Abnormality monitoring method and device of storage device, storage medium and electronic device

Abstract

The application discloses an abnormality monitoring method and device of a storage device, a storage medium and an electronic device, relates to the field of storage devices, and comprises the following steps: collecting a first power consumption parameter, a first temperature parameter and a first correlation parameter of the storage device, and detecting a second power consumption parameter and a second temperature parameter in a current running scene; determining a power consumption deviation parameter according to the first power consumption parameter and the second power consumption parameter, determining a temperature deviation parameter according to the first temperature parameter and the second temperature parameter, and converting the first correlation parameter into a correlation deviation parameter; detecting an equipment abnormality state of the storage device according to the power consumption deviation parameter, the temperature deviation parameter and the correlation deviation parameter, and detecting an abnormality cause of the storage device in the case that the storage device is found to be abnormal, so that the technical problems of low fault root cause identification efficiency of the abnormality monitoring method of the storage device are solved, and the technical effects of improving the fault root cause identification efficiency of the abnormality monitoring of the storage device are achieved.

Description

Technical Field

[0001] This application relates to the field of storage devices, and more specifically, to a method and apparatus for abnormal monitoring of storage devices, a storage medium, and an electronic device. Background Technology

[0002] In related technologies, anomaly monitoring of storage devices typically relies on temperature detection. For example, the temperature of the storage device is collected and compared to a preset fixed temperature threshold. Once the temperature exceeds the threshold, the storage device's performance is reduced to attempt cooling, thereby preventing damage from high temperatures. However, this method has significant limitations: Firstly, effective cooling by reducing performance requires that the temperature anomaly is caused by normal high load. If the anomaly stems from a hardware failure, this operation cannot solve the root cause and may even exacerbate the damage due to delayed fault handling, resulting in permanent device damage or data loss. Secondly, the above method can only identify the anomaly of temperature exceeding the threshold; it cannot distinguish between benign temperature increases caused by normal high load and malignant temperature increases caused by hardware failure. The root cause of the fault still requires manual investigation, resulting in low fault handling efficiency and difficulty in ensuring the reliable operation of the storage device.

[0003] In response to the technical problems of low efficiency in identifying the root cause of failures in the anomaly monitoring methods of storage devices in related technologies, no effective solution has yet been proposed. Summary of the Invention

[0004] This application provides a method and apparatus for monitoring the anomalies of storage devices, a storage medium, and an electronic device, in order to at least solve the technical problems of low efficiency in identifying the root cause of failures in the related art.

[0005] According to one embodiment of this application, a method for monitoring anomalies in a storage device is provided, including:

[0006] The system collects a first power consumption parameter of the storage device, collects a first temperature parameter of the storage device, and determines a first correlation parameter of the storage device. It also detects a second power consumption parameter and a second temperature parameter of the storage device under the current operating scenario. The first power consumption parameter is used to indicate the first device power consumption of the storage device, the first temperature parameter is used to indicate the first device temperature of the storage device, the first correlation parameter is used to indicate the first correlation between the first power consumption parameter and the first temperature parameter, the second power consumption parameter is used to indicate the second device power consumption that the storage device is allowed to generate under the current operating scenario, and the second temperature parameter is used to indicate the second device temperature that the storage device is allowed to generate under the current operating scenario.

[0007] A power consumption deviation parameter is determined based on a first power consumption parameter and a second power consumption parameter. A temperature deviation parameter is determined based on a first temperature parameter and a second temperature parameter. A first correlation parameter is converted into a correlation deviation parameter. The power consumption deviation parameter indicates the degree of deviation of the power consumption of the first device relative to the power consumption of the second device. The temperature deviation parameter indicates the degree of deviation of the temperature of the first device relative to the temperature of the second device. The correlation deviation parameter indicates the degree of deviation of the first correlation relative to the second correlation. The second correlation indicates the degree of correlation between the power consumption and temperature of the storage device.

[0008] The abnormal status of the storage device is detected based on power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters. When an abnormal status is detected to indicate that there is an abnormality in the storage device, the cause of the abnormality is determined based on the power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters.

[0009] According to another embodiment of the present application, an anomaly monitoring device for a storage device is also provided, comprising:

[0010] The acquisition module is used to acquire a first power consumption parameter of the storage device, acquire a first temperature parameter of the storage device, and determine a first correlation parameter of the storage device, and detect a second power consumption parameter and a second temperature parameter of the storage device in the current operating scenario. The first power consumption parameter is used to indicate the first device power consumption of the storage device, the first temperature parameter is used to indicate the first device temperature of the storage device, the first correlation parameter is used to indicate the first correlation between the first power consumption parameter and the first temperature parameter, the second power consumption parameter is used to indicate the second device power consumption allowed to be generated by the storage device in the current operating scenario, and the second temperature parameter is used to indicate the second device temperature allowed to be generated by the storage device in the current operating scenario.

[0011] The determination module is used to determine a power consumption deviation parameter based on a first power consumption parameter and a second power consumption parameter, determine a temperature deviation parameter based on a first temperature parameter and a second temperature parameter, and convert a first correlation parameter into a correlation deviation parameter. The power consumption deviation parameter is used to indicate the degree of deviation of the power consumption of the first device relative to the power consumption of the second device, the temperature deviation parameter is used to indicate the degree of deviation of the temperature of the first device relative to the temperature of the second device, the correlation deviation parameter is used to indicate the degree of deviation of the first correlation relative to the second correlation, and the second correlation is used to indicate the degree of correlation between the power consumption and temperature of the storage device.

[0012] The first detection module is used to detect the abnormal state of the storage device based on power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters. When the abnormal state of the device is detected to indicate that there is an abnormality in the storage device, the module is used to detect the cause of the abnormality of the storage device based on the power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters.

[0013] This application also provides an electronic device, including: a memory for storing a computer program; and a processor for implementing the abnormal monitoring method of any of the above-described storage devices when executing the computer program.

[0014] This application also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the steps of the abnormal monitoring method of any of the above-described storage devices.

[0015] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the above-described methods for monitoring the anomalies of a storage device.

[0016] This application first collects the first power consumption parameter and the first temperature parameter of the storage device, determines the first correlation parameter of the storage device, and detects the second power consumption parameter and the second temperature parameter that the storage device is allowed to generate under the current operating scenario. Then, a power consumption deviation parameter is determined based on the first power consumption parameter and the second power consumption parameter, and a temperature deviation parameter is determined based on the first temperature parameter and the second temperature parameter. The first correlation parameter is then converted into a correlation deviation parameter. Finally, the abnormal state of the storage device is detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter. If an abnormality is detected, the cause of the abnormality is further detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter. By monitoring and locating the cause of abnormality from the deviations of power consumption, temperature, and their correlation, the shortcomings of related technologies that rely on fixed thresholds and cannot distinguish between benign temperature rises and malignant failures are avoided. Therefore, it can solve the technical problems of low efficiency in fault root cause identification in the abnormal monitoring methods of storage devices in related technologies, and achieve the technical effect of improving the efficiency of fault root cause identification in abnormal monitoring of storage devices. Attached Figure Description

[0017] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a hardware structure block diagram of a computer device for an abnormal monitoring method of a storage device according to an embodiment of this application;

[0019] Figure 2 This is a schematic diagram of an anomaly monitoring method for storage devices based on related technologies;

[0020] Figure 3 This is a flowchart of an anomaly monitoring method for a storage device according to an embodiment of this application;

[0021] Figure 4 This is an example diagram of an anomaly monitoring system for a storage device according to an embodiment of this application;

[0022] Figure 5 This is an example diagram of a power consumption detection module circuit according to an embodiment of this application;

[0023] Figure 6 This is an example diagram of a temperature monitoring module circuit according to an embodiment of this application;

[0024] Figure 7 This is an example diagram of the backup power management module circuit according to an embodiment of this application;

[0025] Figure 8 This is a flowchart of SSD anomaly monitoring according to an embodiment of this application;

[0026] Figure 9 This is a flowchart of the fault location decision-making process according to an embodiment of this application;

[0027] Figure 10 This is a flowchart of an anomaly monitoring method based on multi-node temperature according to an embodiment of this application;

[0028] Figure 11 This is a flowchart of an anomaly monitoring method based on multi-node power consumption according to an embodiment of this application;

[0029] Figure 12 This is a flowchart of a multi-level power consumption level control method according to an embodiment of this application;

[0030] Figure 13 This is a structural block diagram of an anomaly monitoring device for a storage device according to an embodiment of this application;

[0031] Figure 14 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0032] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0033] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0034] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] The methods and embodiments provided in this application can be executed on a server device or a similar computing device. Taking running on a server device as an example, Figure 1 This is a hardware structure block diagram of a computer device for an abnormal monitoring method of a storage device according to an embodiment of this application. Figure 1 As shown, the server device may include one or more ( Figure 1 Only one is shown in the image. A processor 102 (which may include, but is not limited to, a central processing unit (CPU), microprocessor (MCU), or programmable logic device (FPGA), etc.) and a memory 104 for storing data are also shown. The server device may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the server equipment described above. For example, the server equipment may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0036] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the abnormal monitoring method of the storage device in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thus implementing the above-described method. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to server devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0037] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider for the server device. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0038] This application proposes a method for monitoring anomalies in storage devices. Before describing the optional embodiments of this application, in order to better understand the inventive concept and the inventiveness of this solution, the relevant technologies are first explained: Figure 2 This is a schematic diagram of a method for monitoring the anomalies of a storage device based on related technologies, such as... Figure 2 As shown, related technologies primarily utilize a Baseboard Management Controller (BMC) to monitor the temperature status of storage devices. For example, temperature information of the storage device is obtained through the System Management Bus (SMBUS), and temperature control strategies (such as controlling the workload or cooling system of the storage device) are executed based on this information to maintain the temperature of the storage device within a certain range. However, the above method has significant limitations: judging storage device anomalies solely based on single-point temperature data cannot distinguish whether the anomaly is a benign temperature rise caused by normal high load or a malignant temperature rise caused by hardware failure. The root cause of the fault must be manually investigated one by one, resulting in low fault handling efficiency and difficulty in ensuring the reliable operation of the storage device.

[0039] To address the above issues, this embodiment provides a method for monitoring anomalies in storage devices. Figure 3 This is a flowchart of an anomaly monitoring method for a storage device according to an embodiment of this application, such as... Figure 3 As shown, the process includes the following steps:

[0040] Step S12: Collect the first power consumption parameter of the storage device, collect the first temperature parameter of the storage device, and determine the first correlation parameter of the storage device. Detect the second power consumption parameter and the second temperature parameter of the storage device in the current operating scenario. The first power consumption parameter is used to indicate the first device power consumption of the storage device, the first temperature parameter is used to indicate the first device temperature of the storage device, the first correlation parameter is used to indicate the first correlation between the first power consumption parameter and the first temperature parameter, the second power consumption parameter is used to indicate the second device power consumption allowed to be generated by the storage device in the current operating scenario, and the second temperature parameter is used to indicate the second device temperature allowed to be generated by the storage device in the current operating scenario.

[0041] Step S14: Determine a power consumption deviation parameter based on the first power consumption parameter and the second power consumption parameter; determine a temperature deviation parameter based on the first temperature parameter and the second temperature parameter; and convert the first correlation parameter into a correlation deviation parameter. The power consumption deviation parameter indicates the degree of deviation between the power consumption of the first device and the power consumption of the second device; the temperature deviation parameter indicates the degree of deviation between the temperature of the first device and the temperature of the second device; the correlation deviation parameter indicates the degree of deviation between the first correlation and the second correlation; and the second correlation indicates the degree of correlation between the power consumption and temperature of the storage device.

[0042] Step S16: Detect the abnormal state of the storage device based on the power consumption deviation parameter, temperature deviation parameter, and correlation deviation parameter. If the abnormal state of the storage device is detected to indicate that there is an abnormality in the storage device, detect the cause of the abnormality of the storage device based on the power consumption deviation parameter, temperature deviation parameter, and correlation deviation parameter.

[0043] This application first collects the first power consumption parameter and the first temperature parameter of the storage device, determines the first correlation parameter of the storage device, and detects the second power consumption parameter and the second temperature parameter that the storage device is allowed to generate under the current operating scenario. Then, a power consumption deviation parameter is determined based on the first power consumption parameter and the second power consumption parameter, and a temperature deviation parameter is determined based on the first temperature parameter and the second temperature parameter. The first correlation parameter is then converted into a correlation deviation parameter. Finally, the abnormal state of the storage device is detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter. If an abnormality is detected, the cause of the abnormality is further detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter. By monitoring and locating the cause of abnormality from the deviations of power consumption, temperature, and their correlation, the shortcomings of related technologies that rely on fixed thresholds and cannot distinguish between benign temperature rises and malignant failures are avoided. Therefore, it can solve the technical problems of low efficiency in fault root cause identification in the abnormal monitoring methods of storage devices in related technologies, and achieve the technical effect of improving the efficiency of fault root cause identification in abnormal monitoring of storage devices.

[0044] Optionally, in this embodiment, the storage device may be, but is not limited to, physical hardware used for persistent or temporary storage of digital information. Examples include solid-state drives (SSDs) and hybrid hard drives (HHDs). This embodiment uses an SSD as an example to illustrate this anomaly monitoring method, but it does not limit the specific type of storage device.

[0045] Optionally, in this embodiment, the anomaly monitoring method for the storage device can be applied to, but is not limited to, the anomaly monitoring system for the storage device. Figure 4 This is an example diagram of an anomaly monitoring system for a storage device according to an embodiment of this application, such as... Figure 4 As shown, the anomaly monitoring system may include, but is not limited to, a multi-channel power consumption detection module circuit, a multi-channel temperature monitoring module circuit, a backup power management module circuit, an out-of-band interface, an external controller, and an internal controller. The multi-channel power consumption detection module is connected to multiple core detection nodes one-to-one, the multi-channel temperature monitoring module circuit is connected to multiple core detection nodes one-to-one, the multi-channel power consumption detection module, the multi-channel temperature detection module circuit, and the internal controller are all connected to the backup power management module circuit, the internal controller is connected to the multi-channel power consumption detection module and the multi-channel temperature detection module circuit respectively, and the internal controller is also connected to the external controller through an out-of-band interface. The system includes a multi-channel power consumption detection module circuit for real-time monitoring of the power consumption of multiple nodes in the storage device; a multi-channel temperature monitoring module circuit for real-time monitoring of the temperature of multiple nodes in the storage device; a backup power management module circuit for monitoring and managing the backup power status and controlling the enabling or disabling of the disk-level power supply when an abnormal signal is received; an in-disk controller for acquiring and processing health information (including power consumption monitored by the power consumption detection module circuit and temperature monitored by the temperature monitoring module circuit), the operating status of the storage device, executing internal decisions and controlling the storage device, and reporting health information to an external controller via an out-of-band interface (such as SMBUS), which could be the SSD's main controller; and an external controller for receiving health information reported by the SSD and executing decisions, which can be, but is not limited to, a BMC with server hardware monitoring and control capabilities.

[0046] It should be noted that the execution entity of the method described in this embodiment can be flexibly configured: it can be executed independently by the in-panel controller, for example, the in-panel controller directly completes anomaly judgment and root cause location based on locally acquired power consumption and temperature data; it can also be executed independently by the external controller, for example, the external controller receives health information uploaded by the in-panel controller through an out-of-band interface and independently performs anomaly analysis and decision-making; or it can be executed collaboratively by both, for example, the in-panel controller is responsible for preliminary anomaly screening and basic data preprocessing, and then sends key information to the external controller, which combines the global system status to complete in-depth root cause analysis and final decision-making, in order to adapt to the monitoring accuracy and efficiency requirements of different scenarios.

[0047] Optionally, in this embodiment, such as Figure 4 As shown, multiple nodes in a storage device can indicate, but are not limited to, the heat-generating and power-consuming units within the SSD, such as the SSD main controller, NAND (Non-Volatile Memory) flash memory chips, DRAM (Dynamic Random Access Memory) cache chips, and the PCB (Printed Circuit Board) environment.

[0048] Optionally, in this embodiment, Figure 5 This is an example diagram of the power consumption detection module circuit according to an embodiment of this application, such as... Figure 5 As shown, the power consumption detection module circuit measures the current flowing through the sampling resistor and the voltage across it, thereby calculating the actual power consumption of the storage device and using this as the input source for the first power consumption parameter. The power consumption detection module circuit mainly includes: a sampling resistor R1, used to sample the current of the input power supply and generate a sampling voltage across it; an analog-to-digital converter unit, used to convert the sampling voltage into a digital quantity; and power consumption, current, and voltage registers, used to store the measurement results after analog-to-digital conversion. The on-disk controller (e.g., the SSD main controller) reads the digital quantities in these registers through the IIC (Inter-Integrated Circuit) interface, and after calculation or processing, obtains the first device power consumption of the storage device. The first power consumption parameter is a digital and parameterized representation of the first device power consumption.

[0049] Optionally, in this embodiment, since the anomaly monitoring system deploys multiple power consumption detection module circuits, it can simultaneously monitor the total power consumption of the entire panel and the power consumption of the core device's power supply branch. Therefore, the first power consumption parameter can be a single total power consumption value or a set of real-time power consumption values ​​for multiple nodes. Simultaneously, the power consumption detection module circuit also includes an Alert register and an ALERT (alarm) output pin, which can be configured with power consumption or current thresholds. When these thresholds are exceeded, a pre-alarm signal is output through the ALERT pin to the panel controller or directly triggers the backup power management module circuit, enabling rapid response and protection.

[0050] Optionally, in this embodiment, Figure 6 This is an example diagram of a temperature monitoring module circuit according to an embodiment of this application, such as... Figure 6 As shown, the temperature monitoring module circuit is used to monitor the temperature of multiple nodes in the storage device in real time. The temperature monitoring module circuit acquires analog signals through local temperature sensors, processes them through an amplifier and analog-to-digital converter, converts them into digital temperature signals, and then verifies them against the calibration of the storage unit to ensure accuracy. The on-disk controller reads these digital signals through the IIC interface to obtain the current temperature of the first device in the storage device. The first temperature parameter is a digital and parameterized representation of the temperature of this first device.

[0051] Optionally, in this embodiment, since the anomaly monitoring system deploys multiple temperature monitoring module circuits, it can monitor the temperature of multiple key nodes such as the SSD main controller, NAND, DRAM, and SSD board. Therefore, the first temperature parameter can be a single temperature value or a set of real-time temperatures from multiple nodes. Simultaneously, the temperature monitoring module circuit uses digital logic to implement temperature warning judgments, and when the monitored temperature is too high or too low, it outputs a pre-alarm signal to the disk controller via the ALERT pin or directly triggers the backup power management module circuit to achieve rapid response and protection.

[0052] Optionally, in this embodiment, Figure 7 This is an example diagram of the backup power management module circuit according to an embodiment of this application, such as... Figure 7As shown, the backup power management module circuit is mainly used to initiate the backup power process when the main power supply is abnormal or the power consumption / temperature exceeds a preset threshold, ensuring data security and providing power enable control and health detection functions. This module includes: an overcurrent protection module to provide protection on the normal power supply path; a boost / buck circuit module that boosts the input power (P12V_SENSE+) to P32V to charge the external energy storage capacitor (C3) during normal power supply; when a drop in the main power supply voltage is detected (e.g., below 10V), this module steps down the P32V on the energy storage capacitor, outputting the voltage to continue powering the subsequent DC (Direct Current) / DC buck module circuit to complete the backup power process. Furthermore, the EN (enable) pin is used to receive control signals from the disk controller (such as GPIO1) to enable timely disk-level power-off processing of the storage device when abnormal power consumption or temperature is detected. For example, when an abnormal device state is detected, the disk controller outputs a low level to the EN pin via GPIO1 to achieve hardware-level power-off protection and prevent further escalation of risks.

[0053] Optionally, in this embodiment, the first correlation parameter may be, but is not limited to, a value used to quantify the degree of statistical linear correlation between the real-time power consumption (first power consumption parameter) sequence and the real-time temperature (first temperature parameter) sequence of the storage device. Specifically, the first correlation parameter is periodically calculated within a continuous sliding time window (e.g., the most recent 60 seconds) to characterize the consistency of the trends in power consumption and temperature data. For example, it may be a value between -1 and +1 obtained using the Pearson correlation coefficient, or the peak value of the cross-correlation function obtained based on time series analysis. For instance, in the SSD "2GB / s stable sequential write" scenario, the first correlation parameter calculated through the sliding time window is 0.9 (strong positive correlation), indicating that the real-time power consumption and temperature trends are highly consistent in this scenario.

[0054] Optionally, in this embodiment, the first correlation can be, but is not limited to, indicating the physical meaning directly represented by the first correlation parameter, namely, the statistical correlation between actual power consumption and actual temperature change trends. For example, whether the temperature also increases when power consumption increases (positive correlation), whether the temperature decreases when power consumption increases (negative correlation), or whether there is no obvious correlation between the two changes (near-zero correlation).

[0055] Optionally, in this embodiment, the second correlation may, but is not limited to, indicating the expected and typical statistical correlation between power consumption changes and temperature changes when the storage device is in a healthy and normal operating state under a specific "current operating scenario". For example, in the "2GB / s stable sequential write" scenario, the average Pearson correlation coefficient value between power consumption and temperature calculated from a large amount of historical normal data, for example, a preferred value is 1.

[0056] Optionally, in this embodiment, the current operating scenario can be, but is not limited to, the specific workload and environmental conditions under which the storage device is monitored. These states directly affect the power consumption and heat dissipation mode of the device. They can be divided according to dimensions such as data read / write type (sequential write / random read), bandwidth requirements (2GB / s, 5GB / s), IOPS (Input / Output Operations Per Second) pressure (low IOPS / high IOPS). For example, they can be, but are not limited to: high-intensity continuous read operation scenario, high-intensity random write operation scenario, mixed read / write scenario, idle standby scenario, or background garbage collection execution scenario. Under different current operating scenarios, the load intensity of the core components of the storage device (main controller, NAND, DRAM) varies, and the corresponding allowable power consumption (second power consumption parameter) also varies. For example, in the "low IOPS random read" scenario, the DRAM load is low, and its second power consumption parameter is 0.5W. In the "high IOPS random write" scenario, the DRAM load is high, and its second power consumption parameter is 4W. Moreover, this scenario information can be obtained or calculated by the on-disk controller to retrieve or dynamically generate the corresponding second power consumption parameter and second temperature parameter.

[0057] In the embodiment provided in step S12 above, the detection of the second power consumption parameter and the second temperature parameter can be achieved by querying a predefined reference mapping table. For example, the in-disk controller can analyze the characteristics of input / output requests in real time, such as request type, data block size, and the randomness or sequence of access, to classify the current complex operating scenario into several predefined workload modes. These workload modes may include, but are not limited to, "high-bandwidth sequential read," "high-IOPS random write," and "idle standby," and based on the identified workload mode, query the corresponding preset power consumption reference value and temperature reference value from the mapping table pre-stored in non-volatile memory as the second power consumption parameter and the second temperature parameter in the current scenario.

[0058] In another implementation, the detection of the second power consumption parameter and the second temperature parameter can also be achieved through dynamic calculation based on a physical model. For example, the in-disk controller acquires accurate workload metrics in real time, including bandwidth, number of read / write operations, and ambient temperature. Subsequently, the controller uses pre-calibrated physical model parameters, such as the idle power consumption of the device, power consumption per unit bandwidth, power consumption per unit of operation, and thermal resistance coefficient, to calculate the second power consumption parameter by summing the idle power consumption of each core device, the power consumption generated by the current bandwidth, and the power consumption generated by the current number of read / write operations. The second temperature parameter is obtained by summing the ambient temperature of each core device and the temperature generated under the expected power consumption.

[0059] Optionally, in this embodiment, it should be further noted that those skilled in the art will understand that the methods for detecting the second power consumption parameter and the second temperature parameter are not limited to the two methods described above. For example, it can also be achieved through prediction based on machine learning models. As long as a reasonable expected power consumption value and temperature value can be determined based on the current operating scenario, they all fall within the protection scope of this application.

[0060] Optionally, in this embodiment, the power consumption deviation parameter may be, but is not limited to, indicating the degree of deviation between the measured power consumption (first power consumption parameter) of the storage device or its internal core components and its expected power consumption (second power consumption parameter) under the current operating scenario. The power consumption deviation parameter may be a dimensionless ratio or percentage used to quantify the difference between actual power consumption and normal power consumption; the larger the value, the higher the probability of abnormal power consumption.

[0061] Optionally, in this embodiment, the temperature deviation parameter may be, but is not limited to, indicating the degree of deviation of the measured temperature (first temperature parameter) of the storage device or its internal core components from its expected temperature (second temperature parameter) under the current operating scenario. The temperature deviation parameter is also presented in the form of a relative value (such as a ratio or percentage), reflecting the difference between the actual temperature and the normal temperature.

[0062] Optionally, in this embodiment, the correlation deviation parameter may be, but is not limited to, indicating the degree of deviation between the measured correlation between power consumption and temperature of the storage device (first correlation parameter) and the expected correlation under healthy conditions (second correlation). The correlation deviation parameter is mainly used to detect whether the normal physical relationship between power consumption and temperature has been disrupted, and to further identify hidden problems such as heat dissipation system failure and sensor coordination failure.

[0063] In the embodiment provided in step S14 above, the power consumption deviation parameter can be determined by calculating the absolute difference between the first power consumption parameter and the second power consumption parameter. For example, but not limited to, a computing unit can be used to calculate the absolute difference between the first power consumption parameter and the second power consumption parameter as the power consumption deviation parameter. If the first power consumption parameter is 15W and the second power consumption parameter is 12W, then the power consumption deviation parameter is positive 3W. Alternatively, the normalized deviation (Z-score) between the first power consumption parameter and the second power consumption parameter can be calculated. For example, but not limited to, a statistical module can be used to calculate the quotient of the absolute difference between the first power consumption parameter and the second power consumption parameter and the historical power consumption standard deviation under the current operating scenario as the power consumption deviation parameter.

[0064] In the embodiment provided in step S14 above, the temperature deviation parameter can be determined by calculating the absolute difference between the first temperature parameter and the second temperature parameter. Alternatively, it can be determined by calculating the normalized deviation between the first and second temperature parameters, similar to the method used to determine the power consumption deviation parameter.

[0065] In the embodiment provided in step S14 above, the first correlation parameter is converted into a correlation deviation parameter, which can be achieved by calculating the absolute difference between the first correlation parameter and the second correlation parameter, wherein the second correlation parameter is used to indicate the second correlation. For example, but not limited to, a calculation unit can be used to calculate the absolute difference between the first correlation parameter and the second correlation parameter as the correlation deviation parameter. Since the power consumption and temperature of a healthy device should be strongly positively correlated in the operating scenario, the second correlation parameter is close to 1, so that the calculated correlation deviation parameter should be close to 0 under normal circumstances. Alternatively, the normalized deviation between the first power consumption parameter and the second power consumption parameter can be calculated. For example, but not limited to, a statistical module can be used to calculate the quotient of the absolute difference between the first power consumption parameter and the second power consumption parameter and the historical power consumption standard deviation in the current operating scenario as the power consumption deviation parameter, which is used to quantify the difference between the actual correlation degree and the expected correlation degree.

[0066] Optionally, in this embodiment, the cause of the anomaly can be, but is not limited to, an indication: the specific physical root cause or systemic problem that leads to the abnormal health status of the storage device, identified by analyzing the comprehensive performance of power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters. These causes are directly related to hardware failure, performance degradation, or system failure, providing clear guidance for accurate maintenance. For example, they can include, but are not limited to: abnormal heat dissipation system (e.g., excessively low cooling fan speed, abnormal application of thermal pads, or failure of heat conduction paths), aging or degradation of core components (e.g., early aging of NAND flash memory chips, performance degradation of the main controller), firmware or logic anomalies (e.g., firmware defects causing abnormally frequent background tasks, abnormal or leaky DRAM memory access), sensor or hardware failures (e.g., sensor (such as sampling resistor) failure in the power consumption detection module circuit, sensor failure in the temperature monitoring module circuit, abnormal power supply (e.g., unstable power supply on the host side or abnormality in the backup power management module circuit), etc.

[0067] In the embodiment provided in step S16 above, when an abnormal device state is detected to indicate an anomaly in the storage device, i.e., a hardware-level fault is determined, the cause of the storage device's anomaly is detected based on power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters. This can be achieved, but is not limited to, by querying a predefined anomaly cause rule base. For example, the anomaly cause rule base stores various deviation parameter combination patterns and their corresponding most likely anomaly causes. For instance, pattern A (high power consumption deviation, high temperature deviation, low correlation deviation) maps to "heat dissipation system failure"; pattern B (high power consumption deviation, low temperature deviation, high correlation deviation) maps to "specific component failure." After detecting an abnormal device state, the on-disk controller queries the corresponding deviation parameter combination pattern from the anomaly cause rule base pre-stored in non-volatile memory to determine the most likely anomaly cause.

[0068] In another implementation, the cause of storage device anomalies can be detected by using the contribution ratio of deviation parameters and physical logic reasoning. This can be, but is not limited to, being executed by the on-board controller according to a preset procedure: First, the on-board controller calls its built-in threshold comparison logic to compare power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters with preset normal thresholds, marking the parameter with the most severe deviation as the core anomaly. Then, deep localization is performed based on the type of the core anomaly: If the core anomaly is a power consumption deviation parameter, the on-board controller further retrieves individual power consumption deviation data for each core component and determines the target component with the largest contribution through numerical comparison; if the core anomaly is a temperature deviation parameter, the corresponding abnormal temperature source is located similarly. If the core anomaly is a correlation deviation parameter, and its value indicates that the correlation coefficient between power consumption and temperature has seriously deviated from the healthy baseline, the on-board controller can, according to preset rules, locate the cause of the anomaly as a systemic anomaly such as a power sensor failure, a temperature sensor failure, or a cooling system failure.

[0069] As an optional approach, the causes of storage device anomalies can be detected based on power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters, including:

[0070] S21, Detect abnormal correlation status of storage devices based on correlation deviation parameters;

[0071] S22, when an abnormal correlation state is detected to indicate that the storage device is abnormal in the correlation between power consumption and temperature, the cause of the abnormality is determined to be an abnormal temperature sensor, an abnormal power consumption sensor, or an abnormal heat dissipation system.

[0072] S23, when a correlation abnormality is detected to indicate that the storage device is not abnormal in the correlation between power consumption and temperature, the power consumption abnormality of the storage device is detected based on the power consumption deviation parameter, and the temperature abnormality of the storage device is detected based on the temperature deviation parameter.

[0073] S24, when an abnormal power consumption state is detected to indicate that the storage device has an abnormal power consumption, a first target device causing the abnormal power consumption state is located from the device components based on the device power consumption of the device components included in the storage device as the cause of the abnormality; when an abnormal temperature state is detected to indicate that the storage device has an abnormal temperature, a second target device causing the abnormal temperature state is located from the device components based on the device temperature and / or device temperature relationship of the device components included in the storage device as the cause of the abnormality.

[0074] Optionally, in this embodiment, the correlation anomaly state may be, but is not limited to, a logical judgment signal or level indicator used to indicate that the actual correlation between the power consumption and temperature of the storage device (characterized by a first correlation parameter) deviates from the expected correlation under healthy conditions (characterized by a second correlation). The aforementioned correlation anomaly state is a direct output of the on-disk controller after evaluating the correlation deviation parameter, used to characterize the degree of deviation of the real-time statistical correlation between power consumption and temperature from the healthy baseline (the correlation that should exist under healthy conditions).

[0075] Optionally, in this embodiment, detecting the correlation anomaly state of the storage device can be achieved, but is not limited to, through the following methods: A comparison logic based on a single threshold. For example, the in-disk controller calls an internally stored fixed threshold to compare the calculated correlation deviation parameter with the fixed threshold. If the correlation deviation parameter is greater than or equal to the threshold, the correlation anomaly state is determined to be "yes" or "high-level anomaly"; if it is less than the threshold, it is determined to be "no" or "normal". Alternatively, a classification logic based on threshold ranges can be used. For example, the in-disk controller presets multiple threshold ranges, mapping the correlation deviation parameter to different anomaly levels, such as "normal", "minor anomaly", and "serious anomaly", as correlation anomaly states.

[0076] Optionally, in this embodiment, if the correlation is abnormal, it indicates that the inherent physical relationship between power consumption and temperature has been disrupted. The cause of the fault is highly likely to be a system-level fault affecting the global data correlation or thermal balance. For example, a temperature sensor malfunction means that the sensor in the temperature monitoring module circuit is inaccurate, the signal transmission is faulty, or the calibration fails, resulting in distorted temperature values. A power consumption sensor malfunction means that the sampling resistor or analog-to-digital conversion unit in the power consumption detection module circuit malfunctions, resulting in distorted power consumption values. A heat dissipation system malfunction means that the heat dissipation conditions inside the SSD (such as the effectiveness of the heat sink and the airflow of the fan) have deteriorated significantly, leading to a sharp increase in thermal resistance. If the correlation is determined to be normal, it indicates that the inherent physical relationship between power consumption and temperature is normal. The problem may originate from a local fault in one or more components, thus requiring in-depth analysis of the power consumption deviation parameters and temperature deviation parameters.

[0077] Optionally, in this embodiment, determining the cause of the anomaly includes temperature sensor anomaly, power consumption sensor anomaly, and heat dissipation system anomaly, which can be achieved, but is not limited to, through the following methods: The in-panel controller executes a predefined rule mapping. For example, the in-panel controller has the following built-in decision logic: once the relevant anomaly state is "abnormal," the set of anomaly causes is immediately set to {temperature sensor anomaly, power consumption sensor anomaly, heat dissipation system anomaly}. Alternatively, the in-panel controller executes auxiliary reasoning based on data trends. For example, after determining a systemic anomaly, the in-panel controller further analyzes the original data sequences (i.e., the first power consumption parameter and the first temperature parameter) from the multi-channel power consumption detection module circuit and the multi-channel temperature monitoring module circuit. If the power consumption data sequence is found to be stable while the temperature data sequence jumps sharply, the in-panel controller increases the confidence level of "temperature sensor anomaly" in the reported cause; if the opposite is true, it increases the confidence level of "power consumption sensor anomaly"; if both data are abnormal but the trends are decoupled, it increases the confidence level of "heat dissipation system anomaly."

[0078] Optionally, in this embodiment, the in-panel controller can report relevant abnormal states and auxiliary analysis data to the out-of-band controller via an out-of-band interface.

[0079] Optionally, in this embodiment, the abnormal power consumption state and the abnormal temperature state may be, but are not limited to, independent judgment results generated by the in-panel controller for the overall power consumption and the overall temperature, respectively, based on whether the power consumption deviation parameter or the temperature deviation parameter exceeds their respective preset thresholds.

[0080] Optionally, in this embodiment, detecting abnormal power consumption states of the storage device based on power consumption deviation parameters is similar to detecting abnormal temperature states of the storage device based on temperature deviation parameters, and can be based on parallel independent threshold comparisons. For example, the in-disk controller has built-in power consumption abnormality thresholds and temperature abnormality thresholds, and compares the power consumption deviation parameters and temperature deviation parameters with their corresponding thresholds. If any parameter exceeds the threshold, its corresponding abnormal state (power consumption abnormality state or temperature abnormality state) is marked as "abnormal".

[0081] Optionally, in this embodiment, the device may refer to, but is not limited to, the key components inside the storage device that perform core functions and directly affect power consumption and temperature, and are the main source of power consumption and temperature data, such as the main controller in a solid-state drive, DRAM cache chips, NAND flash memory chips, and the PCB board environment.

[0082] Optionally, in this embodiment, the device power consumption and device temperature can be, but are not limited to, indicators: power consumption and temperature values ​​obtained independently for each device through a multi-channel power consumption detection module circuit and a multi-channel temperature monitoring module circuit.

[0083] Optionally, in this embodiment, the device temperature relationship may refer to, but is not limited to, the temperature difference or relative trend between different devices in the storage device. For example, the temperature difference between the main controller and the NAND flash memory chips. Device temperature relationships can be used to determine localized heat dissipation problems.

[0084] Optionally, in this embodiment, the first target device and the second target device can be, but are not limited to, the single or multiple core device devices that contribute the most to the current power consumption anomaly or temperature anomaly. For example, the device identified as the first target device may be a NAND flash memory block with abnormally high power consumption.

[0085] Optionally, in this embodiment, locating the first target device based on the device power consumption can be achieved, but is not limited to, through the following methods: Based on maximum value comparison logic. For example, the in-panel controller reads the real-time power consumption and expected power consumption of all core devices (main controller, DRAM, NAND), and calculates the individual power consumption deviation for each device. Subsequently, by sorting or comparing, the device with the largest individual power consumption deviation value is directly determined as the first target device. Alternatively, it can be based on threshold filtering logic. For example, the in-panel controller sets a threshold for individual power consumption deviation. All devices with individual power consumption deviations exceeding this threshold are filtered out to form a "suspected device set," and this set is used as the location result.

[0086] Optionally, in this embodiment, locating the second target device based on the device temperature and / or device temperature relationship of the device components can be achieved, but is not limited to, through the following methods: Location based on individual temperature. For example, similar to locating abnormal power consumption, the in-panel controller compares the individual temperature deviations of each component and identifies the component with the highest temperature or the most abnormal temperature rise as the second target device. Location can also be based on temperature relationships. For example, the in-panel controller calculates the measured temperature difference between key component pairs (such as the main controller and NAND) and compares it with the expected temperature difference. If the temperature difference is abnormal (e.g., much smaller than expected), it indicates a failure in the heat dissipation system leading to global heat accumulation; if the temperature of a component is significantly mismatched with its power consumption (e.g., low power consumption but extremely high temperature), it indicates a faulty temperature sensor or severe blockage of a local heat dissipation path for that component.

[0087] Based on the above, when the correlation between power consumption and temperature is disrupted, the root cause of the fault can be quickly pinpointed to a limited system-level component (sensor or cooling system), and reported preferentially via out-of-band interface, avoiding unnecessary complex calculations at the component level and significantly improving the early warning and response speed for major risks. When the correlation is normal, the in-panel controller can quickly identify abnormal power consumption of specific devices by analyzing the real-time power consumption and dynamic expected values ​​of the devices, refining the cause of the fault from "abnormal power consumption of the entire panel" to "abnormal power consumption of a specific device." Similarly, when an abnormal temperature state is triggered, by analyzing the temperature data and temperature difference relationship of each sensor node, it is possible to accurately distinguish whether it is local overheating of the main controller, obstruction of the cooling airflow, or heat accumulation in a specific device. The above-mentioned diagnostic capabilities at the device level change the maintenance action from "replacing the entire panel" to "replacing a component" or "targeted cooling optimization," greatly reducing maintenance costs and business downtime.

[0088] As an optional approach, based on the power consumption of the devices included in the storage device, the primary target device causing the abnormal power consumption state is identified as the cause of the anomaly, including:

[0089] S31, obtain the device power consumption deviation of each device, wherein the device power consumption deviation is used to indicate the degree of deviation of the first device power consumption of the corresponding device relative to the second device power consumption in the current operating scenario, the device power consumption deviation is the quotient between the device power consumption difference and the second device power consumption, and the device power consumption difference is the absolute value of the difference between the first device power consumption and the second device power consumption.

[0090] S32 compares the power consumption deviation of each device with the deviation parameter threshold;

[0091] S33, the device whose power consumption deviation is greater than or equal to the deviation parameter threshold is identified as the first target device, and the cause of the abnormality is obtained.

[0092] Optionally, in this embodiment, the power consumption of the first device may refer to, but is not limited to, the actual power consumption value of a single device in the storage device under the current operating state. It needs to be acquired by the power consumption detection module circuit of the corresponding branch, such as: the real-time power consumption of the NAND Flash chip, the real-time power consumption of the DRAM chip, the real-time power consumption of the SSD main controller, etc.

[0093] Optionally, in this embodiment, the power consumption of the second device may refer to, but is not limited to, the expected power consumption value of a single device within the storage device when it is in a healthy and normal operating state under the current operating scenario. This can be achieved by querying a predefined benchmark mapping table, or by dynamic calculation based on a physical model.

[0094] Optionally, in this embodiment, the deviation parameter threshold may be, but is not limited to, a critical value used to determine whether the device power consumption deviation exceeds the normal range. This threshold is calibrated based on a large amount of device health operation data. For example, if the deviation parameter threshold is set to 0.3, when the device power consumption deviation is greater than or equal to 0.3, it is determined that the device power consumption is abnormal. This solution does not limit this.

[0095] Optionally, in this embodiment, devices with power consumption deviations greater than or equal to a deviation parameter threshold are identified as the first target devices to determine the cause of the anomaly. This can be achieved, but is not limited to, by directly filtering based on comparison results. For example, the in-panel controller uses a built-in data filtering module to traverse all calculated device power consumption deviations and compares the power consumption deviation of each device with a preset deviation parameter threshold. Once a device's deviation value is found to be greater than or equal to the threshold, the device is immediately identified as a suspicious object. After traversing all devices, all identified devices are identified as the first target device set causing this power consumption anomaly. The in-panel controller can further generate an anomaly report from this result, for example: "Anomaly cause: DRAM power consumption deviation exceeds the standard."

[0096] Based on the above, by calculating and comparing the power consumption deviation of each device with the threshold, and using the quantified device power consumption deviation as the core basis, the first target device causing the power consumption anomaly was accurately located. This solved the problem that traditional monitoring could not locate the specific device from the overall power consumption anomaly, providing a clear direction for subsequent targeted maintenance and further improving the accuracy and efficiency of power consumption anomaly diagnosis of storage devices.

[0097] Optionally, in this embodiment, the method of locating the second target device causing the abnormal temperature state from the device components included in the storage device based on the device temperature and / or device temperature relationship is similar to the method of locating the first target device causing the abnormal power consumption state from the device components included in the storage device based on the device power consumption, and will not be described again here.

[0098] As an optional approach, the storage device's first power consumption parameter, first temperature parameter, and first correlation parameter are collected, including:

[0099] S41, collect the current power consumption of the storage device and the first device power consumption of each device in the storage device to obtain the first power consumption parameter, and collect the first device temperature of each device in the storage device;

[0100] S42, calculate the temperature difference parameter and average temperature parameter between the first device temperature of the first device and the first device temperature of the second device, wherein the first temperature parameter includes the first device temperature, temperature difference parameter and average temperature parameter of each device, and the influence of the device temperature of the first device and the device temperature of the second device on the device temperature of the storage device is greater than or equal to that of other devices included in the storage device.

[0101] S43, calculate the correlation between the current power consumption and the average temperature parameter to obtain the first correlation parameter.

[0102] Optionally, in this embodiment, the current power consumption may refer to, but is not limited to, the overall real-time power consumption value of the storage device in its current operating state, which is acquired by the whole disk power consumption detection module circuit.

[0103] Optionally, in this embodiment, the first device and the second device may refer to, but are not limited to, core components in the storage device that have a high degree of influence on the overall device temperature, determined based on the device function and thermal characteristics, such as: SSD main controller (responsible for data processing and control, generating a large amount of heat and having a wide thermal impact range) and NAND Flash chips (numerous, prone to accumulating heat during long-term read and write operations).

[0104] Optionally, in this embodiment, the temperature difference parameter may, but is not limited to, a parameter used to characterize the temperature difference between the first device and the second device, reflecting the temperature uniformity between the core devices. It may, but is not limited to, be obtained by calculating the temperature difference between the two first devices. For example, if the temperature of the first device of the SSD main controller (first device) is 90°C and the temperature of the first device of the NAND Flash chip (second device) is 80°C, then the temperature difference parameter is 10°C.

[0105] Optionally, in this embodiment, the average temperature parameter may be, but is not limited to, a parameter used to characterize the overall thermal state of the storage device. For example, it can be obtained by calculating the average of the measured temperatures of the two core devices most sensitive to temperature (such as the main controller and NAND), or it can be a weighted average calculated based on the first device temperature of the first device and the second device temperature according to a preset weight (such as 0.5 for both).

[0106] Optionally, in this embodiment, correlation may refer to, but is not limited to, a statistical indicator used to quantify the degree of linear correlation between the current power consumption and average temperature parameters of the storage device, such as the Pearson correlation coefficient.

[0107] Optionally, in this embodiment, the acquisition of the first power consumption parameter and the first device temperature of each device can be achieved, but is not limited to, through synchronous acquisition by a multi-channel detection module. For example, such as Figure 4 As shown, the multi-channel power consumption detection module circuits within the storage device (such as the sampling resistor R1 branch connected to the overall disk power supply, the sampling resistor R2 branch connected to the main controller Vcore power supply, the sampling resistor R3 branch connected to the NAND Vcc power supply, and the sampling resistor R4 branch connected to the DRAM VDD power supply) operate synchronously. They acquire the current signals of each branch through the sampling resistors, convert them into digital quantities via an analog-to-digital converter, and then the SSD main controller reads the data through the IIC1 interface. The overall disk power consumption data and the power consumption data of each device branch are summarized into the first power consumption parameter. Simultaneously, four temperature monitoring module circuits (corresponding to the main controller, NAND, DRAM, and PCB board level environments, respectively) collect real-time analog temperature signals of each device through local temperature sensors. After processing and calibration by amplifiers and analog-to-digital converters, the main controller reads the data through the IIC2 interface to obtain the first device temperature of each device. This can also be achieved through time-segmented polling. For example, the main controller polls and reads the total power consumption data of the multi-channel power consumption detection module circuit and the power consumption data of each device branch through the IIC1 interface according to a preset period (e.g., 1 second / time), and integrates them to generate the first power consumption parameter; at the same time, it polls and reads the temperature data of each device collected by the 4-channel temperature monitoring module circuit through the IIC2 interface to obtain the first device temperature of each device. During the polling process, the enable signal of each detection module is controlled through the GPIO (General Purpose Input / Output) interface of the main controller to ensure stable acquisition timing and avoid data conflicts.

[0108] Based on the above, after collecting power consumption and temperature data from the devices, the system further calculates and generates a temperature difference parameter reflecting the system-level heat distribution characteristics (reflecting the temperature uniformity among core components) and an average temperature parameter reflecting the overall thermal state (reflecting the overall heat level of the entire system). These two parameters together constitute the first temperature parameter. Statistical analysis is then used to obtain the first correlation parameter quantifying the relationship between power consumption and temperature. This provides comprehensive and accurate basic data support for subsequent abnormal state detection of storage devices (such as determining whether power consumption or temperature exceeds normal ranges) and abnormal cause localization (such as distinguishing between device failure and heat dissipation anomalies). Simultaneously, relying on modular power consumption detection circuits, temperature monitoring circuits, and the modular calculation logic of the main controller, the real-time performance and reliability of data acquisition are effectively ensured.

[0109] As an optional approach, the first correlation parameter is obtained by calculating the correlation between the current power consumption and the average temperature parameter, including:

[0110] S51, the first correlation parameter is calculated using the following formula:

[0111] ;

[0112] in, Current power consumption, The average temperature parameter, This represents the covariance between the current power consumption and the average temperature parameter. The variance of the current power consumption. This represents the variance of the average temperature parameter.

[0113] Optionally, in this embodiment, Used to measure current power consumption With average temperature parameters The statistical index of the overall error between two variables reflects the degree of consistency in their changing trends, and its calculation formula is as follows:

[0114] ;

[0115] in, The mathematical expectation of the current power consumption, This represents the mathematical expectation of the average temperature parameter. and This refers to the average value of the current power consumption or average temperature parameter over a continuous period of time. It is used to characterize the central trend of a variable. For example, within a sliding time window (such as 60 seconds), the expected value of the current power consumption is the arithmetic mean of all current power consumption data within that window, and the expected value of the average temperature parameter is the arithmetic mean of all current temperature data within that window.

[0116] Optionally, in this embodiment, variance , A statistical indicator used to measure the dispersion of data for a single variable (current power consumption or average temperature parameter), reflecting the fluctuation of the variable data. Its calculation formula is as follows:

[0117] ;

[0118] .

[0119] Optionally, in this embodiment, the calculation of the first correlation parameter can be implemented, but is not limited to, through hardware implementation of a statistical operation unit built into the main controller. For example, the SSD main controller integrates a dedicated statistical operation hardware unit, which reads the current power consumption data sequence and the average temperature parameter data sequence within a sliding time window (e.g., the past 60 seconds) from the data cache and performs the above calculation. It can also be implemented, but is not limited to, through a firmware software algorithm. For example, a statistical analysis program can be written in the firmware of the SSD main controller. The program reads the current power consumption data and the average temperature parameter data within a sliding time window from a specified storage area; calculates the expected value of the current power consumption and the expected value of the average temperature parameter through software loop statements; then calculates the covariance, the variance of the current power consumption, and the variance of the average temperature parameter through nested loops; finally, substitutes these statistics into the formula, calculates the first correlation parameter through a software floating-point arithmetic library, and stores the result in a dedicated register.

[0120] The above content, namely, by accurately calculating the linear correlation between the current power consumption and the average temperature parameter, obtains the first correlation parameter, which provides a quantitative basis for subsequent anomaly detection of the correlation between power consumption and temperature of storage devices, ensuring the accuracy and scientific nature of the correlation analysis, further supporting the multi-dimensional diagnosis of abnormal states of storage devices, and improving the refinement and intelligence level of storage device health management.

[0121] As an optional approach, the current power consumption of the storage device and the first device power consumption of each device in the storage device are collected to obtain a first power consumption parameter, and the first device temperature of each device in the storage device is collected, including:

[0122] S61, receive the current power consumption reported by the first power consumption detection module, receive the first device power consumption reported by the second power consumption detection module corresponding to each device, and obtain the first power consumption parameter, wherein the first power consumption detection module is used to detect the power consumption generated by the storage device, and the second power consumption detection module is used to detect the power consumption generated by the corresponding device.

[0123] S62, receive the temperature of the first device reported by the first temperature sensor corresponding to each device, wherein the first temperature sensor is used to collect the temperature of the corresponding device.

[0124] Optionally, in this embodiment, the first power consumption detection module and the second power consumption detection module may, but are not limited to, consist of discrete sampling resistors (such as...). Figure 4 The circuit consists of R1, R2, R3, and R4 in the circuit, along with a differential amplifier and an ADC (analog-to-digital converter). The first power consumption detection module is used to monitor the total input power consumption (i.e., current power consumption) of the entire storage device (such as an SSD), while the second power consumption detection module is differentially deployed on the power supply branches of each core device (such as the main controller, NAND, and DRAM) to accurately measure the first device power consumption of each core device and report the power consumption data to the disk controller through the IIC interface.

[0125] Optionally, in this embodiment, the first temperature sensor (such as...) Figure 4 The 4-channel temperature monitoring module circuit in the storage device can be, but is not limited to, temperature measuring elements deployed on or near the surface of various devices inside the storage device. It is used to collect the real-time temperature of the first device at the corresponding location. Its temperature measurement range is matched with the operating temperature requirements of the device, enabling accurate temperature measurement. It also reports the temperature data to the controller inside the disk through the IIC interface.

[0126] Optionally, in the embodiments of this application, the step of obtaining the first power consumption parameter may include, but is not limited to, the following: the onboard controller reads the digital value in the register of the first power consumption detection module through the IIC interface to obtain the current power consumption; at the same time, it reads the register data of each second power consumption detection module through different IIC slave addresses to obtain the first device power consumption of the corresponding device; and it summarizes the current power consumption with the power consumption of all first devices to generate a first power consumption parameter that includes the power consumption of multiple nodes.

[0127] Through the above-mentioned methods—namely, the first and second power consumption detection modules deployed on the entire disk and each key device, and the first temperature sensor deployed on each device—the overall power consumption, key device power consumption, and key device temperature of the storage device can be collected in real time and synchronously. Compared to related technologies that only collect single-point temperature or overall disk power consumption, this solution obtains multi-dimensional and multi-node data, thus providing a more comprehensive and detailed characterization of the real-time operating status and internal heat distribution of the storage device. This improves the comprehensiveness of the monitoring data and provides the necessary data foundation for subsequent refined deviation calculations and accurate fault root cause localization.

[0128] As an optional approach, the storage device is detected under the current operating scenario, including:

[0129] S71, obtain the idle power consumption, device bandwidth, bandwidth unit power consumption, device read / write parameters and number of times unit power consumption of each device in the storage device, as well as the ambient temperature of the storage device and the thermal resistance coefficient of each device. Among them, the device read / write parameters are the number of read / write operations per second, the bandwidth unit power consumption is the power consumption generated per unit bandwidth, and the number of times unit power consumption is the power consumption generated per second by the device read / write parameters.

[0130] S72 calculates the second power consumption parameter corresponding to each device based on idle power consumption, device bandwidth, power consumption per unit bandwidth, device read / write parameters, and power consumption per unit of number of reads / writes.

[0131] S73 calculates the second temperature parameters corresponding to each device based on the ambient temperature, thermal resistance coefficient, and second power consumption parameter.

[0132] Optionally, in this embodiment, idle power consumption may refer to, but is not limited to, the basic power consumption of a single device in the storage device when there is no workload (such as no read / write operations or background tasks), which is the static power consumption determined by the physical characteristics of the device itself.

[0133] Optionally, in this embodiment, device bandwidth may refer to, but is not limited to, the actual data bandwidth processed by a single device in the storage device under the current operating scenario, used to reflect the load intensity of the device, and is measured in GB / s. Bandwidth unit power consumption may refer to, but is not limited to, the power consumption generated by a single device in the storage device per unit bandwidth of data processed, and is a dynamic power consumption characteristic parameter of the device.

[0134] Optionally, in this embodiment, the device read / write parameters may refer to, but are not limited to, parameters used to measure the intensity of read / write operations of a single device in the storage device, such as the number of read / write operations per second. The power consumption per unit of number of operations may refer to, but is not limited to, the power consumption generated by the device read / write parameters per second for a single device in the storage device, reflecting the dynamic power consumption of the device due to the number of read / write operations. Examples include the power consumption generated by a NAND Flash chip for each read / write operation, the power consumption generated by a DRAM chip for each read / write operation, etc., with the unit being W / IOPS (where IOPS is the number of input / output operations per second).

[0135] Optionally, in this embodiment, the ambient temperature may refer to, but is not limited to, the real-time temperature of the external environment where the storage device is located, such as the ambient temperature inside the server chassis, the operating ambient temperature of the edge computing device, etc.

[0136] Optionally, in this embodiment, the thermal resistivity may refer to, but is not limited to, the resistance parameter in which heat is transferred from the inside of a single device in the storage device to the external environment, reflecting the heat dissipation characteristics of the device. The unit is °C / W (degrees per watt), indicating how many degrees Celsius the core temperature of the device will rise relative to the ambient temperature for every watt of power consumed. For example, a thermal resistivity of 6.0°C / W means that for every watt of power consumed, the core temperature of the device will rise by 6 degrees Celsius relative to the ambient temperature.

[0137] Optionally, in this embodiment, the acquisition of parameters such as idle power consumption and device bandwidth of each device can be achieved, but is not limited to, through the following methods: Before the storage device is used, the anomaly monitoring system pre-stores static parameters such as idle power consumption, bandwidth unit power consumption, cycle unit power consumption, and thermal resistance coefficient of each device in the non-volatile memory (such as EEPROM (Electrically Erasable Programmable Read-Only Memory)) of the storage device; during operation, the on-disk controller of the storage device retrieves these static parameters from the non-volatile memory through the IIC interface; at the same time, the multi-channel power consumption detection module circuit of the anomaly monitoring system collects the device bandwidth (such as the actual read and write bandwidth of NAND) in real time, the multi-channel temperature monitoring module circuit or external sensors collect the ambient temperature in real time, and the internal counting module of the storage device counts the device read and write parameters (such as IOPS). For example, the on-disk controller of the storage device retrieves the idle power consumption of NAND Flash from EEPROM as 1W, bandwidth unit power consumption as 2W / (GB / s), number of times unit power consumption as 0.0001W / IOPS, and thermal resistance coefficient as 2℃ / W; at the same time, the multi-channel power consumption detection module circuit of the abnormal monitoring system collects the device bandwidth of NAND in real time as 2GB / s, the multi-channel temperature monitoring module circuit collects the ambient temperature as 30℃, and the internal counting module of the storage device counts the device read and write parameters (IOPS) as 10000.

[0138] Optionally, in this embodiment, the idle power consumption, device bandwidth, and other parameters of each device can also be obtained in the following ways, but are not limited to: the disk controller of the storage device calculates the theoretical bandwidth power consumption per unit and the power consumption per number of times of each device according to the current operating scenario (such as 2GB / s sequential write); at the same time, the power consumption detection module circuit of the anomaly monitoring system collects the device bandwidth, the temperature monitoring module circuit collects the ambient temperature, and the internal counting module of the storage device counts the device read and write parameters.

[0139] Optionally, in this embodiment, the calculation of the second power consumption parameters of each device based on idle power consumption, etc., can be implemented, but is not limited to, the following methods: the on-board controller of the storage device has a built-in calculation unit that directly calculates the second power consumption parameters according to a preset formula. Alternatively, the storage device pre-stores power consumption models of each device under different operating scenarios, and the on-board controller calls the corresponding power consumption model based on the currently collected device bandwidth and device read / write parameters to calculate the second power consumption parameters of each device.

[0140] Optionally, in this embodiment, the calculation of the second temperature parameters of each device based on ambient temperature, etc., can be achieved, but is not limited to, by the following methods: the storage device's in-disk controller has a built-in calculation unit that directly calculates the second temperature parameters according to a preset formula. Alternatively, the storage device pre-stores temperature models for each device, and the in-disk controller, based on the ambient temperature and the calculated second power consumption parameters, calls the corresponding temperature model to calculate the second temperature parameters of each device.

[0141] Based on the above, by obtaining the static characteristic parameters (idle power consumption, thermal resistance coefficient, etc.) and dynamic load parameters (device bandwidth, device read / write parameters, etc.) of each device included in the storage device, the second power consumption parameter and the second temperature parameter allowed to be generated by each device under the current operating scenario are accurately obtained. This provides an accurate expected benchmark for the subsequent calculation of power consumption deviation and temperature deviation of the storage device, effectively distinguishing between normal high temperature under high load and true abnormal heat generation, and significantly reducing false alarms.

[0142] As an optional approach, the second power consumption parameters for each device are calculated based on idle power consumption, device bandwidth, power consumption per unit bandwidth, device read / write parameters, and power consumption per unit of operation, including:

[0143] S81 calculates the second power consumption parameter for each device using the following formula:

[0144] ;

[0145] in, This is the idle power consumption. For device bandwidth, Power consumption per unit bandwidth For device read / write parameters, Power consumption is measured in units of frequency.

[0146] Optionally, in this embodiment, the power consumption per unit bandwidth of the device is calculated using the following formula:

[0147] ;

[0148] in, This represents the maximum average power consumption of the corresponding device components. This represents the maximum bandwidth that the corresponding device can achieve under maximum sequential write load.

[0149] Optionally, in this embodiment, the power consumption per unit of the device is calculated using the following formula:

[0150] ;

[0151] in, The maximum value that the corresponding device can achieve under maximum random write load. .

[0152] Optionally, in this embodiment, "Can" refers to, but is not limited to, the expected power consumption that a single device within a storage device is allowed to generate when it is in a healthy and normal operating state under the current operating scenario.

[0153] Based on the above, using pre-stored static parameters of devices and real-time acquired dynamic load parameters, the second power consumption parameters of each device are accurately calculated using standardized formulas. This provides a unified and accurate expected benchmark for subsequent power consumption deviation calculations of storage devices, ensuring the scientific and accurate nature of power consumption anomaly detection.

[0154] As an optional approach, the second temperature parameters corresponding to each device are calculated based on the ambient temperature, thermal resistance coefficient, and second power consumption parameter, including:

[0155] S91, calculate the second temperature parameter corresponding to each device component using the following formula:

[0156] ;

[0157] in, For ambient temperature, The thermal resistance coefficient, This is the second power consumption parameter.

[0158] Optionally, in this embodiment, This can refer to, but is not limited to, the expected temperature value that a single device within a storage device is allowed to generate when it is in a healthy and normal operating state under the current operating scenario.

[0159] Based on the above, and considering the ambient temperature, thermal resistance characteristics of devices, and the expected power consumption of devices under the current operating scenario, the second temperature parameters of each device are accurately calculated using standardized formulas. This provides a unified and accurate expected benchmark for subsequent storage device temperature deviation calculations and can effectively distinguish between abnormal storage device temperatures caused by normal high loads and abnormal temperatures caused by hardware-level faults.

[0160] As an optional approach, abnormal device states of storage devices are detected based on power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters, including:

[0161] S101, detect the operating deviation parameters of the storage device based on the power consumption deviation parameters, temperature deviation parameters and correlation deviation parameters, wherein the operating deviation parameters are used to indicate the degree of deviation of the operating state of the storage device from the healthy operating state;

[0162] S102, when the degree of deviation indicated by the operating deviation parameter is greater than or equal to the deviation degree threshold, determine the abnormal state of the device to indicate that there is an abnormality in the storage device;

[0163] S103, if the degree of deviation indicated by the operating deviation parameter is less than the deviation degree threshold, determine the abnormal state of the device to indicate that there is no abnormality in the storage device.

[0164] Optionally, in this embodiment, the operating deviation parameter may refer to, but is not limited to, a parameter used to comprehensively quantify the degree of deviation of the current operating state of the storage device from the overall healthy operating state. It is obtained by integrating power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters, and can comprehensively reflect the deviation of the device in power consumption, temperature, and the logical relationship between the two. For example, dimensionless values ​​such as 0.1777 and 0.44397 calculated by weighted summation formula. The larger the value, the more serious the deviation of the device's operating state from the healthy state.

[0165] Optionally, in this embodiment, a healthy operating state may refer to, but is not limited to, the operating state of the storage device within its design specifications, where the power consumption, temperature, and power consumption-temperature correlation of each core component are all within the normal range. For example, the overall power consumption meets the expected value under the current scenario, the temperature of the core components does not exceed the maximum operating threshold (e.g., the maximum operating threshold of the main controller is 105°C, the maximum operating threshold of NAND is 85°C, etc.), and power consumption and temperature are strongly positively correlated (the Pearson correlation coefficient is close to 1).

[0166] Optionally, in this embodiment, the deviation threshold may be, but is not limited to, a critical value used to determine whether the operating deviation of the storage device exceeds the normal range. This threshold can be flexibly adjusted according to the application scenario (such as server, edge computing device) and reliability requirements of the storage device, and is pre-stored in the non-volatile memory of the storage device.

[0167] Optionally, in this embodiment, the operating deviation parameters of the storage device can be detected in the following ways, but are not limited to: weighted summation integration, where power consumption deviation weight, temperature deviation weight, and correlation deviation weight are pre-set (the sum of the weights is 1), and the three deviation parameters are multiplied by their corresponding weights and then summed to obtain the comprehensive operating deviation parameters. Alternatively, hierarchical superposition integration, where each deviation parameter is first divided into deviation levels (e.g., mild, moderate, severe) and assigned a corresponding score, and then the scores of the three parameters are superimposed, with the final superposition result serving as the operating deviation parameter. Alternatively, threshold-triggered integration, where a trigger threshold is set for each deviation parameter, and when any one or more deviation parameters exceed the corresponding threshold, the operating deviation parameters are quantified and generated according to the number and severity of the triggered deviation parameters.

[0168] Optionally, in this embodiment, after obtaining the operating deviation parameters, the comparison with the deviation degree threshold can be completed through either hardware logic or software program to determine the abnormal state of the equipment.

[0169] Optionally, in this embodiment, the abnormal state of the device is determined by the threshold comparison logic circuit built into the disk controller. This includes: the disk controller of the storage device retrieves a preset deviation threshold from the non-volatile memory and transmits the calculated operating deviation parameter to the built-in threshold comparison logic circuit; the circuit compares the hardware-level signals, and if the operating deviation parameter is detected to be ≥ the threshold, it immediately outputs the judgment result "the device is abnormal", triggers an abnormality flag, and stores the result in the health status register for subsequent abnormality cause location steps; if the operating deviation parameter is detected to be < the deviation threshold, it outputs the judgment result "the device is not abnormal", and outputs a normal status signal through the GPIO interface to instruct the storage device to maintain the current working mode (e.g., performance and bandwidth are running at 100%).

[0170] Optionally, in this embodiment, the device abnormal state is determined by the firmware program, including: the firmware program of the disk controller reads the deviation threshold from the specified storage address, compares the operating deviation parameter with the threshold through software condition judgment statements; if the operating deviation parameter is greater than or equal to the deviation threshold, the firmware program generates a "device is abnormal" status code, reports it to the BMC through the SMBUS out-of-band interface, and simultaneously starts the abnormal log recording process to record the specific values ​​of the current power consumption deviation parameter, temperature deviation parameter, correlation deviation parameter, and operating deviation parameter; if the device is determined to be normal, the firmware program generates a "device is operating normally" status code, updates the health status database of the storage device, and notifies each detection module circuit (power consumption detection module circuit, temperature monitoring module circuit) to maintain the current data acquisition frequency through the internal communication link, without needing to start an additional abnormal response process.

[0171] Based on the above, the operational deviation parameters formed by integrating multi-dimensional deviation parameters are used to determine anomalies. This allows for the adaptation of health benchmarks under different operating conditions, thereby achieving accurate determination of abnormal device states. Furthermore, the operational deviation parameters are generated based on multi-dimensional deviation parameters, which can trace back to the specific dimensions affecting the deviation (power consumption, temperature, or their correlation), providing direction for fault root cause location. This avoids the shortcomings of related technologies that can only rely on manual fault root cause location, effectively improving the reliability of solid-state drive health management.

[0172] As an optional approach, the operating deviation parameters of the storage device are detected based on power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters, including:

[0173] S111, the operating deviation parameter is calculated using the following formula:

[0174] ;

[0175] in, This is a power consumption deviation parameter. This is a temperature deviation parameter. This is the correlation deviation parameter. The first weighting coefficient for the power consumption deviation parameter is... This is the second weighting coefficient for the temperature deviation parameter. The third weighting coefficient is the correlation deviation parameter. The weighting coefficient is used to indicate the degree of influence of the corresponding deviation parameter on the operation of the storage device in a healthy operating state. The sum of the first weighting coefficient, the second weighting coefficient and the third weighting coefficient is 1.

[0176] Optionally, in this embodiment, the weighting coefficient may refer to, but is not limited to, the parameter used to distinguish power consumption deviation. Temperature deviation parameters Correlation deviation parameter The coefficients that determine the degree of impact on the healthy operating status of storage devices can be determined in two ways: one is a fixed coefficient that is pre-set and stored in EEPROM; the other is a coefficient that is dynamically determined based on the current operating scenario of the storage device. For example, in high-reliability scenarios (such as server storage and critical business data storage), where priority needs to be given to the impact of temperature anomalies on device stability, the coefficients can be dynamically set. And regardless of whether the setting is fixed or dynamic, it satisfies The constraints.

[0177] Optionally, in this embodiment, the operating deviation parameters of the storage device can be detected, but are not limited to, through the following methods: the on-disk controller calls pre-stored weighting coefficients, reads the calculated power consumption deviation parameters, temperature deviation parameters, and correlation deviation parameters from the data register, and executes the formula for calculating the operating deviation parameters through the built-in hardware processing unit or firmware program to obtain the operating deviation parameters. Alternatively, depending on the application scenario of the storage device (such as a high-reliability storage scenario), the weighting coefficients can be adjusted to increase the influence ratio of key deviation parameters. If temperature anomalies are given priority, they can be set... , , The controller inside the control panel reads the three types of deviation parameters and performs a weighted summation according to the adjusted weights to obtain the operating deviation parameters. For example: , , Substituting the adjusted weights, we obtain the operating deviation parameter. This value is the current operating deviation parameter.

[0178] By employing the above method, the three single-dimensional deviation parameters—power consumption, temperature, and correlation—are transformed into a comprehensive operational deviation parameter through a weighted summation formula. This solves the problem in related technologies that rely solely on a single threshold for judgment and cannot identify whether a device is faulty. Furthermore, the flexible setting of weighting coefficients can adapt to the priority requirements of different application scenarios regarding device health dimensions, ensuring that the operational deviation parameter accurately focuses on key influencing factors and improving the comprehensiveness and accuracy of storage device operational status assessment.

[0179] As an optional approach, the power consumption deviation parameter is determined based on the first power consumption parameter and the second power consumption parameter, including:

[0180] S121, obtain the device power consumption deviation of each device in the storage device, and the power consumption weight coefficient of each device. The device power consumption deviation is used to indicate the degree of deviation of the first device power consumption of the corresponding device relative to the second device power consumption in the current operating scenario. The device power consumption deviation is the quotient between the device power consumption difference and the second device power consumption. The device power consumption difference is the absolute value of the difference between the first device power consumption and the second device power consumption. The power consumption weight coefficient is used to indicate the degree of impact of the device power consumption deviation of the corresponding device on the operation of the storage device in a healthy operating state.

[0181] S122, the sum of the abnormal power consumption parameters of each device is determined as the power consumption deviation parameter, wherein the abnormal power consumption parameter is the product of the power consumption deviation of the corresponding device and the power consumption weighting coefficient.

[0182] Optionally, in this embodiment, device power consumption deviation may refer to, but is not limited to, a parameter used to quantify the degree of deviation between the current power consumption of the first device and the power consumption of the second device in a single device. The larger the value, the more serious the deviation of the power consumption of a single device from the healthy expectation. The calculation formula is as follows:

[0183] .

[0184] Optionally, in this embodiment, the power consumption deviation parameter is calculated using the following formula:

[0185] ;

[0186] in, This refers to the total number of core components involved in power consumption deviation calculations within the storage device. For example, if an SSD contains three types of core components: a main controller, NAND, and DRAM, then... ; Less than or equal to A positive integer used to distinguish different core device components within the storage device; For the first The power consumption weighting coefficient of the first device component is used to indicate the power consumption weighting coefficient of the first component. The extent to which the power consumption deviation of individual devices affects the overall healthy operating status of the storage device; For the first The first device power consumption of a device refers to the power consumption value generated by the device during actual operation, which is collected in real time by the multi-channel power consumption detection module circuit in the anomaly monitoring system. For the first The second device power consumption of a device refers to the expected power consumption value of the device when it is in a healthy state under the current operating scenario. It is calculated based on parameters such as idle power consumption, device bandwidth, and power consumption per unit bandwidth. For the first The power consumption deviation of each device is calculated by dividing the absolute difference between the actual power consumption and the expected power consumption by the expected power consumption. This is used to eliminate the problem of quantification injustice caused by differences in the power consumption base of different devices.

[0187] Optionally, in this embodiment, the power consumption weighting coefficient may be, but is not limited to, a coefficient used to distinguish the degree of influence of the power consumption deviation of different devices on the overall healthy operation of the storage device. For example, the power consumption weighting coefficient of the SSD main controller as the core device is set to 0.5; the weighting coefficient of the NAND Flash chip as the storage core is set to 0.3; the weighting coefficient of the DRAM chip is set to 0.2, and the sum of the power consumption weighting coefficients of all devices is 1.

[0188] Optionally, in this embodiment, the abnormal power consumption parameter can be, but is not limited to, the product of the power consumption deviation of a single device and the corresponding power consumption weighting coefficient, used to quantify the actual impact of the power consumption deviation of a single device on the overall health status of the disk. For example, if the power consumption deviation of the SSD main controller is 0.1765 and the power consumption weighting coefficient is 0.5, then the abnormal power consumption parameter is calculated as 0.1765 × 0.5 = 0.08825.

[0189] As an optional approach, the temperature deviation parameter is determined based on the first temperature parameter and the second temperature parameter, including:

[0190] S131, obtain the device temperature deviation and temperature weighting coefficient of each device in the storage device, and obtain the temperature difference abnormal parameter and temperature difference weighting coefficient between the first device and the second device. The device temperature deviation is used to indicate the degree of deviation of the first device temperature of the corresponding device relative to the second device temperature in the current operating scenario. The device temperature deviation is the quotient between the device temperature difference and the second device temperature. The device temperature difference is the absolute value of the difference between the first device temperature and the second device temperature. The temperature weighting coefficient is used to indicate the degree of influence of the device temperature deviation of the corresponding device on the operation of the storage device in a healthy operating state. The influence of the device temperature of the first device and the device temperature of the second device on the device temperature of the storage device is greater than or equal to that of other devices in the storage device. The temperature difference abnormal parameter is used to indicate the degree of deviation of the first temperature difference parameter between the first device and the second device relative to the second temperature difference parameter in the current operating scenario. The temperature difference weighting coefficient is used to indicate the degree of influence of the temperature difference abnormal parameter on the operation of the storage device in a healthy operating state.

[0191] S132, the sum of the device temperature abnormality parameter and the device temperature difference abnormality parameter of each device is determined as the temperature deviation parameter, wherein the device temperature abnormality parameter is the product of the device temperature deviation of the corresponding device and the temperature weighting coefficient, and the device temperature difference abnormality parameter is the product of the temperature difference abnormality parameter and the temperature difference weighting coefficient.

[0192] Optionally, in this embodiment, device temperature deviation may refer to, but is not limited to, a parameter used to quantify the degree of deviation between the current temperature of the first device and the temperature of the second device in a single device. A larger value indicates a more severe deviation of the single device temperature from the expected healthy temperature. The calculation formula is as follows:

[0193] .

[0194] Optionally, in this embodiment, the temperature deviation parameter is calculated using the following formula:

[0195] ;

[0196] in, This refers to the total number of core components within the storage device involved in temperature deviation calculations. For example, if an SSD contains three types of core components: the main controller, NAND, and DRAM, then... ; Less than or equal to A positive integer used to distinguish different core device components within the storage device; For the first The temperature weighting coefficient of each device component is used to indicate the temperature weighting coefficient of the first component. The extent to which the temperature deviation of individual devices affects the overall healthy operating status of the storage device; For the first The first device temperature of a device refers to the temperature value generated by the device during actual operation, which is collected in real time by the multi-channel temperature detection module circuit of the anomaly monitoring system. For the first The second device temperature of a device refers to the expected temperature value of the device when it is in a healthy state under the current operating scenario, which is calculated based on the link temperature, thermal resistance coefficient and expected power consumption. For the first The device temperature deviation of each device is calculated by dividing the absolute difference between the actual device temperature and the expected temperature by the expected temperature. This is used to eliminate the problem of inconsistent deviation quantification caused by differences in the temperature base of different devices. This is the temperature difference weighting coefficient, used to quantify the impact of abnormal temperature difference parameters of core components on the overall healthy operating status of the disk. The first temperature difference parameter refers to the current actual temperature difference between the two types of core components (first device components and second device components, such as the main controller and NAND Flash) that have the greatest impact on the overall temperature within the storage device. The second temperature difference parameter refers to the expected temperature difference between the first and second devices when they are in a healthy state under the current operating scenario. This refers to the abnormal parameters of temperature difference.

[0197] Optionally, in this embodiment, the temperature weighting coefficient may be, but is not limited to, a coefficient used to distinguish the degree of influence of the temperature deviation of different devices on the overall healthy operation of the storage device. The weighting coefficient of core devices (such as SSD main controller and NAND Flash) is higher than that of other devices, and the sum of the temperature weighting coefficient and the temperature difference weighting coefficient of all devices is 1. For example, the temperature weighting coefficient of the main controller is 0.4, the temperature weighting coefficient of NAND Flash is 0.3, and the total temperature weighting coefficient of other devices is 0.3.

[0198] Optionally, in this embodiment, the temperature difference abnormality parameter may, but is not limited to, a parameter used to quantify the degree of deviation between the current actual temperature difference (first temperature difference parameter) between the first device and the second device and the expected healthy temperature difference (second temperature difference parameter), and its calculation formula is as follows:

[0199] .

[0200] Through the above, it is found that abnormal monitoring of storage devices can quantify the degree of deviation between power consumption and temperature at the device level, and distinguish the impact of temperature differences of different devices and core devices on the overall health status of the disk by using weighting coefficients. This enables refined identification of abnormalities in storage devices (it can identify hardware-level faults and pinpoint the specific device or core device with abnormal temperature difference). This avoids the shortcomings of related technologies that cannot identify hardware-level faults, leading to untimely fault handling and equipment damage. It improves the accuracy of abnormal monitoring of storage devices and enables preliminary location of fault root causes, providing device-level data support for early risk prediction.

[0201] As an optional approach, after acquiring the first power consumption parameter of the storage device, the method further includes:

[0202] S141, Receive the power consumption configuration of the storage device, wherein the power consumption configuration is used to configure the power consumption level of the storage device to the desired power consumption level;

[0203] S142, detect the target power level corresponding to the current power consumption in the first power consumption parameter from the power consumption and power consumption level with corresponding relationship, wherein the current power consumption is the power consumption of the storage device collected;

[0204] S143, adjust the operating parameters of the storage device from the target operating parameters that meet the target power consumption level to the reference operating parameters that meet the desired power consumption level.

[0205] Optionally, in this embodiment, the power consumption configuration can be, but is not limited to, commands or settings from the upper-level management system (such as the off-board controller BMC or the host operating system). For example, a "Set Power Consumption" command sent via the out-of-band SMBUS bus, or a value issued by the "Set Features - Power Management" command in the in-band NVMe (Non-Volatile Memory Express) command set, can be used to specify the desired power consumption level.

[0206] Optionally, in this embodiment, the power consumption levels may be, but are not limited to, a series of predefined operating state levels for the storage device to balance performance and power consumption. These levels (e.g., S0 to S5) and their corresponding power consumption values ​​(e.g., 6W to 22W) are predefined and stored in the firmware, forming a mapping table or lookup table of power consumption and power consumption levels with corresponding relationships (e.g., {S0-6W; S1-9W; S2-12W; S3-15W; S4-18W; S5-22W}).

[0207] Optionally, in this embodiment, the target power consumption level can be, but is not limited to, the current power consumption level obtained by the storage device based on the current real-time power consumption (i.e., the total power consumption in the first power consumption parameter) and by looking up the mapping table of the above-mentioned "power consumption and power consumption level correspondence". For example, if the current power consumption is 17W, after looking up the mapping table, it is determined that it falls between S3 (15W) and S4 (18W), and the current target power consumption level can be determined to be S4 (or S3, depending on the definition of the interval). This solution does not limit this.

[0208] Optionally, in this embodiment, the operating parameters may be, but are not limited to, various adjustable configurations used within the storage device to control its performance and power consumption. Examples include: the operating frequency of the main controller, the channel concurrency or interface timing of the NAND Flash, the operating frequency or refresh rate of the DRAM, the link speed or link width of the PCIe (Peripheral Component Interconnect Express) interface, and the operating mode of the internal DC / DC converter.

[0209] Optionally, in this embodiment, both the target operating parameters and the reference operating parameters are specific sets of values ​​for the aforementioned operating parameters under specific states. The target operating parameters refer to a set of operating parameter values ​​currently used by the storage device when operating at a target power consumption level; the reference operating parameters refer to a new set of operating parameter values ​​that the storage device needs to be adjusted to in order to achieve the desired power consumption level. For example, the target operating parameters might be {controller frequency of 1.2GHz, NAND concurrency of 8}, and the reference operating parameters might be {controller frequency of 1.0GHz, NAND concurrency of 6}.

[0210] As an optional solution, adjusting the operating parameters of the storage device from the target operating parameters that meet the target power consumption level to the reference operating parameters that meet the desired power consumption level includes:

[0211] S151, compare the target power consumption level with the desired power consumption level, wherein a higher power consumption level indicates a higher power consumption of the storage device;

[0212] S152, if the target power consumption level is higher than the desired power consumption level, determine the first operating parameter corresponding to the storage service being run by the storage device; reduce the first operating parameter until the power consumption of the storage device is collected and falls into the desired power consumption level;

[0213] S153, if the target power consumption level is lower than the desired power consumption level, determine the second operating parameter corresponding to the storage service being run by the storage device; increase the second operating parameter until the power consumption of the storage device is collected and falls into the desired power consumption level.

[0214] Optionally, in this embodiment, the storage service can be, but is not limited to, the specific I / O task type currently being executed by the storage device. For example, high-bandwidth sequential read / write services (such as video stream processing), high-IOPS random read / write services (such as database queries), and mixed read / write services.

[0215] Optionally, in this embodiment, the first operating parameter and the second operating parameter may be, but are not limited to, subsets or specific combinations of operating parameters. The first operating parameter is a set of one or more operating parameters that need to be reduced when the target power consumption is higher than the expected power consumption, such as the number of concurrent NAND channels and the operating frequency of the main control core. The second operating parameter is a set of one or more operating parameters that need to be increased when the target power consumption is lower than the expected power consumption, such as the operating frequency of the main control core and the PCIe link speed. To achieve different adjustment targets (reduce power consumption / increase power consumption), the set of parameters included in the first operating parameter and the set of parameters included in the second operating parameter may be the same or different.

[0216] Optionally, in this embodiment, comparing the target power consumption level and the desired power consumption level can be achieved, but is not limited to, by the following methods: direct numerical comparison. For example, power consumption levels are encoded as S0, S1, S2...S5, where a larger number after S indicates higher power consumption. The in-panel controller can directly compare the numerical values ​​of the target power consumption level (e.g., S4) and the desired power consumption level (e.g., S2) (4>2) to determine that the target power consumption level is higher than the desired power consumption level. Alternatively, comparing the corresponding power consumption values. For example, the controller queries a mapping table to obtain the center power consumption value (18W) corresponding to the target power consumption level (S4) and the center power consumption value (12W) corresponding to the desired power consumption level (S2), and determines the level by comparing 18W>12W.

[0217] Optionally, in this embodiment, determining and reducing the first operating parameters corresponding to the storage service being run by the storage device can be, but is not limited to, downgrading according to a preset strategy table. For example, the power reduction strategy table is stored in the firmware of the disk controller. This table defines which first operating parameters should be reduced first under different storage services (such as "sequential write" at present) (e.g., reducing NAND concurrency first, and then reducing the main control frequency). The controller queries this table according to the current storage service to determine that the NAND concurrency (first operating parameter) should be reduced first, and then it is reduced step by step from 8 channels to 6 channels. The current power consumption is re-collected after each reduction step, and it is checked whether the power consumption falls within the range of the expected power consumption level (such as S2) until the condition is met.

[0218] Optionally, in this embodiment, determining and increasing the second operating parameters corresponding to the storage service being run by the storage device can be, but is not limited to, upgrading according to a preset strategy table. For example, assuming the current target is S1 and the desired value is S3, the firmware queries the power consumption upgrade strategy table to determine the second operating parameter that should be prioritized for the current storage service (such as "random read") (e.g., prioritizing increasing the DRAM frequency, then increasing the main control frequency). The controller increases these parameters step by step, and re-collects the current power consumption after each step until the power consumption value falls within the range of the desired power consumption level (S3).

[0219] Through the above methods, based on real-time data feedback from multi-channel power consumption detection circuits, storage device power management is achieved, enabling users to proactively adjust power consumption levels according to business needs. Storage devices no longer passively operate in a single power consumption mode but can respond to power configurations issued by the upper-level management system. By detecting the gap between current power consumption and expected power consumption in real time, and proactively reducing or increasing key internal operating parameters based on this gap, the overall power consumption of the storage device is ultimately stabilized at the user's desired level. This achieves a flexible balance between storage device performance and power consumption in different application scenarios (such as pursuing high performance or focusing on energy saving), realizing the technical effect of refined and configurable energy efficiency management.

[0220] As an optional approach, after acquiring the first power consumption parameter and the first temperature parameter of the storage device, the method further includes:

[0221] S161, compare the power consumption of each device included in the first power consumption parameter with the corresponding power consumption threshold, compare the temperature of each device included in the first temperature parameter with the corresponding first temperature threshold, and compare the collected board temperature of the storage device with the second temperature threshold.

[0222] S162, if the power consumption of at least one of the first device power consumptions is greater than or equal to the corresponding power consumption threshold, the temperature of at least one of the first device temperatures is greater than or equal to the corresponding first temperature threshold, and / or the board temperature is greater than or equal to the second temperature threshold, control the storage device to power off.

[0223] Optionally, in this embodiment, the power consumption threshold may be, but is not limited to, an absolute power consumption limit set for each device (such as a controller, NAND, or DRAM). This threshold is typically determined based on the device's electrical specifications; for example, it may be set to 1.25 times the device's rated maximum power consumption. Exceeding this value usually indicates a possible electrical anomaly (such as a transient short circuit or a serious device failure).

[0224] Optionally, in this embodiment, the first temperature threshold and the second temperature threshold may be, but are not limited to, fixed upper temperature limits used to trigger emergency over-temperature protection. The first temperature threshold is set for core device components (such as the controller or NAND flash memory), for example: the first temperature threshold for the controller is 105°C, and the first temperature threshold for the NAND flash memory is 85°C. The second temperature threshold is set for the device board (i.e., the PCB board of the SSD).

[0225] Optionally, in this embodiment, comparing the power consumption of each device included in the first power consumption parameter with the corresponding power consumption threshold can be achieved, but is not limited to, through periodic comparison by the disk controller firmware. For example, in its main loop, the disk controller executes a hardware health check task with high priority (e.g., every 100 milliseconds). This task obtains the latest power consumption and temperature of the first device (including board temperature) from the power consumption detection module circuit and the temperature monitoring module circuit, and reads the corresponding power consumption threshold, first temperature threshold, and second temperature threshold from the EEPROM or firmware code, and then performs software condition judgments one by one (e.g., if(T_ctrl>105)). Correspondingly, when any software comparison result is true (i.e., exceeding the limit), and this state continues for more than a preset time (e.g., 5 minutes), the disk controller confirms that this is a persistent serious fault, and then outputs a control signal (e.g., low level) to the EN pin of the backup power management module circuit through its GPIO1 pin to actively trigger the controlled power-off protection of the storage device.

[0226] Optionally, in this embodiment, comparing the power consumption of each device included in the first power consumption parameter with the corresponding power consumption threshold can also be achieved automatically, but is not limited to, through a hardware comparator in the detection module. For example, the power consumption detection module circuit and the temperature monitoring module circuit typically have programmable threshold registers and hardware comparators internally. During initialization, the in-disk controller writes the aforementioned power consumption threshold and the first / second temperature threshold to these registers via IIC. Subsequently, the hardware comparator automatically and in real-time compares the acquired value with the threshold without firmware intervention. Correspondingly, when the hardware comparator detects an out-of-limit condition, its ALERT pin (e.g., ...) triggers an alarm. Figure 4 The R7 and R8 branches in the circuit are immediately pulled low (or high) by the hardware. This ALERT signal does not go through the main controller firmware, but is directly connected to the EN pin of the backup power management module circuit through the hardware circuit, realizing millisecond-level hardware automatic power-off protection, with a response speed much faster than firmware control.

[0227] Furthermore, those skilled in the art will understand that the methods for implementing this operation are not limited to the two methods described above. For example, power-off can also be achieved by controlling the enable pin of a dedicated external load switch. Any method that can disconnect the main power supply to the storage device or put it into a safety protection state falls within the scope of this invention.

[0228] By comparing the power consumption and temperature of each key component, and whether the board temperature exceeds preset absolute safety limits (power consumption threshold, first temperature threshold, second temperature threshold), extreme electrical or thermal anomalies (such as component short circuits, complete failure of the cooling system, etc.) are determined. Once such an extreme situation threatening hardware safety is detected, regardless of the current operating scenario, the storage device will be immediately (or after a brief debouncing) powered off, effectively preventing permanent damage to the storage device, catastrophic data loss, or even secondary safety accidents such as fires, greatly improving the operational reliability and physical security of the storage device.

[0229] Optionally, in order to better understand the abnormal monitoring process of the above-mentioned storage device, the abnormal monitoring process of the above-mentioned storage device will be described below in conjunction with optional embodiments, but it is not intended to limit the technical solution of the embodiments of this application.

[0230] This embodiment provides a method for monitoring anomalies in a storage device. Taking a PCIe SSD as an example, the PCIe SSD has a capacity of 7.68TB. The core components of the SSD are a main controller, NAND flash memory, and DRAM. The overall maximum average power consumption is 24W (7W for the main controller, 4W for DRAM, 11W for NAND flash memory, and 2W for other power losses or branch circuits), and the idle power consumption is 6W (4.5W for the main controller, 0.5W for DRAM, and 1W for NAND flash memory). The 128K sequential write performance is 5000MB / s, and the 4K random write performance is 200K IOPS. Figure 8 This is a flowchart of SSD anomaly monitoring according to an embodiment of this application, such as... Figure 8 As shown, the main steps include the following:

[0231] Step S701: Hardware initialization and data channel establishment. (See attached document) Figure 4 As shown, the SSD main controller (equivalent to the on-board controller) uses its IIC1 and IIC2 interfaces to establish communication connections with the multi-channel power consumption detection module circuit, the multi-channel temperature monitoring module circuit, and the backup power management module circuit, respectively, and performs initialization configuration for these module circuits. Simultaneously, it establishes a health information reporting channel with the BMC (equivalent to the off-board controller) via the out-of-band interface (SMBUS bus).

[0232] Step S702: Collect health parameters. The main controller periodically polls the storage device (SSD) in real time to obtain all power consumption (first power consumption parameter) and temperature (first temperature parameter) data, including the total disk power consumption and the power consumption of multiple nodes (main controller, NAND, DRAM, etc.), as well as the temperature of multiple nodes (main controller, NAND, DRAM, single board environment), such as the ambient temperature of the SSD. Stable sequential write speed of 2GB / s at 50℃. Actual temperature measurement: It is 90℃. It is 65℃. Temperature: 80℃. Actual power consumption measurement: It is 17.8W. It is 6W. It is 1.8W. It is 8.5W.

[0233] Step S703: Determine the current operating scenario. The main controller (in-disk controller) determines the current operating mode and load type (current operating scenario) of the storage device based on workload characteristics such as IO request type, bandwidth (BW), and IOPS.

[0234] Step S704: Calculate the health baseline. Based on the current load type (current operating scenario), the main controller (in-disk controller) calls the pre-stored temperature / power consumption health baseline function to calculate the expected temperature / power consumption (second power consumption parameter and second temperature parameter) of each core component. The expected power consumption calculation method is consistent for each core component. Taking the expected power consumption and expected temperature calculation process of NAND as an example, the expected power consumption calculation formula is:

[0235] ;

[0236] in, This represents the idle power consumption of the NAND flash memory, which is 1W. The formula for calculating the bandwidth power consumption factor (power consumption per unit bandwidth) for sequential writing is as follows: ,Right now ,get It is 2.0W / (GB / s); The current bandwidth is 2GB / s; The power consumption factor (power consumption per IOPS for random read / write operations) is calculated using the following formula: ,Right now ,get It is 0.0000125W / IOPS; The current IOPS is 200K IOPS. Based on the above calculations, the expected power consumption of the entire SSD is obtained. 13.2, Expected power consumption of the main controller The expected power consumption of DRAM is 5.5. 5. Expected power consumption of NAND flash memory It is 1.9.

[0237] The formula for calculating the expected temperature is:

[0238] ;

[0239] in, The thermal resistivity of NAND was calculated as follows: the main controller's thermal resistivity is 6.0°C / W, the NAND's thermal resistivity is 2.5°C / W, and the DRAM's thermal resistivity is 5.5°C / W. Based on these calculations, the expected temperature of the main controller was determined. The expected temperature for DRAM is 83℃. The expected temperature for NAND flash memory is 60.45℃. The temperature is 62.5℃.

[0240] Step S705: Calculate the deviation parameters. The main controller (internal controller) calculates the power consumption health anomaly function (power consumption deviation parameter), the temperature health anomaly function (temperature deviation parameter), and the power consumption-temperature correlation anomaly function (correlation deviation parameter) based on the real-time measured values ​​and expected values.

[0241] The formula for calculating the power consumption health anomaly function is:

[0242] ;

[0243] in, , , and These are the power consumption weights (equivalent to power consumption weight coefficients) for the entire disk, main controller, DRAM, and NAND, used to reflect the importance of each core electronic component to overall power consumption anomalies. Generally speaking, in PCIe SSDs, DRAM power consumption stability is the most critical, followed by NAND; therefore, the weights can be set as follows: Take 0.1 (for the whole plate). Set the value to 0.3 (main controller). Take 0.4 (DRAM). We take 0.2 (NAND). Based on the above calculations, the device power consumption deviations for the entire disk, main controller, NAND, and DRAM are 0.35, 0.09, 0.70, and 0.05, respectively, with the final power consumption health abnormality being 0.342.

[0244] The formula for calculating the temperature health anomaly function is:

[0245] ;

[0246] in, , , The temperature weights are set for the main controller, DRAM, and NAND, respectively. Generally speaking, in PCIe SSDs, the temperature stability of the main controller is the most critical, followed by DRAM. The weights can be set as follows: Take 0.4 (main controller). Take 0.3 (DRAM). Take 0.2 (NAND). The system temperature difference weighting coefficient has a value of 0.1. The difference between the main controller temperature and the actual NAND temperature is 10°C. The difference between the expected temperature values ​​of the main controller and NAND under the current workload mode is 20.5. Based on the above calculations, the device temperature deviations of the main controller, DRAM, and NAND are 0.084, 0.28, and 0.075, respectively, the device temperature difference deviation between the main controller and NAND is 0.51, and the final temperature health anomaly is 0.1911.

[0247] The formula for calculating the power consumption versus temperature anomaly function is:

[0248] ;

[0249] in, The average of the actual temperature values ​​of the main controller and NAND. Total power consumption of storage devices within a specific time window With average temperature The linear correlation coefficient (or Pearson correlation coefficient) between two data sequences is primarily used to characterize the degree of linear correlation between total power consumption and average temperature, where:

[0250] ;

[0251] Based on the above calculations, we obtain Therefore ≈0.

[0252] Step S705: Summarize and calculate the overall health score. The main controller, based on the power consumption health anomaly function (power consumption deviation parameter), temperature health anomaly function (temperature deviation parameter), and power consumption-temperature correlation anomaly function (correlation deviation parameter), summarizes and calculates the SSD health comprehensive index function (device abnormal state), and analyzes abnormal points in temperature / power consumption health data. The formula for calculating the SSD health comprehensive index function is:

[0253] ;

[0254] in, , , These are weighting coefficients, in the initial stage. .

[0255] Step S706: Detect abnormal states and locate the cause of abnormalities (cause of abnormality). The main controller executes the fault location decision process based on the score of the SSD health comprehensive index function (device abnormal state) and whether the power consumption and temperature correlation abnormal function (correlation deviation parameter) exceeds the threshold. Figure 9 This is a flowchart of the fault location decision-making process according to an embodiment of this application, such as... Figure 9 As shown, the first step is to determine if the correlation is abnormal: if the correlation deviation parameter... If an anomaly occurs (exceeding the threshold), the cause will be identified as: power consumption or temperature sensor malfunction, SSD cooling system malfunction (SSD internal thermal pad malfunction or system fan malfunction), etc. Secondary checks will be made for power consumption and temperature anomalies: if the correlation parameters deviate... If normal, then further check the power consumption health anomaly function. (Power consumption deviation parameter) or temperature health anomaly function Whether the (temperature deviation parameter) is abnormal, and based on its main contributors (e.g., , The cause of the anomaly is identified, such as poor heat dissipation of the main controller, NAND aging, DRAM overheating, or abnormal efficiency of the main controller. In this embodiment, the SSD health comprehensive index is obtained based on the above calculations. The value was 0.1777 (< 0.5 threshold), and no critical alarm was triggered. Among them... =0.342 is the main source of anomalies (>0.3 threshold), and the device power consumption deviation of NAND at 0.70 is the main contributor; in addition, =0.1911, with the device temperature deviation of NAND at 0.28 being the main contributor. NAND is identified as the target device.

[0256] Step S707: Reporting of Abnormal Information and Implementation Decisions. The main controller transmits the located fault information (NAND flash memory degradation detected. Main characteristics: NAND write power consumption abnormally increased by about 70%, and NAND temperature increased by about 28%) to the host (BMC) via the out-of-band interface (SMBUS bus). The host then reports the fault and makes operational implementation decisions, such as replacing the corresponding SSD or performing early NAND repairs.

[0257] Optionally, in order to better understand the abnormal monitoring process of the above-mentioned storage device, the abnormal monitoring process of the above-mentioned storage device will be described below in conjunction with optional embodiments, but it is not intended to limit the technical solution of the embodiments of this application.

[0258] In the above PCIe SSD anomaly monitoring process, it is assumed that a comprehensive health score is calculated. The value is 0.44397, close to the 0.5 threshold, triggering a critical alarm. Among them... =1 is the largest contributor, clearly indicating a complete break in the material relationship between power consumption and temperature. Power consumption health anomaly function. =0.0276 Excludes electrical faults in components such as the main controller, NAND, and DRAM; temperature health anomaly function. =0.3044 (exceeding the 0.3 threshold), and the device temperature difference deviation between the main controller and NAND is 1.69, which is severely abnormal. Under high load, the main controller temperature being lower than the NAND temperature violates the laws of physics. The SSD main controller synchronously outputs system warnings and maintenance suggestions through the SMBUS out-of-band management interface: Critical warning; Since power consumption and temperature correlation parameters are unrelated and the main controller temperature is lower than the NAND temperature, it is determined that the main controller core temperature sensor or external temperature sensor is faulty. If the health scores obtained from power consumption and temperature information are similar for three consecutive times, it is recommended to replace the corresponding SSD and the main controller or the main controller's external temperature sensor for repair.

[0259] Optionally, in this embodiment, by using the aforementioned power consumption health anomaly function, temperature health anomaly function, power consumption and temperature correlation anomaly function, and health comprehensive index function, when the health comprehensive score calculated by the health comprehensive index function indicates that the SSD health status is abnormal, specific system faults can be estimated or predicted based on the proportions of the power consumption and temperature correlation anomaly function, the power consumption health anomaly function, and the temperature health anomaly function, respectively. This achieves technical effects such as improving the efficiency of fault root cause identification in storage device anomaly monitoring.

[0260] Optionally, in this embodiment, this application also provides a variety of optional, simpler and more direct anomaly detection and energy efficiency management implementation methods to meet the differentiated requirements of real-time performance, reliability and energy efficiency in different application scenarios.

[0261] Optionally, in this embodiment, an anomaly monitoring method based on multi-node temperature is also provided. Figure 10 This is a flowchart of an anomaly monitoring method based on multi-node temperature according to an embodiment of this application, such as... Figure 10 As shown, when an abnormal device status is detected to indicate that the storage device is not malfunctioning (i.e., it is determined that no hardware-level fault has occurred, and the abnormal temperature is only caused by high load), the performance of the storage device can be adjusted based on the collected temperature data to cool the storage device and prevent high-temperature damage. Its core decision-making process is based on a two-level judgment of the raw temperature reading and the composite temperature index. This method includes:

[0262] Level 1 Judgment (Absolute Threshold Protection): The storage device's on-board controller reads the measured temperature values ​​from four temperature monitoring modules (Sensor1~Sensor4, corresponding to the main controller, NAND, DRAM, and the board, respectively) in real time via the IIC bus. If the measured temperature value of any sensor exceeds its preset absolute maximum safety threshold (for example, the maximum operating temperature thresholds for the main controller, NAND, DRAM, and board can be: 105℃ 85℃ 95℃ If the temperature reaches 85℃ and the abnormal state persists for more than 5 minutes, it is considered an emergency overheating event. At this time, the ALERT pin of the temperature monitoring module will directly output a low-level signal to the enable (EN) pin of the backup power management module through hardware circuitry (such as resistor R7), triggering automatic hardware-level power-down to achieve mandatory protection of the device and data.

[0263] Secondary judgment (composite temperature control): If the measured temperature values ​​of all sensors do not exceed the absolute threshold, the controller calculates the composite temperature index according to the following formula:

[0264] ;

[0265] in, A preset temperature threshold, such as 77℃, is set. Based on the range of this composite temperature value, the anomaly monitoring system executes the following gradient control strategy:

[0266] < The storage device is in good condition, with its performance and NAND bandwidth operating at 100% full speed, and the BMC fan running at the base speed.

[0267] ≤ ≤ The storage device has entered an initial warning state, with storage device-side services reducing load and performance and NAND bandwidth adjusted to 75% operation. Simultaneously, the BMC increases fan speed to 1.25 times the baseline to enhance heat dissipation. greater than The value, for example, 80℃.

[0268] < ≤ Storage devices have entered a moderate warning state, with further load reduction on the storage devices themselves, and performance and NAND bandwidth adjusted to 50% operation; BMC fan speed has been increased to 1.5 times the baseline. greater than The value, for example, 85℃.

[0269] < ≤ Storage devices have entered a high alert state, with significant load reduction on the storage device side, and performance and NAND bandwidth adjusted to 10% operation; BMC fan speed remains at 1.5 times the baseline. greater than The value, for example, 90℃.

[0270] > If the overheating persists for more than 5 minutes, the storage device is determined to be severely overheated. The internal controller outputs a low level through GPIO1, which controls the backup power management module to trigger a firmware-level power-down.

[0271] Optionally, in this embodiment, an anomaly monitoring method based on multi-node power consumption is also provided. Figure 11 This is a flowchart of an anomaly monitoring method based on multi-node power consumption according to an embodiment of this application, such as... Figure 11 As shown, when an abnormal device status is detected to indicate that the storage device is not malfunctioning, i.e., it is determined that no hardware-level fault has occurred and the temperature abnormality is only caused by high load, the performance of the storage device can be adjusted based on the collected power consumption to achieve the purpose of cooling the storage device and preventing high-temperature damage. This method includes:

[0272] Branch power consumption protection (priority judgment): The on-board controller monitors the branch power consumption of each core component (main controller power consumption). NAND power consumption DRAM power consumption If the power consumption of any branch exceeds 1.25 times its rated maximum power consumption (i.e., 125%), and the abnormal state lasts for more than 5 minutes, it is considered a serious electrical abnormality (such as a short circuit). At this time, the ALERT pin of the power consumption detection module will directly output a low-level signal to the backup power management module through hardware circuitry (such as resistor R8), triggering hardware-level automatic power-down.

[0273] Overall power consumption control: If the power consumption of all branches is normal, then based on the overall power consumption (overall power consumption)... To regulate:

[0274] The overall power consumption is within the rated range, the storage device performance and NAND bandwidth are operating at 100%, and the BMC fan is operating at the base speed. This represents the average maximum power consumption of the entire disk.

[0275] The overall power consumption was slightly exceeded, so the storage device's workload was reduced, and the performance and NAND bandwidth were adjusted to run at 75%; the BMC increased the fan speed to 1.25 times the baseline.

[0276] The overall power consumption of the disk is significantly exceeded, and the storage device is further deloaded, with performance and NAND bandwidth adjusted to 50% operation; the BMC fan speed is increased to 1.5 times the baseline.

[0277] And this continued for more than 5 minutes: the power consumption of the entire disk was seriously exceeded, and the controller inside the disk triggered a firmware-level power-down by controlling the backup power management module through GPIO1.

[0278] Optionally, in this embodiment, a multi-level power consumption level control method is also provided. Figure 12 This is a flowchart of a multi-level power consumption level control method according to an embodiment of this application, such as... Figure 12 As shown, this method aims to allow users to flexibly balance performance and energy efficiency to adapt to different business scenarios and customer needs. The method includes:

[0279] Power level definition: The system predefines multiple power and performance levels with corresponding relationships. For example, for a PCIe 5.0 SSD, six levels can be set: S0 (6W) - Idle, S1 (9W) - First performance, S2 (12W) - Second performance, S3 (15W) - Third performance, S4 (18W) - Fourth performance, and S5 (22W) - Maximum performance. Each level corresponds to a specific performance level (such as sequential read / write bandwidth).

[0280] Power consumption acquisition: The on-board controller acquires the power consumption of the entire board and core branches in real time through the IIC bus.

[0281] Power level selection and adjustment: Users or upper-level management systems select the target power level Sx based on the current business load and energy efficiency requirements.

[0282] Closed-loop regulation: The SSD main controller dynamically adjusts the working state of internal modules (such as main controller frequency, NAND channel concurrency, DRAM power management, etc.) to reduce the actual power consumption of the entire drive. Approaching the target power consumption Sx. The goal of the adjustment is to meet the following conditions: Until the power consumption stabilizes near the target power consumption level Sx.

[0283] Through the various optional implementation methods provided in this embodiment, a multi-dimensional collaborative overall technical effect can be achieved: by constructing a multi-layered, multi-granular protection system ranging from "intelligent early warning based on complex models" to "rapid protection based on absolute thresholds," the system meets the requirements for accuracy and real-time performance in anomaly monitoring. At the same time, it utilizes a hardware circuit to directly trigger a power-down mechanism to achieve a response to severe anomalies within milliseconds, maximizing hardware security and data integrity. Furthermore, through multi-level power consumption management, power consumption control is upgraded from a passive, fixed mode to an active, configurable strategy, significantly improving the product's adaptability and competitiveness in different application scenarios such as cloud computing and edge computing, and comprehensively optimizing the reliability, security, and scenario adaptability of storage device health management.

[0284] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0285] Based on this understanding, the technical solution of this application, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0286] This embodiment also provides an anomaly monitoring device for a storage device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0287] Figure 13 This is a structural block diagram of an anomaly monitoring device for a storage device according to an embodiment of this application; as shown below. Figure 13 As shown, it includes:

[0288] The acquisition module 1302 is used to acquire a first power consumption parameter, a first temperature parameter, and a first correlation parameter of the storage device, and to detect a second power consumption parameter and a second temperature parameter of the storage device in the current operating scenario. The first power consumption parameter is used to indicate the first device power consumption of the storage device, the first temperature parameter is used to indicate the first device temperature of the storage device, the first correlation parameter is used to indicate the first correlation between the first power consumption parameter and the first temperature parameter, the second power consumption parameter is used to indicate the second device power consumption allowed to be generated by the storage device in the current operating scenario, and the second temperature parameter is used to indicate the second device temperature allowed to be generated by the storage device in the current operating scenario.

[0289] The determining module 1304 is used to determine a power consumption deviation parameter based on a first power consumption parameter and a second power consumption parameter, determine a temperature deviation parameter based on a first temperature parameter and a second temperature parameter, and convert a first correlation parameter into a correlation deviation parameter. The power consumption deviation parameter is used to indicate the degree of deviation of the power consumption of the first device relative to the power consumption of the second device, the temperature deviation parameter is used to indicate the degree of deviation of the temperature of the first device relative to the temperature of the second device, the correlation deviation parameter is used to indicate the degree of deviation of the first correlation relative to the second correlation, and the second correlation is used to indicate the degree of correlation between the power consumption and temperature of the storage device.

[0290] The first detection module 1306 is used to detect the abnormal state of the storage device based on power consumption deviation parameters, temperature deviation parameters and correlation deviation parameters, and when the abnormal state of the device is detected to indicate that there is an abnormality in the storage device, it is used to detect the cause of the abnormality of the storage device based on power consumption deviation parameters, temperature deviation parameters and correlation deviation parameters.

[0291] In one exemplary embodiment, the first detection module includes:

[0292] The first detection unit is used to detect abnormal correlation status of the storage device based on the correlation deviation parameter.

[0293] The first determining unit is used to determine the cause of the abnormality when a correlation abnormality state is detected to indicate that the storage device has an abnormality in the correlation between power consumption and temperature. This abnormality includes temperature sensor abnormality, power consumption sensor abnormality, and heat dissipation system abnormality.

[0294] The second detection unit is used to detect the abnormal power consumption state of the storage device based on the power consumption deviation parameter and the abnormal temperature state of the storage device based on the temperature deviation parameter when the abnormal correlation state is detected to indicate that there is no abnormality in the correlation between power consumption and temperature of the storage device.

[0295] The positioning unit is configured to, when a power consumption abnormality is detected, indicating that the storage device has an abnormal power consumption, locate a first target device causing the power consumption abnormality from the device components included in the storage device as the cause of the abnormality, based on the device power consumption of the device components included in the storage device; and when a temperature abnormality is detected, indicating that the storage device has an abnormal temperature, locate a second target device causing the temperature abnormality from the device components included in the storage device as the cause of the abnormality, based on the device temperature and / or device temperature relationship of the device components included in the storage device.

[0296] In one exemplary embodiment, the second detection unit is further configured to:

[0297] Obtain the device power consumption deviation of each device component. The device power consumption deviation is used to indicate the degree of deviation of the first device power consumption of the corresponding device component relative to the second device power consumption in the current operating scenario. The device power consumption deviation is the quotient between the device power consumption difference and the second device power consumption. The device power consumption difference is the absolute value of the difference between the first device power consumption and the second device power consumption.

[0298] The power consumption deviation of each device is compared with the deviation parameter threshold;

[0299] Devices whose power consumption deviation is greater than or equal to the deviation parameter threshold are identified as the first target devices, and the cause of the anomaly is obtained.

[0300] In one exemplary embodiment, the acquisition module includes:

[0301] The acquisition unit is used to acquire the current power consumption of the storage device and the first device power consumption of each device in the storage device to obtain the first power consumption parameter, and to acquire the first device temperature of each device in the storage device.

[0302] The first calculation unit is used to calculate the temperature difference parameter and the average temperature parameter between the first device temperature of the first device and the first device temperature of the second device. The first temperature parameter includes the first device temperature, temperature difference parameter and average temperature parameter of each device. The device temperature of the first device and the device temperature of the second device have a greater influence on the device temperature of the storage device than or equal to that of other devices included in the storage device.

[0303] The second calculation unit is used to calculate the correlation between the current power consumption and the average temperature parameter to obtain the first correlation parameter.

[0304] In one exemplary embodiment, the second computing unit is further configured to:

[0305] The first correlation parameter is calculated using the following formula:

[0306] ;

[0307] in, Current power consumption, The average temperature parameter, This represents the covariance between the current power consumption and the average temperature parameter. The variance of the current power consumption. This represents the variance of the average temperature parameter.

[0308] In one exemplary embodiment, the acquisition module further includes:

[0309] The first receiving unit is used to receive the current power consumption reported by the first power consumption detection module, receive the first device power consumption reported by the second power consumption detection module corresponding to each device, and obtain the first power consumption parameter. The first power consumption detection module is used to detect the power consumption generated by the storage device, and the second power consumption detection module is used to detect the power consumption generated by the corresponding device.

[0310] The second receiving unit is used to receive the temperature of the first device reported by the first temperature sensor corresponding to each device, wherein the first temperature sensor is used to collect the temperature of the corresponding device.

[0311] In one exemplary embodiment, the acquisition module further includes:

[0312] The first acquisition unit is used to acquire the idle power consumption, device bandwidth, power consumption per unit bandwidth, device read / write parameters and power consumption per unit of operation for each device in the storage device, as well as the ambient temperature of the storage device and the thermal resistance coefficient of each device. The device read / write parameters are the number of read / write operations per second, the power consumption per unit bandwidth is the power consumption generated per unit bandwidth, and the power consumption per unit of operation is the power consumption generated per second by the device read / write parameters.

[0313] The third calculation unit is used to calculate the second power consumption parameters corresponding to each device based on idle power consumption, device bandwidth, power consumption per unit bandwidth, device read / write parameters, and power consumption per unit of number of reads / writes.

[0314] The fourth calculation unit is used to calculate the second temperature parameters corresponding to each device based on the ambient temperature, thermal resistance coefficient, and second power consumption parameters.

[0315] In one exemplary embodiment, the third computing unit is further configured to:

[0316] The second power consumption parameter for each device component is calculated using the following formula:

[0317] ;

[0318] in, This is the idle power consumption. For device bandwidth, Power consumption per unit bandwidth For device read / write parameters, Power consumption is measured in units of frequency.

[0319] In one exemplary embodiment, the fourth computing unit is further configured to:

[0320] The second temperature parameter corresponding to each device component is calculated using the following formula:

[0321] ;

[0322] in, For ambient temperature, The thermal resistance coefficient, This is the second power consumption parameter.

[0323] In one exemplary embodiment, the first detection module includes:

[0324] The third detection unit is used to detect the operating deviation parameters of the storage device based on power consumption deviation parameters, temperature deviation parameters and correlation deviation parameters. The operating deviation parameters are used to indicate the degree of deviation of the operating state of the storage device from the healthy operating state.

[0325] The second determining unit is used to determine the abnormal state of the device to indicate that there is an abnormality in the storage device when the degree of deviation indicated by the operating deviation parameter is greater than or equal to the degree of deviation threshold.

[0326] The third determining unit is used to determine the abnormal state of the device when the degree of deviation indicated by the operating deviation parameter is less than the deviation degree threshold, so as to indicate that there is no abnormality in the storage device.

[0327] In an exemplary embodiment, the third detection unit is further configured to calculate the operating deviation parameter using the following formula:

[0328] ;

[0329] in, This is a power consumption deviation parameter. This is a temperature deviation parameter. This is the correlation deviation parameter. The first weighting coefficient for the power consumption deviation parameter is... This is the second weighting coefficient for the temperature deviation parameter. The third weighting coefficient is the correlation deviation parameter. The weighting coefficient is used to indicate the degree of influence of the corresponding deviation parameter on the operation of the storage device in a healthy operating state. The sum of the first weighting coefficient, the second weighting coefficient and the third weighting coefficient is 1.

[0330] In one exemplary embodiment, the determining module includes:

[0331] The second acquisition unit is used to acquire the device power consumption deviation of each device in the storage device and the power consumption weight coefficient of each device. The device power consumption deviation is used to indicate the degree of deviation of the first device power consumption of the corresponding device relative to the second device power consumption in the current operating scenario. The device power consumption deviation is the quotient between the device power consumption difference and the second device power consumption. The device power consumption difference is the absolute value of the difference between the first device power consumption and the second device power consumption. The power consumption weight coefficient is used to indicate the degree of impact of the device power consumption deviation of the corresponding device on the operation of the storage device in a healthy operating state.

[0332] The fourth determining unit is used to determine the sum of the abnormal power consumption parameters of each device as the power consumption deviation parameter, wherein the abnormal power consumption parameter is the product of the device power consumption deviation of the corresponding device and the power consumption weighting coefficient.

[0333] In one exemplary embodiment, the determining module further includes:

[0334] The third acquisition unit is used to acquire the device temperature deviation and temperature weighting coefficient of each device in the storage device, and to acquire the temperature difference abnormal parameter and temperature difference weighting coefficient between the first device and the second device. The device temperature deviation is used to indicate the degree of deviation of the first device temperature of the corresponding device relative to the second device temperature in the current operating scenario. The device temperature deviation is the quotient between the device temperature difference and the second device temperature. The device temperature difference is the absolute value of the difference between the first device temperature and the second device temperature. The temperature weighting coefficient is used to indicate the degree of influence of the device temperature deviation of the corresponding device on the operation of the storage device in a healthy operating state. The influence of the device temperature of the first device and the device temperature of the second device on the device temperature of the storage device is greater than or equal to that of other devices included in the storage device. The temperature difference abnormal parameter is used to indicate the degree of deviation of the first temperature difference parameter between the first device and the second device relative to the second temperature difference parameter in the current operating scenario. The temperature difference weighting coefficient is used to indicate the degree of influence of the temperature difference abnormal parameter on the operation of the storage device in a healthy operating state.

[0335] The fifth determining unit is used to determine the sum of the device temperature abnormality parameter and the device temperature difference abnormality parameter of each device as the temperature deviation parameter. The device temperature abnormality parameter is the product of the device temperature deviation of the corresponding device and the temperature weighting coefficient, and the device temperature difference abnormality parameter is the product of the temperature difference abnormality parameter and the temperature difference weighting coefficient.

[0336] In one exemplary embodiment, the apparatus further includes:

[0337] The receiving module is used to receive the power consumption configuration of the storage device after collecting the first power consumption parameters of the storage device, wherein the power consumption configuration is used to configure the power consumption level of the storage device to the desired power consumption level.

[0338] The second detection module is used to detect the target power level corresponding to the current power level in the first power parameter from the power consumption and power level with corresponding relationship, wherein the current power consumption is the power consumption of the collected storage device.

[0339] The adjustment module is used to adjust the operating parameters of the storage device from the target operating parameters that meet the target power consumption level to the reference operating parameters that meet the desired power consumption level.

[0340] In one exemplary embodiment, the adjustment module includes:

[0341] The comparison unit is used to compare the target power consumption level with the desired power consumption level, where a higher power consumption level indicates higher power consumption of the storage device;

[0342] The sixth determining unit is used to determine the first operating parameters corresponding to the storage service being run by the storage device when the target power consumption level is higher than the expected power consumption level; and to reduce the first operating parameters until the power consumption of the storage device is collected and falls into the expected power consumption level.

[0343] The seventh determining unit is used to determine the second operating parameters corresponding to the storage services run by the storage device when the target power consumption level is lower than the expected power consumption level; and to increase the second operating parameters until the power consumption of the storage device is collected and falls into the expected power consumption level.

[0344] In one exemplary embodiment, the apparatus further includes:

[0345] The comparison module is used to compare the power consumption of each device included in the first power consumption parameter with the corresponding power consumption threshold after collecting the first power consumption parameter and the first temperature parameter of the storage device; compare the temperature of each device included in the first temperature parameter with the corresponding first temperature threshold; and compare the collected board temperature of the storage device with the second temperature threshold.

[0346] The control module is used to control the storage device to power off when the power consumption of at least one device in the first device power consumption is greater than or equal to the corresponding power consumption threshold, the temperature of at least one device in the first device temperature is greater than or equal to the corresponding first temperature threshold, and / or the board temperature is greater than or equal to the second temperature threshold.

[0347] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0348] For a description of the features in the embodiment corresponding to the anomaly monitoring device of the storage device, please refer to the relevant description of the embodiment corresponding to the anomaly monitoring method of the storage device, which will not be repeated here.

[0349] Embodiments of this application also provide an electronic device. Figure 14 This is a schematic diagram of an electronic device according to an embodiment of this application, such as... Figure 14 As shown, the electronic device includes a memory and a processor, the memory storing a computer program, and the processor being configured to run the computer program to perform the steps in any of the above-described embodiments of the abnormal monitoring method for the storage device.

[0350] In one exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0351] Specific examples in this embodiment can be found in the examples described in the above embodiments and exemplary implementations, and will not be repeated here.

[0352] Embodiments of this application also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the steps in any of the above-described embodiments of the abnormal monitoring method for storage devices when running.

[0353] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0354] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the methods described in various embodiments of this application; the computer program product further includes a non-volatile computer-readable storage medium storing the computer program, which, when executed by a processor, implements the steps of the abnormal monitoring method for the storage device described in various embodiments of this application.

[0355] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0356] The above provides a detailed description of an anomaly monitoring method for a storage device provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only intended to help understand the method and core ideas of this application. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this application.

Claims

1. A method for monitoring anomalies in a storage device, characterized in that, include: The system collects a first power consumption parameter of the storage device, collects a first temperature parameter of the storage device, and determines a first correlation parameter of the storage device. It also detects a second power consumption parameter and a second temperature parameter of the storage device under the current operating scenario. The first power consumption parameter indicates the first device power consumption of the storage device, the first temperature parameter indicates the first device temperature of the storage device, the first correlation parameter indicates the first correlation between the first power consumption parameter and the first temperature parameter, the second power consumption parameter indicates the second device power consumption allowed by the storage device under the current operating scenario, and the second temperature parameter indicates the second device temperature allowed by the storage device under the current operating scenario. A power consumption deviation parameter is determined based on the first power consumption parameter and the second power consumption parameter, a temperature deviation parameter is determined based on the first temperature parameter and the second temperature parameter, and the first correlation parameter is converted into a correlation deviation parameter. The power consumption deviation parameter is used to indicate the degree of deviation of the power consumption of the first device relative to the power consumption of the second device, the temperature deviation parameter is used to indicate the degree of deviation of the temperature of the first device relative to the temperature of the second device, the correlation deviation parameter is used to indicate the degree of deviation of the first correlation relative to the second correlation, and the second correlation is used to indicate the degree of correlation between the power consumption and temperature of the storage device. The abnormal state of the storage device is detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter. If the abnormal state of the device is detected to indicate that there is an abnormality in the storage device, the cause of the abnormality of the storage device is detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter. The step of detecting the cause of the storage device's anomaly based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter includes: The correlation anomaly state of the storage device is detected based on the correlation deviation parameter. If the abnormal correlation state is detected to indicate that the storage device is abnormal in the correlation between power consumption and temperature, the cause of the abnormality is determined to include temperature sensor abnormality, power consumption sensor abnormality, and heat dissipation system abnormality. If the abnormal correlation state is detected to indicate that the storage device is not abnormal in the correlation between power consumption and temperature, the abnormal power consumption state of the storage device is detected according to the power consumption deviation parameter, and the abnormal temperature state of the storage device is detected according to the temperature deviation parameter. When the abnormal power consumption state is detected to indicate that the storage device has an abnormal power consumption, a first target device causing the abnormal power consumption state is located from the device components according to the device power consumption of the device components included in the storage device as the cause of the abnormality; when the abnormal temperature state is detected to indicate that the storage device has an abnormal temperature, a second target device causing the abnormal temperature state is located from the device components according to the device temperature and / or device temperature relationship of the device components included in the storage device as the cause of the abnormality.

2. The method according to claim 1, characterized in that, The step of locating the first target device causing the abnormal power consumption state from the device devices included in the storage device as the cause of the abnormality includes: Obtain the device power consumption deviation of each of the aforementioned devices, wherein the device power consumption deviation is used to indicate the degree of deviation of the first device power consumption of the corresponding device relative to the second device power consumption in the current operating scenario, the device power consumption deviation is the quotient between the device power consumption difference and the second device power consumption, and the device power consumption difference is the absolute value of the difference between the first device power consumption and the second device power consumption; The power consumption deviation of each device is compared with the deviation parameter threshold; Devices whose power consumption deviation is greater than or equal to the deviation parameter threshold are identified as the first target device, and the cause of the anomaly is obtained.

3. The method according to claim 1, characterized in that, The acquisition of the first power consumption parameter of the storage device, the acquisition of the first temperature parameter of the storage device, and the determination of the first correlation parameter of the storage device include: The first power consumption parameter is obtained by collecting the current power consumption of the storage device and the first device power consumption of each device in the storage device, and the first device temperature of each device in the storage device is also collected. Calculate the temperature difference parameter and the average temperature parameter between the first device temperature of the first device and the first device temperature of the second device, wherein the first temperature parameter includes the first device temperature of each device, the temperature difference parameter and the average temperature parameter, and the influence of the device temperature of the first device and the device temperature of the second device on the device temperature of the storage device is greater than or equal to that of other devices included in the storage device. The first correlation parameter is obtained by calculating the correlation between the current power consumption and the average temperature parameter.

4. The method according to claim 3, characterized in that, The calculation of the correlation between the current power consumption and the average temperature parameter to obtain the first correlation parameter includes: The first correlation parameter is calculated using the following formula: ； in, The current power consumption, The average temperature parameter is... This represents the covariance between the current power consumption and the average temperature parameter. The variance of the current power consumption. The variance of the average temperature parameter is denoted as .

5. The method according to claim 3, characterized in that, The first power consumption parameter is obtained by collecting the current power consumption of the storage device and the first device power consumption of each device in the storage device, and the first device temperature of each device in the storage device is collected, including: The system receives the current power consumption reported by the first power consumption detection module, receives the power consumption of the first device reported by the second power consumption detection module corresponding to each device, and obtains the first power consumption parameter. The first power consumption detection module is used to detect the power consumption generated by the storage device, and the second power consumption detection module is used to detect the power consumption generated by the corresponding device. The device receives the temperature of the first device reported by the first temperature sensor corresponding to each of the aforementioned devices, wherein the first temperature sensor is used to collect the temperature of the corresponding device.

6. The method according to claim 1, characterized in that, The detection of the second power consumption parameter and the second temperature parameter of the storage device under the current operating scenario includes: The idle power consumption, device bandwidth, power consumption per unit bandwidth, device read / write parameters, and power consumption per unit of operation are obtained for each device in the storage device, as well as the ambient temperature of the storage device and the thermal resistance coefficient of each device. The device read / write parameters are the number of read / write operations per second, the power consumption per unit bandwidth is the power consumption generated per unit bandwidth, and the power consumption per unit of operation is the power consumption generated per second by the device read / write parameters. The second power consumption parameter corresponding to each device is calculated based on the idle power consumption, the device bandwidth, the power consumption per unit bandwidth, the device read / write parameters, and the power consumption per unit of the number of times. The second temperature parameter corresponding to each device is calculated based on the ambient temperature, the thermal resistance coefficient, and the second power consumption parameter.

7. The method according to claim 6, characterized in that, The step of calculating the second power consumption parameter corresponding to each device based on the idle power consumption, the device bandwidth, the power consumption per unit bandwidth, the device read / write parameters, and the power consumption per unit of the number of operations includes: The second power consumption parameter corresponding to each device is calculated using the following formula: ; in, The idle power consumption is... The bandwidth of the device. The power consumption per unit bandwidth. The read / write parameters for the device are... The power consumption is the unit of the number of times.

8. The method according to claim 6, characterized in that, The step of calculating the second temperature parameter corresponding to each device based on the ambient temperature, the thermal resistance coefficient, and the second power consumption parameter includes: The second temperature parameter corresponding to each device component is calculated using the following formula: ; in, The ambient temperature is... The thermal resistance coefficient is... This refers to the second power consumption parameter.

9. The method according to claim 1, characterized in that, The step of detecting the abnormal state of the storage device based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter includes: The operating deviation parameter of the storage device is detected based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter, wherein the operating deviation parameter is used to indicate the degree of deviation of the operating state of the storage device from the healthy operating state; If the degree of deviation indicated by the operational deviation parameter is greater than or equal to the deviation degree threshold, the abnormal state of the device is determined to indicate that there is an abnormality in the storage device; If the degree of deviation indicated by the operating deviation parameter is less than the deviation degree threshold, the abnormal state of the device is determined to indicate that the storage device is not abnormal.

10. The method according to claim 9, characterized in that, The step of detecting the operating deviation parameters of the storage device based on the power consumption deviation parameter, the temperature deviation parameter, and the correlation deviation parameter includes: The operating deviation parameter is calculated using the following formula: ； in, The power consumption deviation parameter is... The temperature deviation parameter is... The correlation deviation parameter is... The first weighting coefficient for the power consumption deviation parameter is... This is the second weighting coefficient for the temperature deviation parameter. The third weighting coefficient is the correlation deviation parameter. The weighting coefficient is used to indicate the degree of influence of the corresponding deviation parameter on the operation of the storage device in the healthy operating state. The sum of the first weighting coefficient, the second weighting coefficient and the third weighting coefficient is 1.

11. The method according to claim 1, characterized in that, The step of determining the power consumption deviation parameter based on the first power consumption parameter and the second power consumption parameter includes: The device power consumption deviation and power consumption weighting coefficient of each device in the storage device are obtained. The device power consumption deviation is used to indicate the degree of deviation of the first device power consumption of the corresponding device relative to the second device power consumption in the current operating scenario. The device power consumption deviation is the quotient between the device power consumption difference and the second device power consumption. The device power consumption difference is the absolute value of the difference between the first device power consumption and the second device power consumption. The power consumption weighting coefficient is used to indicate the degree of influence of the device power consumption deviation of the corresponding device on the operation of the storage device in a healthy operating state. The first power consumption parameter includes the first device power consumption of each device. The second power consumption parameter includes the second device power consumption of each device in the current operating scenario. The sum of the abnormal power consumption parameters of each of the aforementioned devices is determined as the power consumption deviation parameter, wherein the abnormal power consumption parameter is the product of the power consumption deviation of the corresponding device and the power consumption weighting coefficient.

12. The method according to claim 1, characterized in that, The step of determining the temperature deviation parameter based on the first temperature parameter and the second temperature parameter includes: The device temperature deviation and temperature weighting coefficient of each device in the storage device are obtained, as well as the temperature difference abnormality parameter and temperature difference weighting coefficient between the first device and the second device. The device temperature deviation indicates the degree of deviation of the first device temperature relative to the second device temperature in the current operating scenario. The device temperature deviation is the quotient between the device temperature difference and the second device temperature, and the device temperature difference is the absolute value of the difference between the first and second device temperatures. The temperature weighting coefficient indicates the degree of impact of the device temperature deviation of the corresponding device on the operation of the storage device in a healthy operating state. The device temperature deviation of the first device and the temperature weighting coefficient are also obtained. The influence of the device temperature of the second device on the device temperature of the storage device is greater than or equal to that of other devices included in the storage device. The temperature difference abnormality parameter is used to indicate the degree of deviation of the first temperature difference parameter between the first device and the second device relative to the second temperature difference parameter under the current operating scenario. The temperature difference weighting coefficient is used to indicate the degree of influence of the temperature difference abnormality parameter on the operation of the storage device under the healthy operating state. The first temperature parameter includes the first device temperature and the first temperature difference parameter of each device. The second temperature parameter includes the second device temperature and the second temperature difference parameter of each device under the current operating scenario. The sum of the device temperature anomaly parameter and the device temperature difference anomaly parameter of each of the aforementioned devices is determined as the temperature deviation parameter, wherein the device temperature anomaly parameter is the product of the device temperature deviation of the corresponding device and the temperature weighting coefficient, and the device temperature difference anomaly parameter is the product of the temperature difference anomaly parameter and the temperature difference weighting coefficient.

13. The method according to claim 1, characterized in that, After acquiring the first power consumption parameter of the storage device, the method further includes: Receive the power consumption configuration of the storage device, wherein the power consumption configuration is used to configure the power consumption level of the storage device to the desired power consumption level; The target power level corresponding to the current power consumption in the first power consumption parameter is detected from the corresponding power consumption and power consumption level, wherein the current power consumption is the power consumption of the storage device collected. The operating parameters of the storage device are adjusted from the target operating parameters that meet the target power consumption level to the reference operating parameters that meet the desired power consumption level.

14. The method according to claim 13, characterized in that, The step of adjusting the operating parameters of the storage device from the target operating parameters that meet the target power consumption level to the reference operating parameters that meet the desired power consumption level includes: Compare the target power consumption level with the desired power consumption level, wherein a higher power consumption level indicates a higher power consumption of the storage device; If the target power consumption level is higher than the desired power consumption level, determine the first operating parameter corresponding to the storage service being run by the storage device; reduce the first operating parameter until the power consumption of the storage device is collected and falls into the desired power consumption level; If the target power consumption level is lower than the desired power consumption level, determine the second operating parameter corresponding to the storage service being run by the storage device; increase the second operating parameter until the power consumption of the storage device is collected and falls into the desired power consumption level.

15. The method according to claim 1, characterized in that, After acquiring the first power consumption parameter and the first temperature parameter of the storage device, the method further includes: The power consumption of each device included in the first power consumption parameter is compared with the corresponding power consumption threshold, the temperature of each device included in the first temperature parameter is compared with the corresponding first temperature threshold, and the board temperature of the storage device's board is compared with the second temperature threshold. If the power consumption of at least one of the first devices is greater than or equal to the corresponding power consumption threshold, the temperature of at least one of the first devices is greater than or equal to the corresponding first temperature threshold, and / or the board temperature is greater than or equal to the second temperature threshold, the storage device is powered off.

16. An anomaly monitoring device for a storage device, characterized in that, include: The acquisition module is used to acquire a first power consumption parameter of the storage device, acquire a first temperature parameter of the storage device, and determine a first correlation parameter of the storage device, and detect a second power consumption parameter and a second temperature parameter of the storage device in the current operating scenario. The first power consumption parameter indicates the first device power consumption of the storage device, the first temperature parameter indicates the first device temperature of the storage device, the first correlation parameter indicates the first correlation between the first power consumption parameter and the first temperature parameter, the second power consumption parameter indicates the second device power consumption allowed by the storage device in the current operating scenario, and the second temperature parameter indicates the second device temperature allowed by the storage device in the current operating scenario. The determination module is configured to determine a power consumption deviation parameter based on the first power consumption parameter and the second power consumption parameter, determine a temperature deviation parameter based on the first temperature parameter and the second temperature parameter, and convert the first correlation parameter into a correlation deviation parameter. The power consumption deviation parameter indicates the degree of deviation between the power consumption of the first device and the power consumption of the second device, the temperature deviation parameter indicates the degree of deviation between the temperature of the first device and the temperature of the second device, and the correlation deviation parameter indicates the degree of deviation between the first correlation and the second correlation. The second correlation indicates the degree of correlation between the power consumption and temperature of the storage device. The first detection module is used to detect the abnormal state of the storage device based on the power consumption deviation parameter, the temperature deviation parameter and the correlation deviation parameter, and when the abnormal state of the device is detected to indicate that the storage device is abnormal, the module is used to detect the cause of the abnormality of the storage device based on the power consumption deviation parameter, the temperature deviation parameter and the correlation deviation parameter. The first detection module includes: The first detection unit is used to detect the abnormal correlation state of the storage device based on the correlation deviation parameter. The second detection unit is used to determine the cause of the abnormality when the abnormal correlation state is detected as indicating that the storage device has an abnormality in the correlation between power consumption and temperature. This abnormality includes temperature sensor abnormality, power consumption sensor abnormality, and heat dissipation system abnormality. The positioning unit is used to detect the abnormal power consumption state of the storage device based on the power consumption deviation parameter and the abnormal temperature state of the storage device based on the temperature deviation parameter when the abnormal correlation state is detected as indicating that the storage device has no abnormality in the correlation between power consumption and temperature. The first detection unit is configured to, when detecting the abnormal power consumption state indicating an abnormality in the power consumption of the storage device, locate a first target device causing the abnormal power consumption state from the device components included in the storage device as the cause of the abnormality, based on the device power consumption of the device components included in the storage device; and when detecting the abnormal temperature state indicating an abnormality in the temperature of the storage device, locate a second target device causing the abnormal temperature state from the device components included in the storage device as the cause of the abnormality, based on the device temperature and / or device temperature relationship of the device components included in the storage device.

17. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the anomaly monitoring method for a storage device as described in any one of claims 1 to 15 when executing the computer program.

18. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the abnormal monitoring method for the storage device as described in any one of claims 1 to 15.

19. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the abnormal monitoring method for the storage device as described in any one of claims 1 to 15.

Images