Problem handling methods, devices, electronic equipment, and storage media based on heapdump

Abstract

This application discloses a problem processing method, apparatus, electronic device, and storage medium based on heapdump. The method includes the following steps: acquiring a heapdump file and extracting indicator data of the problem corresponding to the heapdump file from the heapdump file; inputting the feature information corresponding to the indicator data into a problem analysis model to obtain a target prediction probability; wherein, the problem analysis model is obtained by training a model to be trained based on feature training samples and standard prediction probability labels of the feature training samples; and determining a first solution corresponding to the problem based on the target prediction probability. That is, the embodiments of this application determine the first solution corresponding to the problem through a problem analysis model, thereby improving the efficiency of identifying and solving problems.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a problem processing method, apparatus, electronic device and storage medium based on heapdump. Background Technology

[0002] With the rise and development of container technology, more and more enterprises are running applications in containers. In this environment, container memory management becomes particularly critical.

[0003] Currently, issues related to container memory management (including but not limited to memory leaks and memory overflows) are resolved by developers through heapdump files. This requires developers to have a certain level of knowledge and analytical proficiency, but the varying levels of knowledge and analytical proficiency among developers contribute to their inefficiency in troubleshooting and resolving problems.

[0004] Therefore, in practical applications, a solution is needed that can improve the efficiency of troubleshooting and resolving problems. Summary of the Invention

[0005] The main objective of this application is to provide a problem handling method, apparatus, electronic device, and storage medium based on heapdump, aiming to solve the technical problem of low efficiency in troubleshooting and resolving problems.

[0006] To achieve the above objectives, this application provides a problem-solving method based on heapdump, which includes the following steps:

[0007] Obtain the heapdump file and extract the metric data corresponding to the problem in the heapdump file;

[0008] The feature information corresponding to the indicator data is input into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples;

[0009] Based on the predicted probability of the target, a first solution to the problem is determined.

[0010] For example, before inputting the feature information corresponding to the indicator data into the problem analysis model to obtain the analysis results, the following steps are included:

[0011] Obtain feature training samples and standard predicted probability labels for the feature training samples;

[0012] Based on the feature training samples and the standard predicted probability labels, the model to be trained is iteratively trained to obtain a problem analysis model that meets the accuracy requirements.

[0013] For example, the step of iteratively training the model to be trained based on the feature training samples and the standard predicted probability labels to obtain a problem analysis model that meets the accuracy requirements includes:

[0014] The feature training samples are input into the model to be trained to obtain the training prediction probability;

[0015] The difference between the trained predicted probability and the standard predicted probability label is calculated to obtain the error result;

[0016] Based on the error result, determine whether the error result meets the error standard indicated by the preset error threshold range;

[0017] If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the feature training sample into the model to be trained to obtain the training prediction probability, and stop training until the error result meets the error standard indicated by the preset error threshold range, and obtain the problem analysis model.

[0018] For example, after determining the first solution corresponding to the problem based on the target prediction probability, the process includes:

[0019] Obtain feedback from developers regarding the first solution;

[0020] The first solution is updated based on the feedback information.

[0021] For example, the method further includes:

[0022] The indicator data is input into the rule engine, and a second solution is obtained through matching.

[0023] Output the second solution.

[0024] For example, obtaining the heapdump file includes:

[0025] The receiving terminal sends a heapdump file, which is generated by the terminal when it detects that the application is in a dead state.

[0026] For example, the heapdump file is sent by an upload script, the upload script receives the heapdump file from a shell script, and the shell script receives the heapdump file from the terminal; the heapdump file sent by the receiving terminal includes:

[0027] Receive the heapdump file sent by the upload script.

[0028] For example, to achieve the above objectives, this application also provides a heapdump-based problem processing apparatus, which includes:

[0029] The first acquisition module is used to acquire a heapdump file and extract the indicator data of the problem corresponding to the heapdump file from the heapdump file.

[0030] The input module is used to input the feature information corresponding to the indicator data into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples;

[0031] The determination module is used to determine the first solution corresponding to the problem based on the predicted probability of the target.

[0032] For example, the heapdump-based problem processing apparatus further includes:

[0033] The second acquisition module is used to acquire feature training samples and standard prediction probability labels of the feature training samples;

[0034] The training module is used to iteratively train the model to be trained based on the feature training samples and the standard predicted probability labels to obtain a problem analysis model that meets the accuracy requirements.

[0035] For example, the training module is specifically used for:

[0036] The feature training samples are input into the model to be trained to obtain the training prediction probability;

[0037] The difference between the trained predicted probability and the standard predicted probability label is calculated to obtain the error result;

[0038] Based on the error result, determine whether the error result meets the error standard indicated by the preset error threshold range;

[0039] If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the feature training sample into the model to be trained to obtain the training prediction probability, and stop training until the error result meets the error standard indicated by the preset error threshold range, and obtain the problem analysis model.

[0040] For example, the heapdump-based problem processing apparatus further includes:

[0041] The third acquisition module is used to acquire feedback information input by developers for the first solution;

[0042] An update module is used to update the first solution based on the feedback information.

[0043] For example, the heapdump-based problem processing apparatus further includes:

[0044] The matching module is used to input the indicator data into the rule engine and match it to obtain a second solution;

[0045] The output module is used to output the second solution.

[0046] For example, obtaining the heapdump file includes:

[0047] The receiving terminal sends a heapdump file, which is generated by the terminal when it detects that the application is in a dead state.

[0048] For example, the heapdump file is sent by an upload script, the upload script receives the heapdump file from a shell script, and the shell script receives the heapdump file from the terminal; the first acquisition module is specifically used for:

[0049] Receive the heapdump file sent by the upload script.

[0050] For example, to achieve the above objectives, this application also provides an electronic device, the device comprising: a memory, a processor, and a heapdump-based problem handler stored on the memory and executable on the processor, the heapdump-based problem handler being configured to implement the steps of the heapdump-based problem handling method as described above.

[0051] For example, to achieve the above objectives, this application also provides a computer-readable storage medium storing a heapdump-based problem handling program, which, when executed by a processor, implements the steps of the heapdump-based problem handling method as described above.

[0052] Compared to related technologies that require developers to possess a certain level of knowledge and analytical proficiency, the varying levels of such proficiency among developers often lead to inefficiencies in problem identification and resolution. This application, however, obtains a heapdump file and extracts the corresponding problem's indicator data from it. The feature information corresponding to the indicator data is then input into a problem analysis model to obtain a target prediction probability. This problem analysis model is trained using feature training samples and their standard prediction probability labels. Based on the target prediction probability, a first solution to the problem is determined. In other words, this application determines the first solution to the problem through a problem analysis model, eliminating the need for developers with specific knowledge and analytical proficiency to identify and resolve the issue. This avoids the impact of varying developer knowledge and analytical proficiency on problem identification and resolution efficiency, thereby improving overall efficiency. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating an embodiment of the problem handling method based on heapdump in this application;

[0054] Figure 2 This is a schematic diagram of the module of the problem processing device based on heapdump in this application;

[0055] Figure 3 This is a schematic diagram of the hardware operating environment involved in the embodiments of this application.

[0056] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0058] This application provides a problem handling method based on heapdump, referring to... Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the problem handling method based on heapdump in this application.

[0059] This application provides an embodiment of a problem-handling method based on heapdump. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order. This heapdump-based problem-handling method is applied to electronic devices, including terminals, servers, etc., and includes:

[0060] Step S110: Obtain the heapdump file and extract the indicator data of the problem corresponding to the heapdump file from the heapdump file.

[0061] Step S120: Input the feature information corresponding to the indicator data into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples.

[0062] Step S130: Based on the target prediction probability, determine the first solution corresponding to the problem.

[0063] The following is a detailed description of each step:

[0064] Step S110: Obtain the heapdump file and extract the indicator data of the problem corresponding to the heapdump file from the heapdump file.

[0065] It should be noted that in practical applications, Linux cgroups are used for container memory management, Prometheus + Grafana is used for memory monitoring, and Heapster + InfluxDB + Grafana is used for overall container monitoring. However, these solutions cannot track and analyze problems related to memory management. This application's embodiments, through a problem analysis model, achieve the tracking and analysis of problems related to memory management.

[0066] The heapdump file contains information including, but not limited to, module information, thread information, stack traces, and exception information. This heapdump file is used to capture and preserve the state of the problem at the time it occurred. A snapshot of the heapdump file allows for quick access to the entire heapdump file. A snapshot is a reference or pointer to data stored on a storage device (i.e., the heapdump file), representing the state of the data at a specific moment. Its core working principle involves creating a list of pointers that indicates the address for reading the data, providing a snapshot of the data at a given instant.

[0067] It should be noted that the indicator data is extracted specifically for the aforementioned issues; that is, the indicator data is problem-related, and anomalies in the corresponding indicator data may lead to the occurrence of the corresponding problems. It should also be noted that when performing focused screening of heapdump files based on JVM (Java Virtual Machine) principles, the following indicators are screened: large objects, GCRoot nodes and GC (Garbage Collection) chains, and thread health status, thereby obtaining the corresponding indicator data, such as the name of large objects, the size of large objects, GCRoot node information and GC chain information, and thread health.

[0068] In addition to memory leaks and memory overflows, the aforementioned problems may also include abnormal object lifecycles and abnormal GC performance. An object lifecycle refers to the complete process of an object's creation, survival, and destruction. When an abnormality occurs, the object fails to complete this complete process, for example, its lifespan is shorter than expected.

[0069] Step S120: Input the feature information corresponding to the indicator data into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples.

[0070] It should be noted that feature training samples are feature samples used for training and include a certain amount of feature information. The standard prediction probability label of a feature training sample is the standard prediction probability of the corresponding feature training sample. For example, there are feature training samples X and Y, where the standard prediction probability of feature training sample X is 0.7 and the standard prediction probability of feature training sample Y is 0.3. That is, the standard prediction probability label of feature training sample X is 0.7 and the standard prediction probability label of feature training sample Y is 0.3.

[0071] It should be noted that the prediction probability is the probability of predicting whether the problem corresponding to the heapdump file is a memory leak, memory overflow, object lifecycle abnormality, or GC performance abnormality. For example, given a feature training sample X, the prediction probabilities for feature training sample X include a memory leak probability of 0.3, a memory overflow probability of 0.2, an object lifecycle abnormality probability of 0.1, and a GC performance abnormality probability of 0.4.

[0072] It should be noted that the problem analysis model is a classification model, that is, the model to be trained is constructed by a classification algorithm, which includes, but is not limited to, decision trees and random forests.

[0073] It should be noted that the feature information is a feature vector generated from the indicator data. After inputting the feature vector into the problem analysis model, the probability that the problem corresponding to the heapdump file belongs to memory leak, memory overflow, object lifecycle abnormality, and GC performance abnormality can be obtained, that is, the target prediction probability can be obtained.

[0074] For example, before inputting the feature information corresponding to the indicator data into the problem analysis model to obtain the analysis result, the process includes: obtaining feature training samples and standard prediction probability labels of the feature training samples; and iteratively training the model to be trained based on the feature training samples and the standard prediction probability labels to obtain a problem analysis model that meets the accuracy requirements.

[0075] For example, the step of iteratively training the model to be trained based on the feature training samples and the standard prediction probability labels to obtain a problem analysis model that meets the accuracy conditions includes: inputting the feature training samples into the model to be trained to obtain training prediction probabilities; calculating the difference between the training prediction probabilities and the standard prediction probability labels to obtain an error result; determining whether the error result meets the error standard indicated by a preset error threshold range based on the error result; if the error result does not meet the error standard indicated by the preset error threshold range, returning to the step of inputting the feature training samples into the model to be trained to obtain training prediction probabilities, until the error result meets the error standard indicated by the preset error threshold range, then stopping the training to obtain the problem analysis model.

[0076] It should be noted that the feature training samples are input into the model to be trained to obtain the training prediction probability, which is obtained by performing prediction tests on the model during training.

[0077] Since there is an error between the training result of the model and the actual result, the error result is allowed to be within a preset error threshold range, so as to further determine whether the error result meets the error standard indicated by the preset error threshold range.

[0078] If the error result does not meet the error standard indicated by the preset error threshold range, it indicates that the error of the model to be trained is too large in this training. The process then returns to the step of inputting the feature training samples into the model to obtain the training prediction probability, i.e., iterative training is performed until the error result meets the error standard indicated by the preset error threshold range. Training then stops, resulting in the test prediction model that meets the accuracy requirements, thereby improving the accuracy of the model's predictions. The preset error threshold can be set as needed, and this embodiment does not impose specific limitations.

[0079] Step S130: Based on the target prediction probability, determine the first solution corresponding to the problem.

[0080] It's understandable that each problem corresponds to a primary solution. By finding the problem with the highest probability from the target's predicted probabilities, we can identify the primary solution for that problem. This solution is then pushed to the developers for their reference, allowing them to solve the problem using this primary solution.

[0081] Compared to related technologies that require developers to possess a certain level of knowledge and analytical proficiency, the varying levels of such proficiency among developers often lead to inefficiencies in problem identification and resolution. This application, however, obtains a heapdump file and extracts the corresponding problem's indicator data from it. The feature information corresponding to the indicator data is then input into a problem analysis model to obtain a target prediction probability. This problem analysis model is trained using feature training samples and their standard prediction probability labels. Based on the target prediction probability, a first solution to the problem is determined. In other words, this application determines the first solution to the problem through a problem analysis model, eliminating the need for developers with specific knowledge and analytical proficiency to identify and resolve the issue. This avoids the impact of varying developer knowledge and analytical proficiency on problem identification and resolution efficiency, thereby improving overall efficiency.

[0082] For example, after determining the first solution corresponding to the problem based on the target prediction probability, the process includes: obtaining feedback information input by developers regarding the first solution; and updating the first solution based on the feedback information.

[0083] It should be noted that the first solution is intended to provide suggestions for solving the above problem; it is only for developers' reference and does not mean that developers must solve the problem according to the first solution. If developers find any shortcomings in the first solution after solving it, they can provide feedback to help the electronic device improve upon it.

[0084] For example, the method further includes: inputting the indicator data into a rule engine to obtain a second solution; and outputting the second solution.

[0085] It is understood that this application embodiment provides a multi-dimensional solution, enabling developers to combine the first and second solutions to solve the problem, thereby further improving the efficiency of problem-solving. The rule engine includes noun matching; that is, if a noun in the rule engine exists in the metric data, the second solution corresponding to that noun can be matched from the rule engine.

[0086] It should be noted that the output methods of the second solution include speaker output, display output, etc. It can be understood that the first solution can also output in the same way.

[0087] For example, obtaining the heapdump file includes: receiving a heapdump file sent by a terminal, wherein the heapdump file is generated by the terminal when it detects that the application is in a frozen state.

[0088] Terminals include, but are not limited to, mobile terminals and PCs (Personal Computers). When a target application running on this terminal triggers an Out Of Memory (OOM) error, it enters a frozen state, generating a heapdump file. This heapdump file is then uploaded to the electronic device, typically a cloud server. Only after receiving the heapdump file does the cloud server investigate the problem and provide corresponding solutions.

[0089] It is understandable that storing heapdump files on a cloud server for unified management can greatly simplify the heapdump file management process and make it easier to troubleshoot, resolve, and accumulate experience.

[0090] For example, the heapdump file is sent by an upload script, the upload script receives the heapdump file from a shell script, and the shell script receives the heapdump file from the terminal; receiving the heapdump file sent by the terminal includes: receiving the heapdump file sent by the upload script.

[0091] It should be noted that the embodiments of this application participate in the lifecycle of the container corresponding to the application in a scripted manner, thereby achieving low intrusion into the container and no burden on the application operation, thus realizing the lightweighting of the application.

[0092] In addition, this application also provides a problem processing apparatus based on heapdump, referring to... Figure 2 , Figure 2This is a schematic diagram of the heapdump-based problem processing apparatus of this application. The heapdump-based problem processing apparatus includes:

[0093] The first acquisition module 20 is used to acquire a heapdump file and extract the indicator data of the problem corresponding to the heapdump file from the heapdump file.

[0094] Input module 21 is used to input the feature information corresponding to the indicator data into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples;

[0095] The determination module 22 is used to determine the first solution corresponding to the problem based on the target prediction probability.

[0096] For example, the heapdump-based problem processing apparatus further includes:

[0097] The second acquisition module is used to acquire feature training samples and standard prediction probability labels of the feature training samples;

[0098] The training module is used to iteratively train the model to be trained based on the feature training samples and the standard predicted probability labels to obtain a problem analysis model that meets the accuracy requirements.

[0099] For example, the training module is specifically used for:

[0100] The feature training samples are input into the model to be trained to obtain the training prediction probability;

[0101] The difference between the trained predicted probability and the standard predicted probability label is calculated to obtain the error result;

[0102] Based on the error result, determine whether the error result meets the error standard indicated by the preset error threshold range;

[0103] If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the feature training sample into the model to be trained to obtain the training prediction probability, and stop training until the error result meets the error standard indicated by the preset error threshold range, and obtain the problem analysis model.

[0104] For example, the heapdump-based problem processing apparatus further includes:

[0105] The third acquisition module is used to acquire feedback information input by developers for the first solution;

[0106] An update module is used to update the first solution based on the feedback information.

[0107] For example, the heapdump-based problem processing apparatus further includes:

[0108] The matching module is used to input the indicator data into the rule engine and match it to obtain a second solution;

[0109] The output module is used to output the second solution.

[0110] For example, obtaining the heapdump file includes:

[0111] The receiving terminal sends a heapdump file, which is generated by the terminal when it detects that the application is in a dead state.

[0112] For example, the heapdump file is sent by an upload script, the upload script receives the heapdump file from a shell script, and the shell script receives the heapdump file from the terminal; the first acquisition module 20 is specifically used for:

[0113] Receive the heapdump file sent by the upload script.

[0114] The specific implementation of the problem processing device based on heapdump in this application is basically the same as the various embodiments of the problem processing method based on heapdump described above, and will not be repeated here.

[0115] In addition, this application also provides an electronic device. For example... Figure 3 As shown, Figure 3 This is a schematic diagram of the hardware operating environment involved in the embodiments of this application.

[0116] like Figure 3 As shown, the electronic device may include a processor 301, a communication interface 302, a memory 303, and a communication bus 304. The processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304. The memory 303 is used to store computer programs. When the processor 301 executes the program stored in the memory 303, it implements the steps of the problem handling method based on heapdump.

[0117] The communication bus 304 mentioned in the aforementioned electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus 304 can be divided into an address bus, a data bus, and a control bus, etc. For ease of illustration, it is represented by only one thick line in the figure, but this does not indicate that there is only one bus or one type of bus.

[0118] Communication interface 302 is used for communication between the above-mentioned electronic device and other devices.

[0119] The memory 303 may include random access memory (RMD) or non-volatile memory (NM), such as at least one disk storage device. Optionally, the memory 303 may also be at least one storage device located remotely from the aforementioned processor 301.

[0120] The processor 301 mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0121] The specific implementation of the electronic device in this application is basically the same as the various embodiments of the problem handling method based on heapdump described above, and will not be repeated here.

[0122] Furthermore, embodiments of this application also propose a computer-readable storage medium storing a heapdump-based problem handling program, which, when executed by a processor, implements the steps of the heapdump-based problem handling method described above.

[0123] The specific implementation of the computer-readable storage medium in this application is basically the same as the embodiments of the problem handling method based on heapdump described above, and will not be repeated here.

[0124] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system 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 system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0125] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0126] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of 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. Based on this understanding, the technical solution of this application, in essence, 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) as described above, 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.

[0127] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance. It should also be understood that although the terms "first," "second," etc., are used in the text to describe various elements in some embodiments of this application, these elements should not be limited by these terms. These terms are merely used to distinguish one element from another. For example, a first table may be named a second table, and similarly, a second table may be named a first table, without departing from the scope of the various described embodiments. Both the first table and the second table are tables, but they are not the same table.

[0128] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A problem-solving method based on heapdump, characterized in that, The problem handling method based on heapdump includes the following steps: Obtain the heapdump file and extract the metric data corresponding to the problem in the heapdump file; The feature information corresponding to the indicator data is input into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples; Based on the predicted probability of the target, a first solution to the problem is determined; Before inputting the feature information corresponding to the indicator data into the problem analysis model to obtain the analysis results, the following steps are included: Obtain feature training samples and standard predicted probability labels for the feature training samples; Based on the feature training samples and the standard predicted probability labels, the model to be trained is iteratively trained to obtain a problem analysis model that meets the accuracy requirements. The step of iteratively training the model to be trained based on the feature training samples and the standard predicted probability labels to obtain a problem analysis model that meets the accuracy requirements includes: The feature training samples are input into the model to be trained to obtain the training prediction probability; The difference between the trained predicted probability and the standard predicted probability label is calculated to obtain the error result; Based on the error result, determine whether the error result meets the error standard indicated by the preset error threshold range; If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the feature training sample into the model to be trained to obtain the training prediction probability, and stop training until the error result meets the error standard indicated by the preset error threshold range, and obtain the problem analysis model.

2. The problem handling method based on heapdump as described in claim 1, characterized in that, After determining the first solution to the problem based on the target prediction probability, the process includes: Obtain feedback from developers regarding the first solution; The first solution is updated based on the feedback information.

3. The problem handling method based on heapdump as described in claim 1, characterized in that, The method further includes: The indicator data is input into the rule engine, and a second solution is obtained through matching. Output the second solution.

4. The problem handling method based on heapdump as described in claim 1, characterized in that, The process of obtaining the heapdump file includes: The receiving terminal sends a heapdump file, which is generated by the terminal when it detects that the application is in a dead state.

5. The problem handling method based on heapdump as described in claim 4, characterized in that, The heapdump file is sent by the upload script, the upload script receives the heapdump file from the shell script, and the shell script receives the heapdump file from the terminal; The heapdump file sent by the receiving terminal includes: Receive the heapdump file sent by the upload script.

6. A problem processing device based on heapdump, characterized in that, The heapdump-based problem processing device includes: The first acquisition module is used to acquire a heapdump file and extract the indicator data of the problem corresponding to the heapdump file from the heapdump file. The input module is used to input the feature information corresponding to the indicator data into the problem analysis model to obtain the target prediction probability; wherein, the problem analysis model is obtained by training the model to be trained based on the feature training samples and the standard prediction probability labels of the feature training samples; The determination module is used to determine a first solution to the problem based on the predicted probability of the target. The heapdump-based problem processing device is also used to achieve: Obtain feature training samples and standard predicted probability labels for the feature training samples; Based on the feature training samples and the standard predicted probability labels, the model to be trained is iteratively trained to obtain a problem analysis model that meets the accuracy requirements. The heapdump-based problem processing device is also used to achieve: The feature training samples are input into the model to be trained to obtain the training prediction probability; The difference between the trained predicted probability and the standard predicted probability label is calculated to obtain the error result; Based on the error result, determine whether the error result meets the error standard indicated by the preset error threshold range; If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the feature training sample into the model to be trained to obtain the training prediction probability, and stop training until the error result meets the error standard indicated by the preset error threshold range, and obtain the problem analysis model.

7. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and a heapdump-based problem handler stored in the memory and executable on the processor, the heapdump-based problem handler being configured to implement the steps of the heapdump-based problem handling method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a heapdump-based problem handling program, which, when executed by a processor, implements the steps of the heapdump-based problem handling method as described in any one of claims 1 to 5.

Images