Real-time monitoring method for running state of computer software

Abstract

The invention discloses a computer software running state real-time monitoring method, which relates to the technical field of computers, and comprises the following steps: performing hierarchical data acquisition on a resource layer and an application layer of to-be-tested computer software based on a preset acquisition mode to obtain software running parameter information; classifying the software operation parameter information based on the software model to obtain an operation parameter data pool of each piece of software, and arranging operation data in the operation parameter data pool of each piece of software based on the software operation process to obtain an operation parameter relation chain of each piece of software; the method comprises the following steps: performing multi-dimensional analysis on a dependency relationship among software in a computer to be tested in a non-running state to obtain a software static dependency relationship, performing multi-dimensional analysis on a dependency relationship among software in a running state to obtain a software dynamic dependency relationship, and fusing to obtain a dependency relationship network among the software; and constructing a computer software relation network in combination with the operation parameter relation chain. The method has the effect of improving the real-time monitoring efficiency of the running state of the computer software.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method for real-time monitoring of the running status of computer software. Background Technology

[0002] Currently, a computer is a modern electronic computing machine used for high-speed calculations. It can perform numerical calculations, logical calculations, and has storage and memory functions. It is a modern intelligent electronic device capable of automatically and rapidly processing massive amounts of data according to a program. Computer software refers to the collection of computer programs, data, and all documentation required for developing, using, and maintaining programs. Abnormalities in the operation of computer software can affect people's work, life, and entertainment; therefore, real-time monitoring of the operating status of computer software is crucial.

[0003] Existing real-time monitoring methods for computer software operation status involve data collection, transmission, analysis, and alarm notification triggering. However, these methods suffer from data silos and incompatibility between different software programs. Consequently, when one software program malfunctions, it fails to promptly alert other related programs, resulting in low efficiency and room for improvement in real-time monitoring. Summary of the Invention

[0004] To improve the efficiency of real-time monitoring of computer software operating status, this application provides a method for real-time monitoring of computer software operating status.

[0005] This application provides a method for real-time monitoring of the running status of computer software, which adopts the following technical solution:

[0006] A method for real-time monitoring of the running status of computer software includes the following steps:

[0007] Step S1: Based on a preset acquisition method, hierarchical data acquisition is performed on the resource layer and application layer of the computer software under test to obtain software operation parameter information.

[0008] Step S2: Classify the software operation parameter information based on the software model to obtain the operation parameter data pool of each software, and arrange the operation data in the operation parameter data pool of each software based on the software operation process to obtain the operation parameter relationship chain of each software.

[0009] Step S3: Analyze the dependencies between software in the computer under test in the non-running state to obtain the static software dependencies, and analyze the dependencies between software in the running state to obtain the dynamic software dependencies. Combine these to obtain the software dependency network, and construct the computer software relationship network by combining the running parameter relationship chain.

[0010] Step S4: Generate a dynamic threshold dataset using the historical alarm information of the computer under test;

[0011] Step S5: Compare the software running parameter information with the dynamic threshold dataset to determine whether there are any abnormalities in the running data of the software running parameter information, and combine the computer software relationship network to classify the running data that need to be alarmed to obtain the alarm priority result.

[0012] Step S6: Based on the alarm priority results, the software operation parameter information is sent to the backend system in a hierarchical manner and stored.

[0013] Preferably, the monitoring objects of the computer under test include resource layer monitoring objects and application layer monitoring objects;

[0014] The resource layer monitoring objects include CPU utilization, memory usage, disk I / O, network traffic, and number of processes;

[0015] The application layer monitoring objects include metrics, logs, and links;

[0016] Obtain the service mesh's Sidecar agent and establish a signal connection link with the computer under test. Based on the service mesh's Sidecar agent, detect and collect resource layer monitoring objects to obtain resource layer monitoring data information.

[0017] An agent is installed on the computer under test to collect indicator information and log information.

[0018] An SDK is integrated into the computer's software code to collect and report indicator information and link information;

[0019] The resource layer monitoring data, indicator information, log information, and link information are combined to form software operation parameter information.

[0020] Preferably, each software in the computer under test is numbered to obtain a software feature number;

[0021] Based on the software model corresponding to each running data in the software running parameter information, the running data in the software running parameter information is classified to obtain running parameter classification information, which includes the running parameter data pool of each software.

[0022] Based on the operation flow of each software in the computer under test, the operation data in the operation parameter data pool of each software are arranged in parallel and sequentially to obtain the operation parameter relationship chain of each software.

[0023] Preferably, the first static dependency is determined based on the location of each software, the second static dependency is identified based on the configuration file of each software, and the third static dependency is obtained by detecting the installation of newly added dependency libraries during the installation process of each software.

[0024] The first static dependency, the second static dependency, and the third static dependency are combined to form a software static dependency. Based on the software static dependency network, the degree of association between each software and other software when the software is not running is determined to obtain the software static association coefficient.

[0025] Based on the system tools of the computer under test, the correlation between each software process and other software during runtime is obtained to obtain the first dynamic dependency relationship;

[0026] The network connection of the computer under test is monitored to identify whether there are listening ports between the software. If there are listening ports, it is determined that there is a dependency relationship and the second dynamic dependency relationship is output.

[0027] Preferably, the first dynamic dependency relationship and the second dynamic dependency relationship are combined to form a software dynamic dependency relationship. Based on the software dynamic dependency relationship network, the degree of association between each software and other software when the software is not running is determined to obtain the software dynamic association coefficient.

[0028] The software correlation coefficient is obtained by adding the static correlation coefficient between each software to the corresponding dynamic correlation coefficient.

[0029] The combination of static and dynamic software dependencies forms a network of inter-software dependencies;

[0030] A computer software relationship network is constructed based on the inter-software dependency network and the relationship chain of the running parameters of each software.

[0031] Preferably, historical operational alarm information of the computer under test is obtained, including historical alarm data type, historical alarm timestamp, historical alarm software type, and historical operational alarm severity.

[0032] The historical alarm information of the computer under test is calculated based on machine learning algorithms to learn the historical patterns of the alarms in the computer software operation. Based on time series, prediction is performed to generate a dynamic threshold dataset, which includes the dynamic threshold of each running data in the software operation parameter information.

[0033] Preferably, each running data in the software running parameter information is matched one-to-one with each dynamic threshold in the dynamic threshold dataset and compared to determine whether there is any abnormality in each running data in the software running parameter information. If there is an abnormality, the running data is marked as a type of alarm running data, the software containing the type of alarm running data is marked as abnormal software, and a first-level alarm signal is output.

[0034] Based on the operational parameter relationship chain of each software, the operational data other than the operational data of the first type of alarm in the operational parameter relationship chain are marked as the second type of alarm operational data, and a second-level alarm signal is output.

[0035] Based on the software dependency network, software that is associated with abnormal software is marked as abnormal associated software;

[0036] Based on the software correlation coefficient, software with a correlation degree greater than the preset software correlation coefficient threshold is marked as high-risk abnormal software. The running data in the running parameter relationship chain of high-risk abnormal software is marked as three types of alarm running data, and three types of alarm signals are output.

[0037] Based on the software correlation coefficient, software with a correlation degree less than or equal to the preset software correlation coefficient threshold is marked as low-risk abnormal software. The running data in the running parameter relationship chain of low-risk abnormal software is marked as four types of alarm running data, and four types of alarm signals are output.

[0038] The combination of the first type of alarm operation data and the second type of alarm operation data forms a first priority alarm result, wherein the first type of alarm operation data has a higher priority than the second type of alarm operation data. The combination of the third type of alarm operation data and the fourth type of alarm operation data forms a second priority alarm result, wherein the third type of alarm operation data has a higher priority than the fourth type of alarm operation data. The combination of the first priority alarm result and the second priority alarm result forms an alarm priority result, wherein the first priority alarm result has a higher priority than the second priority alarm result.

[0039] Preferably, the maintenance personnel list information is obtained, which includes personnel who inspect and troubleshoot various software when the computer under test experiences abnormal alarms;

[0040] Based on the maintenance personnel list information, when the first priority alarm result is received, the first type of alarm operation data and the second type of alarm operation data are sequentially sent to the mobile terminals of the corresponding maintenance personnel.

[0041] Based on the alarm priority results, the alarm operation data of the first type, the alarm operation data of the second type, the alarm operation data of the third type, the alarm operation data of the fourth type, and other operation data are sequentially sent to the central storage module in the backend system for storage.

[0042] In summary, this application includes at least one of the following beneficial technical effects:

[0043] 1. By collecting data from the resource layer and application layer of the computer software under test, software operation parameter information is obtained. Based on the software model, the operation parameter data pool of each software is obtained. Based on the software process, the operation parameter relationship chain of each software is further arranged. The dependency relationship between software is obtained by multi-dimensional analysis of the dependency relationship between the computer under test in the non-running state and the dependency relationship in the running state. A dynamic threshold dataset is generated by the historical operation alarm information of the computer under test. The software operation parameter information is compared with the dynamic threshold dataset to determine whether there are any abnormalities in each operation data in the software operation parameter information. Combined with the computer software relationship network, the operation data that needs to be alarmed is classified into levels to obtain the alarm priority result. The alarm is sent to the backend system and stored in a hierarchical manner, which improves the efficiency of real-time monitoring of computer software operation status.

[0044] 2. By learning historical alarm patterns through machine learning algorithms, dynamic thresholds are automatically generated to identify truly abnormal alarms that deviate from normal patterns, greatly reducing false alarms and improving alarm detection accuracy. This, in turn, improves the accuracy of monitoring computer software operating status and further enhances the efficiency of real-time monitoring of computer software operating status. Attached Figure Description

[0045] Figure 1 This embodiment is a flowchart illustrating the method for real-time monitoring of computer software operation status. Detailed Implementation

[0046] The present application will be further described in detail below with reference to the accompanying drawings.

[0047] This application discloses a method for real-time monitoring of the running status of computer software.

[0048] A method for real-time monitoring of the running status of computer software includes the following steps:

[0049] Reference Figure 1 Step S1 involves hierarchical data collection of the resource layer and application layer of the computer software under test using a preset collection method to obtain software operating parameter information. Step S1 specifically includes the following sub-steps:

[0050] Step S11: The monitoring objects of the computer under test include resource layer monitoring objects and application layer monitoring objects.

[0051] Step S12: The resource layer monitoring objects include CPU utilization, memory usage, disk I / O, network traffic, and number of processes.

[0052] Step S13: The application layer monitoring objects include metrics, logs, and links.

[0053] Among these, metrics refer to QPS (QPS is a common metric for measuring search traffic received by an information retrieval system, such as a search engine or database, within one second), request response time, error rate, and business metrics (such as order count and payment success rate). Logs refer to the logs output by the application (such as Info, Debug, Warn, and Error logs). The path refers to the complete path and performance data of a single request flowing through various microservices in a distributed system.

[0054] Step S14: Obtain the Sidecar agent of the service mesh (such as Istio) and establish a signal connection link with the computer under test. Based on the Sidecar agent of the service mesh (such as Istio), detect and collect resource layer monitoring objects to obtain resource layer monitoring data information.

[0055] Step S15: Install an agent (such as Telegraf, Datadog Agent, PrometheusNode Exporter) on the computer under test to collect metric information and log information.

[0056] Step S16: Integrate the SDK into the computer's software code to collect and report indicator information (business indicators) and link information (such as using Micrometer, OpenTelemetry).

[0057] Step S17: Resource layer monitoring data, indicator information, log information, and link information are combined to form software operation parameter information.

[0058] Reference Figure 1 Step S2 involves classifying the software operating parameter information based on the software model to obtain an operating parameter data pool for each software. Then, based on the software operating flow, the operating data in each software's operating parameter data pool is arranged to obtain the operating parameter relationship chain for each software. Step S2 specifically includes the following sub-steps:

[0059] Step S21: Number each software in the computer under test to obtain software feature numbers.

[0060] Step S22: Based on the software model corresponding to each running data in the software running parameter information, classify each running data in the software running parameter information to obtain running parameter classification information. The running parameter classification information includes the running parameter data pool of each software.

[0061] Step S23: Based on the operation flow of each software in the computer under test, the operation data in the operation parameter data pool of each software is arranged in parallel and sequentially to obtain the operation parameter relationship chain of each software.

[0062] Reference Figure 1 Step S3 involves multi-dimensional analysis of the dependencies between software programs in the computer under test when they are not running to obtain static software dependencies, and multi-dimensional analysis of the dependencies between software programs in the computer under test when they are running to obtain dynamic software dependencies. These dependencies are then fused to obtain a software dependency network, and combined with the runtime parameter relationship chain to construct a computer software relationship network. Step S3 specifically includes the following sub-steps:

[0063] Step S31: Determine the first static dependency relationship based on the location of each software. In actual use, a software's startup script or service configuration may explicitly specify that other services it depends on must be started first.

[0064] Secondary static dependencies are identified based on the configuration files of each software. In practice, many software programs explicitly configure their dependencies in configuration files (such as .ini, .yml, .xml, .conf).

[0065] During the installation process of each software, the installation status of newly added dependency libraries is identified to obtain the third static dependency relationship.

[0066] Step S32: The first static dependency, the second static dependency, and the third static dependency are combined to form the software static dependency. Based on the software static dependency network, the degree of association between each software and other software when the software is not running is determined to obtain the software static association coefficient.

[0067] Step S33: Based on the system tools of the computer under test, obtain the association between each software process and other software during runtime to obtain the first dynamic dependency relationship.

[0068] Specifically, when the computer under test is a Windows system, you can view process IDs (PIDs), CPU usage, memory usage, etc., through Task Manager, or view dependencies through Resource Monitor. For example, under the "CPU" tab, expanding a process will show its associated handles (open files, registry keys, dynamic link libraries DLLs), and under the "Network" tab, you can see its network connections. When the computer under test is a Linux / macOS system, you can use the pstree command to directly display process relationships in a tree diagram, or use the strace / dtrace commands to trace all system calls executed by the process (such as file operations and network communications), thereby accurately understanding the interaction dependencies with other software.

[0069] Step S34: Monitor the network connection of the computer under test, identify whether there are listening ports between the software, and if there are listening ports, determine that there is a dependency relationship and output the second dynamic dependency relationship.

[0070] Step S35: The first dynamic dependency relationship and the second dynamic dependency relationship are combined to form a software dynamic dependency relationship. Based on the software dynamic dependency relationship network, the degree of association between each software and other software when the software is not running is judged to obtain the software dynamic association coefficient.

[0071] Step S36: Add the static correlation coefficient between each software program to the corresponding dynamic correlation coefficient to obtain the correlation coefficient between the software programs.

[0072] Step S37: The combination of static software dependencies and dynamic software dependencies forms a software dependency network.

[0073] Step S38: Construct a computer software relationship network based on the inter-software dependency network and the operational parameter relationship chain of each software.

[0074] Reference Figure 1 Step S4 involves generating a dynamic threshold dataset using the historical alarm information of the computer under test. Step S4 specifically includes the following sub-steps:

[0075] Step S41: Obtain historical operational alarm information of the computer under test. The historical operational alarm information includes historical alarm data type, historical alarm timestamp, historical alarm software type, and historical operational alarm severity.

[0076] Step S42: Calculate the historical alarm information of the computer under test based on machine learning algorithm, learn the historical pattern of the alarm of the computer under test software operation, make predictions based on time series, and generate a dynamic threshold dataset. The dynamic threshold dataset includes the dynamic threshold of each running data in the software running parameter information.

[0077] In practical applications, alarms based on fixed thresholds are difficult to adapt to fluctuations in business traffic (such as day and night, weekdays and holidays), which can easily lead to a large number of false alarms or missed alarms. In this embodiment, machine learning algorithms are used to learn historical alarm patterns, automatically generate dynamic thresholds, identify abnormal alarms that truly deviate from the normal pattern, greatly reduce false alarms, improve alarm detection accuracy, and thus improve the accuracy of monitoring the operating status of computer software.

[0078] Reference Figure 1Step S5 involves comparing the software operating parameter information with the dynamic threshold dataset to determine if any abnormalities exist in the operating data. Furthermore, it uses the computer software relationship network to classify the operating data requiring alarms into different levels, resulting in alarm priority. Step S5 specifically includes the following sub-steps:

[0079] Step S51: Compare each running data in the software running parameter information with each dynamic threshold in the dynamic threshold dataset to determine whether there is any abnormality in each running data in the software running parameter information. If there is an abnormality, mark the running data as a type of alarm running data, mark the software containing the type of alarm running data as abnormal software, and output a first-level alarm signal.

[0080] Specifically, when a certain running data in the software running parameter information is greater than the corresponding dynamic threshold in the dynamic threshold dataset, it is determined that the running data is abnormal and is marked as a type of alarm running data.

[0081] Step S52: Based on the operating parameter relationship chain of each software, mark the operating data other than the first type of alarm operating data in the operating parameter relationship chain where the first type of alarm operating data is located as the second type of alarm operating data, and output the second-level alarm signal.

[0082] Step S53: Based on the software dependency network, mark the software that is associated with the abnormal software as abnormal associated software.

[0083] Step S54: Based on the software correlation coefficient, software with a correlation degree greater than the preset software correlation coefficient threshold is marked as high-risk abnormal software. The running data in the running parameter relationship chain of the high-risk abnormal software is marked as three types of alarm running data, and three types of alarm signals are output.

[0084] Step S55: Based on the software correlation coefficient, software with a correlation degree less than or equal to the preset software correlation coefficient threshold is marked as low-risk abnormal software. The running data in the running parameter relationship chain of the low-risk abnormal software is marked as four types of alarm running data, and four types of alarm signals are output.

[0085] Step S56: The combination of type 1 alarm operation data and type 2 alarm operation data forms a first priority alarm result, wherein the priority of type 1 alarm operation data is higher than that of type 2 alarm operation data. The combination of type 3 alarm operation data and type 4 alarm operation data forms a second priority alarm result, wherein the priority of type 3 alarm operation data is higher than that of type 4 alarm operation data. The combination of the first priority alarm result and the second priority alarm result forms an alarm priority result, wherein the priority of the first priority alarm result is higher than that of the second priority alarm result.

[0086] In practical applications, when a low-level fault occurs, it may trigger alarms from hundreds of related services, drowning out truly useful information and causing alarm fatigue for operation and maintenance personnel. In this embodiment, by automatically classifying alarm priorities and performing intelligent alarm noise reduction, the efficiency of real-time monitoring and alarming of the operating status of the computer software under test is improved.

[0087] Reference Figure 1 Step S6 involves sending the software operating parameter information to the backend system in a tiered manner based on the alarm priority results and storing it. Step S6 specifically includes the following sub-steps:

[0088] Step S61: Obtain the list of maintenance personnel. The list of maintenance personnel includes the personnel who inspect and troubleshoot the software when the computer under test issues an abnormal alarm.

[0089] Step S62: Based on the maintenance personnel list information, when the first priority alarm result is received, the first type of alarm operation data and the second type of alarm operation data are sequentially sent to the mobile terminals of the corresponding maintenance personnel.

[0090] Step S63: Based on the alarm priority results, the alarm operation data of type 1, type 2, type 3, type 4, and other operation data are sequentially sent to the central storage module in the backend system for storage. It should be noted that in this embodiment, a high-throughput message queue (such as Kafka, RabbitMQ) is used as a buffer to prevent data surges from overwhelming the backend system.

[0091] Those skilled in the art will understand that embodiments of the present invention can be provided as methods or systems. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0092] This invention is described with reference to flowchart illustrations and structural diagrams of methods and systems according to embodiments of the invention. It should be understood that the combination of each process and module in the flowchart and structural diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing device, generate instructions for implementing the process. Figure 1 One or more processes and structures Figure 1 A device for a function specified in one or more modules.

[0093] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and structures Figure 1 The function specified in one or more modules.

[0094] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and structures Figure 1 The steps of a specified function in one or more modules.

[0095] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0096] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for real-time monitoring of the running status of computer software, characterized in that, Includes the following steps: Step S1: Based on a preset acquisition method, hierarchical data acquisition is performed on the resource layer and application layer of the computer software under test to obtain software operation parameter information. Step S2: Classify the software operation parameter information based on the software model to obtain the operation parameter data pool of each software, and arrange the operation data in the operation parameter data pool of each software based on the software operation process to obtain the operation parameter relationship chain of each software. Step S3: Analyze the dependencies between software in the computer under test in the non-running state to obtain the static software dependencies, and analyze the dependencies between software in the running state to obtain the dynamic software dependencies. Combine these to obtain the software dependency network, and construct the computer software relationship network by combining the running parameter relationship chain. Step S4: Generate a dynamic threshold dataset using the historical alarm information of the computer under test; Step S5: Compare the software running parameter information with the dynamic threshold dataset to determine whether there are any abnormalities in the running data of the software running parameter information, and combine the computer software relationship network to classify the running data that need to be alarmed to obtain the alarm priority result. Step S6: Based on the alarm priority results, the software operation parameter information is sent to the backend system in a hierarchical manner and stored.

2. The method for real-time monitoring of computer software operating status according to claim 1, characterized in that, Step S1 specifically includes the following sub-steps: The monitoring objects of the computer under test include resource layer monitoring objects and application layer monitoring objects; The resource layer monitoring objects include CPU utilization, memory usage, disk I / O, network traffic, and number of processes; The application layer monitoring objects include metrics, logs, and links; Obtain the service mesh's Sidecar agent and establish a signal connection link with the computer under test. Based on the service mesh's Sidecar agent, detect and collect resource layer monitoring objects to obtain resource layer monitoring data information. An agent is installed on the computer under test to collect indicator information and log information. An SDK is integrated into the computer's software code to collect and report indicator information and link information; The resource layer monitoring data, indicator information, log information, and link information are combined to form software operation parameter information.

3. The method for real-time monitoring of computer software operating status according to claim 2, characterized in that, Step S2 specifically includes the following sub-steps: Each piece of software in the computer under test is assigned a number to obtain a software feature number; Based on the software model corresponding to each running data in the software running parameter information, the running data in the software running parameter information is classified to obtain running parameter classification information, which includes the running parameter data pool of each software. Based on the operation flow of each software in the computer under test, the operation data in the operation parameter data pool of each software are arranged in parallel and sequentially to obtain the operation parameter relationship chain of each software.

4. The method for real-time monitoring of computer software operating status according to claim 3, characterized in that, Step S3 specifically includes the following sub-steps: The first static dependency is determined based on the location of each software, the second static dependency is identified based on the configuration file of each software, and the third static dependency is obtained by detecting the installation of newly added dependency libraries during the installation process of each software. The first static dependency, the second static dependency, and the third static dependency are combined to form a software static dependency. Based on the software static dependency network, the degree of association between each software and other software when the software is not running is determined to obtain the software static association coefficient. Based on the system tools of the computer under test, the correlation between each software process and other software during runtime is obtained to obtain the first dynamic dependency relationship; The network connection of the computer under test is monitored to identify whether there are listening ports between the software. If there are listening ports, it is determined that there is a dependency relationship and the second dynamic dependency relationship is output.

5. The method for real-time monitoring of computer software operating status according to claim 4, characterized in that, Step S3 further includes the following sub-steps: The first dynamic dependency and the second dynamic dependency are combined to form a software dynamic dependency. Based on the software dynamic dependency network, the degree of association between each software and other software when the software is not running is determined to obtain the software dynamic association coefficient. The software correlation coefficient is obtained by adding the static correlation coefficient between each software to the corresponding dynamic correlation coefficient. The combination of static and dynamic software dependencies forms a network of inter-software dependencies; A computer software relationship network is constructed based on the inter-software dependency network and the relationship chain of the running parameters of each software.

6. The method for real-time monitoring of computer software operating status according to claim 5, characterized in that, Step S4 specifically includes the following sub-steps: Obtain historical operational alarm information of the computer under test, including historical alarm data type, historical alarm timestamp, historical alarm software type, and historical operational alarm severity. The historical alarm information of the computer under test is calculated based on machine learning algorithms to learn the historical patterns of the alarms in the computer software operation. Based on time series, prediction is performed to generate a dynamic threshold dataset, which includes the dynamic threshold of each running data in the software operation parameter information.

7. The method for real-time monitoring of computer software operating status according to claim 6, characterized in that, Step S5 specifically includes the following sub-steps: Each running data in the software running parameter information is compared with each dynamic threshold in the dynamic threshold dataset to determine whether there is any abnormality in each running data in the software running parameter information. If there is an abnormality, the running data is marked as a type of alarm running data, the software containing the type of alarm running data is marked as abnormal software, and a first-level alarm signal is output. Based on the operational parameter relationship chain of each software, the operational data other than the operational data of the first type of alarm in the operational parameter relationship chain are marked as the second type of alarm operational data, and a second-level alarm signal is output. Based on the software dependency network, software that is associated with abnormal software is marked as abnormal associated software; Based on the software correlation coefficient, software with a correlation degree greater than the preset software correlation coefficient threshold is marked as high-risk abnormal software. The running data in the running parameter relationship chain of high-risk abnormal software is marked as three types of alarm running data, and three types of alarm signals are output. Based on the software correlation coefficient, software with a correlation degree less than or equal to the preset software correlation coefficient threshold is marked as low-risk abnormal software. The running data in the running parameter relationship chain of low-risk abnormal software is marked as four types of alarm running data, and four types of alarm signals are output. The combination of the first type of alarm operation data and the second type of alarm operation data forms a first priority alarm result, wherein the first type of alarm operation data has a higher priority than the second type of alarm operation data. The combination of the third type of alarm operation data and the fourth type of alarm operation data forms a second priority alarm result, wherein the third type of alarm operation data has a higher priority than the fourth type of alarm operation data. The combination of the first priority alarm result and the second priority alarm result forms an alarm priority result, wherein the first priority alarm result has a higher priority than the second priority alarm result.

8. The method for real-time monitoring of computer software operating status according to claim 7, characterized in that, Step S6 specifically includes the following sub-steps: Obtain the list of maintenance personnel, which includes personnel who inspect and troubleshoot various software programs when the computer under test experiences abnormal alarms. Based on the maintenance personnel list information, when the first priority alarm result is received, the first type of alarm operation data and the second type of alarm operation data are sequentially sent to the mobile terminals of the corresponding maintenance personnel. Based on the alarm priority results, the alarm operation data of the first type, the alarm operation data of the second type, the alarm operation data of the third type, the alarm operation data of the fourth type, and other operation data are sequentially sent to the central storage module in the backend system for storage.

Images