change rate set;

[0175] The calculation unit 302 is further configured to obtain a state evaluation sample of each electronic device based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type;

[0176] The calculation unit 302 is further configured to perform calculations based on the status evaluation samples of each of the electronic components and the weights of each of the electronic components to obtain a status evaluation result of the storage device.

[0177] In one embodiment, the processing unit is further used to determine the target working state data with the largest value among the working state data at each moment based on each data sampling type of each electronic device;

[0178] The calculation unit 302 is further configured to obtain a first state evaluation sample of each electronic device based on the target working state data of each data sampling type in each electronic device and a weight of each data sampling type;

[0179] The calculation unit 302 is further used to calculate according to the first state evaluation samples of each of the electronic devices and the weight of each of the electronic devices to obtain a first evaluation result of the storage device;

[0180] The determining unit 303 is further configured to determine a target data change rate having a maximum value in the data change rate set;

[0181] The calculation unit 302 is further configured to obtain a second state evaluation sample of each of the electronic devices based on the target data change rate of each of the data sampling types in each of the electronic devices and the weight of each of the data sampling types;

[0182] The calculation unit 302 is further used to calculate according to the second state evaluation samples of each of the electronic devices and the weight of each of the electronic devices to obtain a second evaluation result of the storage device;

[0183] The determining unit 303 is further configured to determine a status evaluation result of the storage device according to the first evaluation result and the second evaluation result.

[0184] In one embodiment, the processing unit is further configured to select the interval range to which the status evaluation result belongs according to a plurality of interval ranges divided by a plurality of preset status evaluation result thresholds; wherein each of the interval ranges corresponds to a preset fault degree;

[0185] The processing unit is further configured to use a preset fault degree corresponding to the interval range to which the status evaluation result belongs as the fault degree of the storage device.

[0186] In one embodiment, the determining unit 303 is further configured to determine whether to perform fault location according to the fault degree of the storage device;

[0187] The determining unit 303 is further configured to determine the electronic device to be maintained according to the weighted result corresponding to the maximum data change rate of each electronic device if the fault degree exceeds a preset fault degree threshold;

[0188] The processing unit is further configured to execute a maintenance strategy corresponding to the electronic component to be maintained according to the electronic component to be maintained.

[0189] In actual application, the processing unit can be implemented by a processor in the storage device in combination with a communication interface, and the acquisition unit 301, the calculation unit 302, and the determination unit 303 can be implemented by a communication interface in a fault prediction device of the storage device.

[0190] It should be noted that: the above embodiment provides a storage device fault prediction device, when performing fault prediction, only uses the division of the above program modules as an example. In actual applications, the above processing can be assigned to different program modules as needed, that is, the internal structure of the device is divided into different program modules to complete all or part of the above-described processing. In addition, the storage device fault prediction device provided in the above embodiment and the storage device fault prediction method embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0191] Based on the hardware implementation of the above program module, and in order to implement a fault prediction method for a storage device provided in an embodiment of the present application, an embodiment of the present application further provides a storage device, such as Figure 4 As shown, the storage device 400 includes:

[0192] CPU 401, memory 402 and input / output interface 403;

[0193] The memory 402 is a temporary storage memory or a permanent storage memory;

[0194] The central processor 401 is configured to communicate with the memory 402 and execute instructions in the memory 402 to perform any one of the above-mentioned methods for predicting failures of storage devices.

[0195] Of course, in actual application, the various components in the storage device 400 are coupled together through the bus system 404. It can be understood that the bus system 404 is used to realize the connection and communication between these components. In addition to the data bus, the bus system 404 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, Figure 4 Various buses are labeled as bus system 404 .

[0196] The memory 402 in the embodiment of the present application is used to store various types of data to support the operation of the storage device 400. Examples of such data include: any computer program used to operate on the storage device 400.

[0197] It is understandable that when the central processing unit 401 in the storage device 400 described above executes the computer program, it can also implement the functions of the various units in the above-mentioned corresponding device embodiments, which will not be repeated here. Exemplarily, the computer program can be divided into one or more modules / units, one or more modules / units are stored in the memory 402, and executed by the central processing unit 401 to complete the various embodiments of the present application. One or more modules / units can be a series of computer program instruction segments that can perform specific functions, and the instruction segments are used to describe the execution process of the computer program in the storage device 400. For example, the computer program can be divided into the various units in the above-mentioned storage device 400, and each unit can implement the specific functions described in the above-mentioned corresponding storage device 400.

[0198] The fault prediction method of the storage device provided in the embodiment of the present application is applied to the storage device 400. Those skilled in the art can understand that the central processor 401 and the memory 402 are merely examples of the storage device 400, and do not constitute a limitation on the storage device 400, which may include more or fewer components, or a combination of certain components, or different components. For example, the storage device 400 may also include an error correction module (ECC), an interface controller (such as SATA, NVMe, PCIe interface), etc.

[0199] The central processor 401 may be a central processing unit (CPU), or other general-purpose processors, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the storage device 400, and uses various interfaces and lines to connect various parts of the entire storage device 400.

[0200] The memory 402 can be used to store computer programs and / or modules. The central processor 401 implements various functions of the storage device 400 by running or executing the computer programs and / or modules stored in the memory 402 and calling the data stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 402 may include a high-speed random access memory, and may also include a non-volatile memory, such as a small-capacity non-volatile storage (such as NOR Flash).

[0201] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, any one of the above-mentioned methods for predicting a fault of a storage device is executed.

[0202] An embodiment of the present application also provides a computer program product having a computer program / instruction stored thereon. When the computer program / instruction is executed by a processor, it is used to implement the fault prediction method for the storage device described in the first aspect of the embodiment of the present application or any specific implementation method of the first aspect.

[0203] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0204] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0205] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0206] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0207] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a storage device to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

Claims

1. A method for predicting a failure of a flash memory device, characterized in that: include: Acquire the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data; Calculating a status evaluation result of the storage device according to the working status data of each electronic device at each moment; The failure degree of the storage device is determined according to the status evaluation result of the storage device.

2. The method for predicting a storage device failure according to claim 1, wherein: The electronic device includes at least two of a memory storage, a serial interface, a flash memory, a bus interface, a system on chip, and a capacitor; the data sampling type of the memory storage includes at least one of an error count and a continuous working time; the data sampling type of the serial interface includes at least one of the number of erase and write times, temperature, and continuous working time; the data sampling type of the flash memory includes at least one of the number of erase and write times, temperature, and continuous working time; the data sampling type of the bus interface includes the number of signal transmission errors; The data sampling type of the system on chip includes at least one of the number of restarts and the temperature; the data sampling type of the capacitor includes the capacitance and the instantaneous state value of the capacitor.

3. The method for predicting a storage device failure according to claim 1, wherein: The calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes: Based on each of the data sampling types of each of the electronic devices, determining the target working state data with the largest value among the working state data at each moment; Based on the target working state data of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a state evaluation sample of each electronic device; The status evaluation result of the storage device is obtained by performing calculation according to the status evaluation samples of each of the electronic components and the weight of each of the electronic components.

4. The method for predicting a failure of a storage device according to claim 1, wherein: The calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes: Based on each of the data sampling types of each of the electronic devices, calculating the data change rate of the working status data at every two adjacent moments to obtain a data change rate set; Determining a target data change rate with the largest value in the data change rate set; Based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a state evaluation sample of each electronic device; The status evaluation result of the storage device is obtained by performing calculation according to the status evaluation samples of each of the electronic components and the weight of each of the electronic components.

5. The method for predicting a failure of a storage device according to claim 1, wherein: The calculating of the state evaluation result of the storage device according to the working state data of each electronic device at each moment includes: For each data sampling type of each electronic device, determining the target working state data with the largest value among the working state data at each moment; Based on the target working state data of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a first state evaluation sample of each electronic device; Calculating according to the first state evaluation samples of each of the electronic devices and the weight of each of the electronic devices to obtain a first evaluation result of the storage device; Based on each of the data sampling types of each of the electronic devices, calculating the data change rate of the working status data at every two adjacent moments to obtain a data change rate set; Determining a target data change rate with the largest value in the data change rate set; Based on the target data change rate of each data sampling type in each electronic device and the weight of each data sampling type, obtaining a second state evaluation sample of each electronic device; Calculating according to the second state evaluation samples of each of the electronic devices and the weight of each of the electronic devices to obtain a second evaluation result of the storage device; A status evaluation result of the storage device is determined according to the first evaluation result and the second evaluation result.

6. The method for predicting a failure of a storage device according to claim 1, wherein: Determining the failure degree of the storage device according to the status evaluation result of the storage device includes: According to a plurality of interval ranges divided by a plurality of preset state evaluation result thresholds, selecting the interval range to which the state evaluation result belongs; wherein each of the interval ranges corresponds to a preset fault degree; The preset fault degree corresponding to the interval range to which the status evaluation result belongs is used as the fault degree of the storage device.

7. The method for predicting a failure of a storage device according to claim 1, wherein: The method further comprises: Determining whether to perform fault location according to the fault degree of the storage device; If the fault degree exceeds a preset fault degree threshold, determining the electronic device to be maintained according to a weighted result corresponding to the maximum data change rate of each electronic device; According to the electronic component to be maintained, a maintenance strategy corresponding to the electronic component to be maintained is executed.

8. A storage device failure prediction device, characterized in that: include: An acquisition unit, used to acquire the working status data of each electronic device in the storage device at each moment; wherein different data sampling types under the same electronic device correspond to different working status data; A calculation unit, used for calculating the status evaluation result of the storage device according to the working status data of each electronic device at each moment; A determination unit is used to determine the failure degree of the storage device according to the status evaluation result of the storage device.

9. A storage device, characterized in that: include: CPU, memory and input / output interface; The memory is a short-term storage memory or a persistent storage memory; The central processing unit is configured to communicate with the memory and execute instruction operations in the memory to perform the fault prediction method for a storage device according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the failure prediction method for a storage device according to any one of claims 1 to 7 is performed.

Images