A fault prediction method, device, equipment and readable storage medium

Abstract

The present application discloses a fault prediction method, device, equipment and readable storage medium. By obtaining the historical parameter data of the first period and the historical parameter data of the second period before the to-be-predicted time node of the server, where the duration of the first period is greater than the duration of the second period, based on the historical parameter data of the first period, the reference parameter value range of the server at the to-be-predicted time node is predicted, and based on the historical parameter data of the second period, the predicted parameter data of the server at the to-be-predicted time node is predicted. It is judged whether the predicted parameter data is within the reference parameter value range. If not, it is determined that there is a potential fault at the to-be-predicted time node of the server. In the present application, by judging whether the predicted parameter data is within the reference parameter value range, it can be determined whether there is a potential fault at the to-be-predicted time node of the server, thereby realizing the prediction of potential faults.

Description

Technical Field

[0001] The present application relates to the field of operation and maintenance technology, and in particular to a fault prediction method, apparatus, device, and readable storage medium. Background Art

[0002] With the rapid development of internet technology, the number of internet users is increasing, business scenarios are becoming more diverse, and the number of servers adapted to each scenario is also increasing. To prevent server failures in each scenario, corresponding thresholds are set to monitor server data transmission volume. However, fault monitoring based on set thresholds only displays faults and issues alerts when the preset threshold is reached. At this point, the fault has already occurred, and the time it takes to resolve it is difficult to predict, which can seriously affect people's use of related service functions. Therefore, predicting faults has always been a concern. Summary of the Invention

[0003] In view of this, the present application provides a fault prediction method, apparatus, device and readable storage medium to facilitate fault prediction.

[0004] In order to achieve the above objectives, the following solutions are proposed:

[0005] A fault prediction method, comprising:

[0006] Obtain historical parameter data of the server in a first period and a second period before the time node to be predicted, wherein the duration of the first period is greater than the duration of the second period;

[0007] Based on the historical parameter data of the first period, predicting a reference parameter value range of the server at the time node to be predicted;

[0008] Based on the historical parameter data of the second period, predict and obtain predicted parameter data of the server at the time node to be predicted;

[0009] Determining whether the predicted parameter data is within the reference parameter value range;

[0010] If not, it is determined that the server has a potential failure at the predicted time node.

[0011] Optionally, also include:

[0012] The predicted parameter data and the time node to be predicted are input into the fault prediction model to determine whether the server has a potential fault at the time node to be predicted. The fault prediction model uses the parameter data of the server at the historical time node as a training sample and whether the server has a fault at the historical time node as a sample label for training.

[0013] Optionally, based on the historical parameter data of the first period, predicting the reference parameter value range of the server at the to-be-predicted time node includes:

[0014] Selecting a first target time node at the same moment as the to-be-predicted time node within the first period;

[0015] Using the data within a preset time range before and after the first target time node to predict the reference parameter value range of the server at the predicted time node.

[0016] Optionally, based on the historical parameter data of the second period, predicting the predicted parameter data of the server at the to-be-predicted time node includes:

[0017] Selecting a second target time node at the same moment as the to-be-predicted time node within the second period;

[0018] Using the data within a preset time range before and after the second target time node to predict the predicted parameter data of the server at the predicted time node.

[0019] Optionally, it further includes:

[0020] Judging whether there is a corresponding handling method for the potential fault;

[0021] If so, handling the potential fault according to the handling method;

[0022] If not, sending an alarm message.

[0023] Optionally, after handling the potential fault according to the handling method, it further includes:

[0024] Judging whether the potential fault is eliminated;

[0025] If so, recording the handling data;

[0026] If not, sending an alarm message.

[0027] A fault prediction device includes:

[0028] A historical data acquisition unit, configured to acquire the historical parameter data of the first period and the historical parameter data of the second period of the server before the to-be-predicted time node, wherein the duration of the first period is greater than the duration of the second period;

[0029] A reference range acquisition unit, configured to predict the reference parameter value range of the server at the to-be-predicted time node based on the historical parameter data of the first period;

[0030] A prediction data acquisition unit, configured to predict the predicted parameter data of the server at the to-be-predicted time node based on the historical parameter data of the second period;

[0031] A fault hidden danger judgment unit, configured to judge whether the predicted parameter data is within the reference parameter value range, and if not, determine that there is a fault hidden danger of the server at the to-be-predicted time node.

[0032] Optionally, it further includes:

[0033] A fault prediction unit, configured to input the predicted parameter data and the to-be-predicted time node into a fault prediction model to determine whether there is a fault hidden danger of the server at the to-be-predicted time node, where the fault prediction model is trained with the parameter data of the server at historical time nodes as training samples and whether the server has a fault at the historical time nodes as sample labels.

[0034] A fault prediction device, including: a memory and a processor;

[0035] The memory is configured to store a program;

[0036] The processor is configured to execute the program to implement each step of the fault prediction method as described above.

[0037] A readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each step of the fault prediction method as described above is implemented.

[0038] As can be seen from the above technical solutions, a fault prediction method, apparatus, device, and readable storage medium provided by an embodiment of the present application obtain historical parameter data of a first period and historical parameter data of a second period before a to-be-predicted time node of a server, where the duration of the first period is greater than the duration of the second period. Based on the historical parameter data of the first period, a reference parameter value range of the server at the to-be-predicted time node is predicted, and based on the historical parameter data of the second period, predicted parameter data of the server at the to-be-predicted time node is predicted. It is determined whether the predicted parameter data is within the reference parameter value range. If not, it is determined that there is a potential fault at the to-be-predicted time node of the server. In the present application, by using the historical parameter data of the first period, a reference parameter value range at the to-be-predicted time node is predicted, and then by using the historical parameter data of the second period, predicted parameter data at the to-be-predicted time node is predicted. Since the first period is greater than the second period, the obtained reference parameter value range will be more general, and the obtained predicted parameter data will be closer to the real data at the to-be-predicted time node. By determining whether the predicted parameter data is within the reference parameter value range, it can be determined whether there is a potential fault at the to-be-predicted time node of the server, thereby realizing the prediction of faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0040] Figure 1 FIG. is a flowchart of a fault prediction method provided by an embodiment of the present application;

[0041] Figure 2 FIG. is a schematic structural diagram of a fault prediction apparatus provided by an embodiment of the present application;

[0042] Figure 3 FIG. is a hardware structure block diagram of a fault prediction device disclosed by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0044] Figure 1The flowchart of a fault prediction method provided by an embodiment of the present application is referred to Figure 1 As shown, the method may include:

[0045] Step S100, obtain the historical parameter data of the first period and the historical parameter data of the second period of the server before the to-be-predicted time node.

[0046] Among them, the duration of the first period is greater than the duration of the second period.

[0047] Specifically, after determining the to-be-predicted time node, the historical parameter data of the first period and the historical parameter data of the second period of the server before the to-be-predicted time node can be obtained. The first period and the second period may be adjacent to the to-be-predicted time node or may be separated from the to-be-predicted time node by a preset time period.

[0048] Step S101, based on the historical parameter data of the first period, predict the reference parameter value range of the server at the to-be-predicted time node.

[0049] Specifically, in the above step, the historical parameter data of the first period of the server before the to-be-predicted time node can be obtained. Using the historical parameter data of the first period, the reference parameter value range of the to-be-predicted time node can be predicted. The reference parameter value range may include the upper limit and the lower limit of the reference parameter value range.

[0050] Step S102, based on the historical parameter data of the second period, predict the predicted parameter data of the server at the to-be-predicted time node.

[0051] Specifically, after obtaining the historical parameter data of the second period of the server before the to-be-predicted time node through the above steps, the predicted parameter data of the to-be-predicted time node can be predicted. The method of predicting the predicted parameter data of the to-be-predicted time node using the historical parameter data of the second period can be implemented by establishing a time series model.

[0052] Among them, the execution order of step S101 and step S102 can be executed simultaneously or can be executed in any order, and the implementation of the present application is not affected by the execution order.

[0053] Step S103, determine whether the predicted parameter data is within the reference parameter value range.

[0054] Specifically, after predicting the predicted parameter data and the reference parameter value range in the above steps, the reference parameter value range can be used as the parameter data standard of the to-be-predicted time node to determine whether the predicted parameter data is within the reference parameter value range. If not, execute step S104.

[0055] Step S104: Determine that there are potential faults in the server at the to-be-predicted time node.

[0056] Specifically, since the predicted parameter data at the to-be-predicted time node is not within the reference parameter value range, it can be determined that there are potential faults in the server at the to-be-predicted time node. Among them, when the predicted parameter data is higher than the upper limit of the reference parameter value range, the possible fault that may occur is that the predicted parameter data will exceed the server's bearing capacity. When the predicted parameter data is lower than the lower limit of the first reference parameter value range, the possible fault that may occur is that a certain server docking interface may malfunction, resulting in the server being unable to provide services.

[0057] In the above embodiment, a fault prediction method is provided. By obtaining the historical parameter data of the first cycle and the historical parameter data of the second cycle before the to-be-predicted time node of the server, where the duration of the first cycle is greater than the duration of the second cycle, based on the historical parameter data of the first cycle, the reference parameter value range of the to-be-predicted time node is predicted, and based on the historical parameter data of the second cycle, the predicted parameter data of the to-be-predicted time node is predicted. It is judged whether the predicted parameter data is within the reference parameter value range. If not, it is determined that there are potential faults in the server at the to-be-predicted time node. In this application, by using the historical parameter data of the first cycle, the reference parameter value range of the to-be-predicted time node is predicted, and then by using the historical parameter data of the second cycle, the predicted parameter data of the to-be-predicted time node is predicted. Since the first cycle is longer than the second cycle, the obtained reference parameter value range will be more general, and the obtained predicted parameter data will be closer to the real data of the to-be-predicted time node. By judging whether the predicted parameter data is within the reference parameter value range, it can be determined whether there are potential faults in the server at the to-be-predicted time node, thus realizing the prediction of faults.

[0058] In some embodiments of the present application, the process of step S101: predicting the reference parameter value range of the to-be-predicted time node based on the historical data of the first cycle is introduced. This process may include:

[0059] S11: Select the first target time node at the same time as the to-be-predicted time node within the first cycle.

[0060] Specifically, after determining the time node to be predicted, a time node that is at the same moment as the time node to be predicted can be selected from the first cycle as the first target time node. For example, if the current time node to be predicted is "Wednesday, April 15th", and the first cycle is from January 14th to April 14th, then the time nodes in the first cycle that are at the same time position as the time node to be predicted on April 15th can be January 15th, February 15th, March 15th, and the time nodes corresponding to each Wednesday in the first cycle. These nodes are used as the first target time nodes.

[0061] S12. Use the data within the preset time range before and after the first target time node to predict the value range of the reference parameter at the predicted time node.

[0062] Specifically, after determining the first target time node in the above step, the data within the preset time range before and after the second target time node can be selected to predict the value range of the reference parameter at the time node to be predicted. For example, if the current time node to be predicted is "Wednesday, April 15th", and the first cycle is from January 14th to April 14th, then January 15th, February 15th, March 15th, and the time nodes corresponding to Wednesdays in the first cycle are used as the first target time nodes. If the preset time range before and after is set to 1 day, the obtained data is the historical parameter data from January 14th to January 16th, from February 14th to February 16th, from March 14th to March 16th, and from Tuesday to Thursday in the first cycle. After obtaining the historical parameter data, the maximum and minimum values can be selected as the reference parameter range, or...

[0063] In some embodiments of the present application, the process of step S102, predicting the predicted parameter data at the time node to be predicted based on the historical data of the second cycle, is introduced. This process may include:

[0064] S21. Select a second target time node that is at the same moment as the time node to be predicted within the second cycle.

[0065] Specifically, after determining the time node to be predicted, a time node that is at the same moment as the time node to be predicted can be selected from the second cycle as the second target time node. For example, if the current time node to be predicted is "Wednesday, April 15th", and the first cycle is from April 1st to April 14th, then the time nodes that are at the same time position as the time node to be predicted on April 15th in the first cycle can be the Wednesdays from April 1st to April 14th. These nodes are used as the second target time nodes.

[0066] S22. Use the data within the preset time range before and after the second target time node to predict the predicted parameter data at the predicted time node.

[0067] Specifically, after determining the second target time node in the above steps, data within a preset time range before and after the second target time node can be selected to predict the predicted parameter data of the to-be-predicted time node. For example, if the current to-be-predicted time node is "Wednesday, April 15th", and the first cycle is from April 1st to April 14th, then the time node corresponding to Wednesday within the period from April 1st to April 14th is used as the second target time node. If the preset time range before and after is set to 1 day, the obtained data is the historical parameter data from Tuesday to Thursday within the period from April 1st to April 14th. After obtaining the historical parameter data, the average method can be used to add up all the data and finally calculate an average data as the predicted parameter data.

[0068] In the current data analysis field, after obtaining data, a neural network model can also be used to analyze the data to obtain corresponding prediction results. Based on this, in some embodiments of the present application, a neural network model can be used to predict whether there are potential fault hazards in the server at the to-be-predicted time node. Through machine learning, relevant features of the fault data can be obtained from historical data, so that when the to-be-predicted time node and data are input into the neural network model, it can be predicted whether there are potential fault hazards in the server.

[0069] Specifically, the parameter data of the server at the historical time node can be used as the training sample, and whether the server has a fault at the historical time node can be used as the sample label to train a fault prediction model. Using the trained fault prediction model, the predicted parameter data and the to-be-predicted time node can be input into the fault prediction model to determine whether there are potential fault hazards in the server at the to-be-predicted time node.

[0070] In the above embodiment, the parameter data of the server at the historical time node can be used as the training sample, and whether the server has a fault at the historical time node can be used as the sample label to train a fault prediction model. Using the trained fault prediction model, the fault prediction of the server at the to-be-predicted time node can be realized.

[0071] Furthermore, in the above embodiment, it can be determined whether there are potential fault hazards in the server at the to-be-predicted time node by judging whether the predicted parameter data is within the reference parameter value range; a neural network model can also be used to determine whether there are potential fault hazards in the server at the to-be-predicted time node. When both of the above methods determine that there are potential fault hazards in the server at the to-be-predicted time node or one of the methods determines that there are potential fault hazards in the server at the to-be-predicted time node, it can be determined that there are potential fault hazards in the server at the to-be-predicted time node; when both of the above methods determine that there are no potential fault hazards in the server at the to-be-predicted time node, it can be determined that there are no potential fault hazards in the server at the to-be-predicted time node.

[0072] Under normal circumstances, for potential server failures, operation and maintenance personnel need to handle them through some script programs or manually switch the server. However, manual handling also requires an increase in the human response time, which may lead to the failure not being processed in a timely manner. Based on this, in some embodiments of the present application, corresponding handling methods can be set for simple potential failures.

[0073] Specifically, the handling methods corresponding to potential failures can be set in advance. After determining that there is a potential failure at the predicted time node, it can be determined whether there is a corresponding handling method for the potential failure. If so, the potential failure is processed according to the handling method. If not, an alarm message is sent.

[0074] Among them, when the potential failure is that the server's load capacity is exceeded, the handling method can divert the data to be processed by the server, so that the data to be processed by the server is reduced; when the potential failure is that the server may fail, the handling method can replace the data to be processed by the server with other servers for processing.

[0075] In the above embodiments, for some potential failures that can be processed by machine automation, relevant handling methods are set in advance. When it is predicted that there is a potential failure at the predicted time node and there is a corresponding handling method, the potential failure can be processed according to the handling method, so that the potential failure is processed in a timely manner.

[0076] Further, in the above embodiments, after processing the potential failure according to the handling method, it can also be determined whether the potential failure is eliminated. If so, the processing data is recorded. If not, an alarm message is sent.

[0077] Next, a failure prediction device provided by an embodiment of the present application is described. The failure prediction device described below can be correspondingly referred to the failure prediction method described above.

[0078] Figure 2 For the structural schematic diagram of a failure prediction device provided by an embodiment of the present application, refer to Figure 2 As shown, the failure prediction device may include:

[0079] A historical data acquisition unit 10, configured to acquire the historical parameter data of the first period and the historical parameter data of the second period of the server before the predicted time node, where the duration of the first period is greater than the duration of the second period;

[0080]

[0081] ​A prediction data acquisition unit 30, configured to predict the predicted parameter data of the server at the to-be-predicted time node based on the historical parameter data of the second period;

[0082] A fault hidden danger judgment unit 40, configured to judge whether the predicted parameter data is within the reference parameter value range. If not, it is determined that there is a fault hidden danger in the server at the to-be-predicted time node.

[0083] In the above embodiment, a fault prediction method is provided. The historical data acquisition unit 10 acquires the historical parameter data of the first period and the historical parameter data of the second period of the server before the to-be-predicted time node. Among them, the duration of the first period is greater than the duration of the second period. The reference range acquisition unit 20 predicts the reference parameter value range of the to-be-predicted time node based on the historical parameter data of the first period, and the prediction data acquisition unit 30 predicts the predicted parameter data of the to-be-predicted time node based on the historical parameter data of the second period. The fault hidden danger judgment unit 40 judges whether the predicted parameter data is within the reference parameter value range. If not, it is determined that there is a fault hidden danger in the server at the to-be-predicted time node. In this application, by using the historical parameter data of the first period, the reference parameter value range of the to-be-predicted time node is predicted, and then by using the historical parameter data of the second period, the predicted parameter data of the to-be-predicted time node is predicted. Since the first period is greater than the second period, the obtained reference parameter value range will be more universal, and the obtained predicted parameter data will be closer to the real data of the to-be-predicted time node. By judging whether the predicted parameter data is within the reference parameter value range, it can be determined whether there is a fault hidden danger in the server at the to-be-predicted time node, so as to realize the prediction of faults.

[0084] Optionally, the fault prediction device may further include:

[0085] A fault hidden danger prediction unit, configured to input the predicted parameter data and the to-be-predicted time node into a fault prediction model, and determine whether there is a fault hidden danger in the server at the to-be-predicted time node. The fault prediction model is trained with the parameter data of the server at the historical time node as the training samples and the whether a fault occurs in the server at the historical time node as the sample labels.

[0086] Optionally, when the reference range acquisition unit 20 executes the step of predicting the reference parameter value range of the server at the to-be-predicted time node based on the historical parameter data of the first period, it may include:

[0087] Select a time node in the first period that is at the same time position as the to-be-predicted time node as the first target time node;

[0088] Using the data within a preset time range before and after the first target time node, the reference parameter value range of the server at the prediction time node is predicted.

[0089] Optionally, the step of the prediction data acquisition unit 30 executing to predict the prediction parameter data of the server at the to-be-predicted time node based on the historical parameter data of the second cycle may include:

[0090] Selecting a time node within the second cycle that is at the same time position as the to-be-predicted time node as the second target time node;

[0091] Using the data within a preset time range before and after the second target time node, the prediction parameter data of the server at the prediction time node is predicted. [[ID=ll]]

[0092] Optionally, the fault prediction device may further include:

[0093] A processing method determination unit, configured to determine whether there is a corresponding processing method for the potential fault; if so, process the potential fault according to the processing method; if not, send an alarm message.

[0094] Optionally, the fault prediction device may further include:

[0095] A processing result determination unit, configured to determine whether the potential fault is eliminated; if so, record the processing data; if not, send an alarm message.

[0096] An embodiment of the present application further provides a fault prediction device, Figure 3 showing a hardware structure block diagram of the fault prediction device, referring to Figure 3 , the hardware structure of the fault prediction device may include: at least one processor 1, at least one communication interface 2, at least one memory 3, and at least one communication bus 4;

[0097] In the embodiment of the present application, the number of the processor 1, the communication interface 2, the memory 3, and the communication bus 4 is at least one, and the processor 1, the communication interface 2, and the memory 3 complete mutual communication through the communication bus 4;

[0098] The processor 1 may be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention, etc.;

[0099] The memory 3 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory;

[0100] Among them, the memory stores a program, and the processor can call the program stored in the memory. The program is used to: implement each processing flow in the foregoing fault prediction method.

[0101] An embodiment of the present application further provides a storage medium, which can store a program suitable for execution by a processor. The program is used to: implement each processing flow in the foregoing fault prediction method.

[0102] Finally, it should also be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0103] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined with each other, and the same or similar parts can be referred to each other.

[0104] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A fault prediction method, characterized in that Including: Obtain the historical parameter data of the first period and the historical parameter data of the second period before the to-be-predicted time node of the server, where the duration of the first period is greater than the duration of the second period; Based on the historical parameter data of the first period, predict the reference parameter value range of the server at the to-be-predicted time node; Based on the historical parameter data of the second period, predict the predicted parameter data of the server at the to-be-predicted time node; Judge whether the predicted parameter data is within the reference parameter value range; If not, determine that there is a potential fault at the to-be-predicted time node of the server; Wherein, when the predicted parameter data is higher than the upper limit of the reference parameter value range, the fault includes: the predicted parameter data exceeds the bearing capacity of the server; when the predicted parameter data is lower than the lower limit of the reference parameter value range, the fault includes: the server docking interface fails.

2. The method according to claim 1, wherein Also including: Input the predicted parameter data and the to-be-predicted time node into the fault prediction model to determine whether there is a potential fault at the to-be-predicted time node of the server. The fault prediction model is trained with the parameter data of the server at the historical time node as the training samples and whether the server has a fault at the historical time node as the sample labels.

3. The method according to claim 1, characterized in that Based on the historical parameter data of the first period, predicting the reference parameter value range of the server at the to-be-predicted time node includes: Select the first target time node at the same moment as the to-be-predicted time node within the first period; Use the data within the preset time range before and after the first target time node to predict the reference parameter value range of the server at the to-be-predicted time node.

4. The method according to claim 1, characterized in that Based on the historical parameter data of the second period, predicting the predicted parameter data of the server at the to-be-predicted time node includes: Select the second target time node at the same moment as the to-be-predicted time node within the second period; Use the data within the preset time range before and after the second target time node to predict the predicted parameter data of the server at the to-be-predicted time node.

5. The method according to any one of claims 1-4, characterized in that Also including: Judge whether there is a corresponding processing method for the potential fault; If so, process the potential fault according to the processing method; If not, send an alarm message.

6. The method according to claim 5, characterized in that, After processing the potential fault according to the processing method, it also includes: Judge whether the potential fault is eliminated; If so, record the processing data; If not, send an alarm message.

7. A fault prediction device, characterized in that, Including: The historical data acquisition unit is used to obtain the historical parameter data of the first period and the historical parameter data of the second period before the to-be-predicted time node of the server, where the duration of the first period is greater than the duration of the second period; The reference range acquisition unit is used to predict the reference parameter value range of the server at the to-be-predicted time node based on the historical parameter data of the first period; The predicted data acquisition unit is used to predict the predicted parameter data of the server at the to-be-predicted time node based on the historical parameter data of the second period; A fault hidden danger judgment unit, configured to judge whether the predicted parameter data is within the reference parameter value range. If not, it is determined that there is a fault hidden danger in the server at the to-be-predicted time node Wherein, when the predicted parameter data is higher than the upper limit of the reference parameter value range, the faults include: the predicted parameter data exceeds the bearing capacity of the server; when the predicted parameter data is lower than the lower limit of the reference parameter value range, the faults include: a server docking interface fault 8. The device according to claim 7, characterized in that It further includes: A fault prediction unit, configured to input the predicted parameter data and the to-be-predicted time node into a fault prediction model, and determine whether there is a fault hidden danger in the server at the to-be-predicted time node. The fault prediction model is trained with the parameter data of the server at historical time nodes as training samples and whether the server has a fault at the historical time nodes as sample labels 9. A fault prediction device, characterized in that, It includes: A memory and a processor; The memory is used to store programs; The processor is configured to execute the program to implement each step of the fault prediction method according to any one of claims 1-6 10. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, each step of the fault prediction method according to any one of claims 1-6 is implemented

Images