Off-line management method and system for software service operation and maintenance information of airport production operation system

Abstract

The invention belongs to the technical field of software service operation and maintenance, and discloses an airport production operation system software service operation and maintenance information offline management method and system. According to the method, an asset management platform is constructed based on java, project information and server information are maintained by using an improved information anomaly identification method, and a production operation system is provided for different middleware covered by the production operation system. Installation directories, port numbers and version number information of scripts corresponding to different middleware are obtained by searching environment variables and a service obtaining mode through a PowerShell script or a Bash script, script files of the different middleware are automatically generated and transmitted to a server side, and data are transmitted back to an asset management platform in a unified mode; and knowledge base maintenance and abstraction are carried out to form corresponding technical processing flows. According to the invention, a perfect informatization management system is established, so that fault processing project personnel can quickly know project conditions and solve project faults.

Description

Technical Field

[0001] The present invention belongs to the technical field of software service operation and maintenance, and particularly relates to an offline management method and system for software service operation and maintenance information of an airport production operation system. Background Art

[0002] Due to network security considerations, the software system services of each production company are deployed in the local intranet system and cannot be uniformly operationally managed through the Internet or cloud services. As a result, the system deployment information and management method records in various places are scattered and lack standardized management, thereby reducing the efficiency of after-sales problem handling.

[0003] Through the above analysis, the problems and defects existing in the prior art are as follows: the offline maintenance of project assets in various places is scattered, lacking centralized management means, and manual maintenance is time-consuming and laborious; the projects in various places lack a perfect management system, and the operation and maintenance guarantee efficiency is low. Summary of the Invention

[0004] To overcome the problems existing in the related art, the disclosed embodiments of the present invention provide an offline management method and system for software service operation and maintenance information of an airport production operation system.

[0005] The technical solutions are as follows: An offline management method for software service operation and maintenance information of an airport production operation system, including:

[0006] S1, Asset Management: Build an asset management platform based on Java, use an improved information anomaly recognition method to maintain project information and server information. For different middleware covered by the production operation system, through PowerShell scripts or Bash scripts, use the method of finding environment variables and services to obtain the installation directory, port number, and version number information of the corresponding scripts for different middleware, automatically generate script files for different middleware, and transmit them to the server side, and uniformly send the data back to the asset management platform;

[0007] S2, Knowledge Base Maintenance: Based on the service information deployed on the asset management platform, use the fault troubleshooting knowledge base to maintain the data of the problem handling process for each project;

[0008] S3, Abstract each fault scenario into a corresponding technical processing process.

[0009] In step S1, the project information includes: project name, related faulty project, maintenance status;

[0010] The server information includes IP, virtual IP, operating system version, and system running time.

[0011] In step S1, building an asset management platform based on Java and using an improved information anomaly recognition method to maintain project information and server information includes:

[0012] The improved information anomaly recognition method includes: an information anomaly recognition method that first recognizes the normal project server information signal and then recognizes the abnormal project server information signal, and an information anomaly recognition method that first recognizes the abnormal project server information signal and then recognizes the normal project server information signal;

[0013] (1) When and hold simultaneously, consider the recognition of abnormal project server information. Among them, When they cannot hold simultaneously, adopt the single recognition mode of normal project server information;

[0014] In the formula, is the maximum transmission power of the radio frequency transmission module of the normal project server information, χ Z is the total amount of the target rate superposition of the normal project server information, is the maximum transmission power of the project server information recognition node, k is the order of derivation, H Z is the target rate of the normal project server information, f Z,H is the link power gain between the radio frequency transmission module of the normal project server information and the project server information recognition node, is the Gaussian white noise of the power spectral density, f H,Z is the link power gain between the project server information recognition node and the receiving module of the normal project server information;

[0015] (2) When and hold, and and hold simultaneously, the transmission powers used by the radio frequency transmission modules of the normal project server information and the abnormal project server information in the first discrimination frequency band are respectively: At the project server information recognition node, first recognize the normal project server information signal and then recognize the abnormal project server information signal;

[0016] In the formula, is the maximum transmission power of the radio frequency transmission module of the abnormal project server information, χ Y is the total amount of the target rate superposition of the abnormal project server information, f Y,H is the link power gain between the radio frequency transmission module of the abnormal project server information and the project server information recognition node;

[0017] The powers allocated by the project server information recognition node in the second discrimination frequency band to the abnormal project server information and the normal project server information signals are respectively: At the normal project server information receiving module, the abnormal project server information signal is directly recognized as noise to identify its own signal. At the abnormal project server information receiving module, the abnormal project server information receiving module first identifies the normal project server information signal, deletes it from the aliased signal, and then identifies its own signal.

[0018] In step S1, the data is uniformly sent back to the asset management platform, including: for airports with tens of millions of passengers, the script file is regularly sent back to the asset management platform in the form of a websocket interface through the established secure link; for airports with less than tens of millions of passengers, the installation directory, port number, and version number information of different middleware received by the server are generated into an excel according to the standard template, and the excel is regularly downloaded to the local and then imported into the asset management platform; the standard template includes running projects, IP, service names, versions, ports, installation directories, startup methods, and process names.

[0019] The different middleware includes Mysql, Oracle, Jboss, ActiveMq, Tomcat, Redis, Nginx, Kettle, FTP, JDK, HTTPD; for Jboss, the installation directory, port number, and version number information of the corresponding script are obtained by searching for environment variables and service acquisition methods through PowerShell scripts or Bash scripts.

[0020] Jboss obtains the installation directory, port number, and version number information of the corresponding script by searching for environment variables and service acquisition methods through PowerShell scripts or Bash scripts, including:

[0021] Search for environment variables and services through the following formula to generate candidate solutions:

[0022]

[0023] s = rand * peri

[0024] where b is the current search iteration algebra for searching environment variables and services, is the d-th environment variable and service dimension element of the i-th solution at iteration b + 1, is the d-th environment variable and service dimension element of the r1-th solution at iteration b, It is the d-th environmental variable and service dimension element of the s2-th solution in iteration b; s is a random number in the environmental variable and service lookup, peri is set to 2 - 20, s1 is a random number representing a random individual in the domain 1 of the environmental variable and service set lookup, and s2 is a random number representing a random individual in the domain 2 of the environmental variable and service set lookup; the candidate solution includes the accurate solution corresponding to the installation directory, port number, and version number information of the script, and the environmental variable and service include the change information of the project information and server information.

[0025] In step S2, the deployed service information includes service name, deployment directory, version, startup method, backup information, and log information;

[0026] Each project problem includes: memory leak, CPU overload, exhaustion of database and middleware connection pools, performance degradation, abnormal crash, upstream and downstream system data problems, upstream and downstream system link problems, resource leak, or cache invalidation;

[0027] Furthermore, CPU overload includes: monitoring CPU utilization rate, analyzing thread occupancy, finding resource competition and blocking, analyzing memory usage, using tools to troubleshoot time-consuming code, and reviewing code logic.

[0028] Furthermore, finding resource competition and blocking includes:

[0029] In the first step, for resource competition detection, the competition result obtained by the resource competition mechanism is processed using the method of quadratic Fourier transform accumulation. Fourier transform is performed by sliding backward at intervals that are not integer multiples of the non-link transmission frequency domain code length; then a judgment threshold is set for troubleshooting. If the maximum value output by the Fourier transform is greater than the threshold, the troubleshooting is successful;

[0030] In the second step, for resource blocking, the method of finding the maximum frequency domain wave peak using a sliding window is adopted. Fourier transform is performed by sliding at intervals starting from two phase points before and after the troubleshooting position, with 1 Fourier transform performed each time; then the maximum value of the frequency domain peaks output by these 8 Fourier transforms and its corresponding position are found to determine the resource blocking position; the starting points of the 8 slides of the Fourier transform window are respectively at two phase points before and after the troubleshooting position. The interval of the first slide of the Fourier transform window is Q×H - 2, the interval of the second slide is Q×H + 1, the interval of the third slide is Q×H + 2, ……, the interval of the eighth slide is Q×H + 6, where Q is the length of the link transmission frequency domain code and H is the upsampling multiple.

[0031] In the first step, the resource competition mechanism includes:

[0032] (A) Competition constraint: Establish the best function, and the expression is:

[0033]

[0034] where l is the vector representation of the project information obtained from the new scenario of the system, and Z is the matrix composed of the training samples used by the model. is the model parameter, and ξ i is the competition weight defined for the i-th project sample, and Z -i is the matrix composed of the samples excluding the i-th faulty project in the training data, and O -i is the decoding coefficient, and E is the total number of faulty projects included in the entire training sample.

[0035] (B) Set the competition weight, and the expression is:

[0036]

[0037] where g max is the maximum value of the distance between the link transmission resource sample and the project sample, and g i is the Euclidean distance between the link transmission resource sample and the i-th project sample, and θ is the weight coefficient.

[0038] (C) Establish an optimization link transmission model to obtain the optimal decoding result of the link transmission resource sample.

[0039] In the second step, the resource blockage adopts the method of finding the maximum frequency domain wave peak by sliding window, including:

[0040] (1) The CPU first performs 8-phase quantization on the received link transmission frequency domain signal, maps the signals in different transmission path ranges to 8 phases respectively, and then identifies the relevant values for the quantized signal; the relevant values are obtained by the following formula:

[0041]

[0042] where corr(j) is the identified relevant value, F is the link transmission frequency domain signal after 8-phase quantization, H PN is the link transmission frequency domain code, k is the sampling point position of the received link transmission frequency domain signal F, and j is the identification sliding position.

[0043] (2) For the identified relevant values, adopt the method of quadratic Fourier transform accumulation; the Fourier transform window slides backward at intervals of Q×H - 2 sampling points, and performs 1 Fourier transform each time it slides.

[0044] (3) Set the decision threshold σ 0 according to the peak value of the Fourier transform output sequence, and perform troubleshooting judgment; when the accumulated peak value is greater than the decision threshold σ 0 , it is considered that the troubleshooting is successful, and the troubleshooting position is recorded at the same time; otherwise, the Fourier transform window slides backward once and repeats the above steps.

[0045] Another object of the present invention is to provide an offline management system for operation and maintenance information of airport production operation system software services. This system is implemented by the above-mentioned offline management method for operation and maintenance information of airport production operation system software services. The system includes:

[0046] The asset management module constructs an asset management platform based on Java, uses an improved information anomaly recognition method to maintain project information and server information. For different middleware covered by the production operation system, through PowerShell scripts or Bash scripts, it uses the method of finding environment variables and services to obtain the installation directory, port number, and version number information of the scripts corresponding to different middleware, automatically generates script files for different middleware, and transmits them to the server side, and uniformly transmits the data back to the asset management platform;

[0047] The knowledge base maintenance module maintains the data of the problem handling process of each project by using the fault troubleshooting knowledge base based on the service information deployed on the asset management platform;

[0048] The abstraction module is used to abstract each fault scenario into a corresponding technical processing process.

[0049] Combining all the above technical solutions, the beneficial effects of the present invention are as follows: Through the centralized management of information such as airports, projects, servers, and deployment information, the present invention establishes a perfect information management system, which is convenient for the fault project personnel to quickly understand the project situation and solve project faults. On the aspect of fault project personnel, a unified guarantee team is established to conduct unified operation and maintenance management of multiple projects responsible for different regions, and the problem handling efficiency is improved through unified standardized process management.

[0050] The present invention establishes an after-sales asset management system to establish a ledger for after-sales projects, servers, deployment information, key asset startup cycles, data flows, etc., and conducts unified maintenance. When problems occur in each project, the unified operation and maintenance fault project personnel perform quick troubleshooting. If they encounter relatively complex technical problems, they contact technical experts and provide relevant information in the asset management system. The technical experts quickly understand the project situation and conduct fault handling. At the same time, some routine inspection reminders such as server timed restart and password timed modification are also maintained through the asset management system and timed reminders are given to prevent service failures caused by forgetting. The present invention provides a method for conveniently collecting and maintaining relevant deployment information for the architecture of airport production operation systems, improving the centralized maintenance and operation and maintenance capabilities of operation and maintenance guarantee information. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure;

[0052] Figure 1It is a flowchart of the method for offline management of operation and maintenance information of airport production operation system software services provided by an embodiment of the present invention;

[0053] Figure 2 It is a schematic diagram of the offline management system for operation and maintenance information of airport production operation system software services provided by an embodiment of the present invention;

[0054] In the figure: 1. Asset management module; 2. Knowledge base maintenance module; 3. Abstraction module. Specific embodiments

[0055] To make the above objects, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to fully understand the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0056] Embodiment 1, as Figure 1 shown, the method for offline management of operation and maintenance information of airport production operation system software services provided by an embodiment of the present invention includes:

[0057] S1. Asset management, building an asset management platform based on Java, using an improved information anomaly recognition method to maintain project information and server information. For different middleware covered by the production operation system, through PowerShell scripts or Bash scripts, the installation directory, port number, and version number information of the corresponding scripts of different middleware are obtained by searching for environment variables and services, and script files of different middleware are automatically generated and transmitted to the server side, and the data is uniformly transmitted back to the asset management platform;

[0058] In maintaining project information, due to the offline environment, project and server information are maintained manually.

[0059] Project information includes project name, related failure project, and maintenance status; the content of server information maintenance includes IP, virtual IP, operating system version, and system running time; the deployed service information includes service name, deployment directory, version, startup method, backup information, and log information; the standard template includes templates for running projects, IP, service name, version, port, installation directory, startup method, and process name;

[0060] In step S1, the project information includes: project name, related failure project, and maintenance status;

[0061] Server information includes IP, virtual IP, operating system version, and system running time.

[0062] Build an asset management platform based on Java, and use an improved information anomaly recognition method to maintain project information and server information, including:

[0063] The improved information anomaly recognition method includes: an information anomaly recognition method that first recognizes the normal project server information signal and then recognizes the abnormal project server information signal, and an information anomaly recognition method that first recognizes the abnormal project server information signal and then recognizes the normal project server information signal;

[0064] (1) When and hold simultaneously, consider the recognition of abnormal project server information, where When they cannot hold simultaneously, adopt the separate recognition mode of normal project server information;

[0065] In the formula, is the maximum transmission power of the normal project server information radio frequency transmission module, X Z is the total amount of the target rate superposition of the normal project server information, is the maximum transmission power of the project server information recognition node, k is the order of derivation, H Z is the target rate of the normal project server information, f Z,H is the link power gain between the normal project server information radio frequency transmission module and the project server information recognition node, is the Gaussian white noise of the power spectral density, f H,Z is the link power gain between the project server information recognition node and the normal project server information receiving module;

[0066] (2) When and hold, and and hold simultaneously, the transmission powers used by the normal project server information radio frequency transmission module and the abnormal project server information radio frequency transmission module in the first discrimination frequency band are respectively: At the project server information recognition node, first recognize the normal project server information signal and then recognize the abnormal project server information signal;

[0067] In the formula, is the maximum transmission power of the abnormal project server information radio frequency transmission module, χ Y is the total amount of the target rate superposition of the abnormal project server information, f Y,H is the link power gain between the abnormal project server information radio frequency transmission module and the project server information recognition node;

[0068] In the second discrimination frequency band, the power allocated to the abnormal project server information and the normal project server information signals by the project server information recognition node is as follows: At the normal project server information receiving module, the abnormal project server information signal is directly regarded as noise to identify its own signal. At the abnormal project server information receiving module, the abnormal project server information receiving module first identifies the normal project server information signal, deletes it from the aliased signal, and then identifies its own signal.

[0069] Through the above innovative method, an asset management platform based on Java uses the improved information anomaly recognition method to maintain the security of project information and server information.

[0070] In step S1, the data is uniformly transmitted back to the asset management platform, including: for airports with tens of millions of passengers, the script file is regularly transmitted back to the asset management platform in the form of a websocket interface through the established secure link; for airports with less than tens of millions of passengers, the installation directories, port numbers, and version numbers of different middleware received by the server are generated into an excel according to the standard template, and the excel is regularly downloaded to the local and then imported into the asset management platform; the standard template includes running projects, IPs, service names, versions, ports, installation directories, startup methods, and process names.

[0071] The different middleware includes Mysql, Oracle, Jboss, ActiveMq, Tomcat, Redis, Nginx, Kettle, FTP, JDK, HTTPD; Jboss uses PowerShell scripts or Bash scripts to find the environment variables and service acquisition methods to obtain the installation directory, port number, and version number information of the corresponding scripts.

[0072] Jboss uses PowerShell scripts or Bash scripts to find the environment variables and service acquisition methods to obtain the installation directory, port number, and version number information of the corresponding scripts, including:

[0073] The following formula is used to find the environment variables and services to generate candidate solutions:

[0074]

[0075] S = rand * peri

[0076] where b is the current search iteration algebra for finding the environment variables and services, is the d-th environment variable and service dimension element of the i-th solution at iteration b + 1, is the d-th environment variable and service dimension element of the r1-th solution at iteration b, It is the d-th environmental variable and service dimension element of the s2-th solution in iteration b; s is a random number in the environmental variable and service search, peri is set to 2 - 20, s1 is a random number representing a random individual in the search for the environmental variable and service set domain 1, and s2 is a random number representing a random individual in the search for the environmental variable and service set domain 2; the candidate solution includes the accurate solution corresponding to the installation directory, port number, and version number information of the script, and the environmental variable and service include the change information of project information and server information.

[0077] Through the above innovative method, the installation directory, port number, and version number information of the corresponding script can be accurately obtained.

[0078] In another example, according to the characteristics of the airport production operation system, through batch processing scripts such as powdershell and bash, the deployment information of 11 middleware covered by the production operation system is obtained in an automated manner and synchronously transmitted to the unified operation and maintenance management system, that is, the asset management platform, at regular intervals.

[0079] For the 11 middleware (Mysql, Oracle, Jboss, ActiveMq, Tomcat, Redis, Nginx, Kettle, FTP, JDK, HTTPD) covered by the production operation system, through specific PowerShell scripts or Bash scripts, information such as the installation directory, port number, and version number of the corresponding script is obtained, and the information is automatically generated into a script file or transmitted to a specific server, and the data is uniformly sent back to the asset management platform.

[0080] Mysql obtains information such as the installation directory, port number, and version number of the corresponding script through the method of searching for environmental variables and services, Oracle obtains information such as the installation directory, port number, and version number of the corresponding script through the method of searching for environmental variables and services, and Jboss obtains information such as the installation directory, port number, and version number of the corresponding script through the method of searching for environmental variables and services;

[0081] For example: Taking the Jboss system as an example, a PowerShell script or a Bash script first queries whether there is a service related to the jboss keyword in the system. If it exists, it finds the corresponding installation directory according to the service, and checks the configuration file at a specific location in the directory to view information such as the port number and version number used in the configuration file. If it cannot be found through the service, it executes the java process query command to find the jboss-related process name. After finding it, it finds the corresponding installation directory, and checks the configuration file at a specific location in the directory to view information such as the port number and version number used in the configuration file. The queried information supports two ways of feedback. For large airports with tens of millions of passengers, a secure specific link is opened, and the query information is regularly (every 10 minutes) sent back to the asset management platform in the form of a websocket interface. If it is an airport with less than tens of millions of passengers, an excel is generated according to the standard template, and the excel is regularly downloaded to the local and then imported into the asset management platform; the standard template includes running projects, IP, service names, versions, ports, installation directories, startup methods, and process names.

[0082] ActiveMq is obtained by means of finding services and querying the process on port 61616;

[0083] Tomcat obtains information such as the installation directory, port number, and version number of the corresponding script by means of finding environment variables and querying java processes;

[0084] Redis obtains information such as the installation directory, port number, and version number of the corresponding script by means of finding environment variables and querying port numbers;

[0085] Nginx obtains information such as the installation directory, port number, and version number of the corresponding script by means of querying port numbers and process names;

[0086] Kettle obtains information such as the installation directory, port number, and version number of the corresponding script by means of querying process names;

[0087] FTP obtains information such as the installation directory, port number, and version number of the corresponding script by means of querying port numbers and process names;

[0088] JDK obtains information such as the installation directory, port number, and version number of the corresponding script by means of querying environment variables;

[0089] HTTPD obtains information such as the installation directory, port number, and version number of the corresponding script by means of querying services, port numbers, and process names.

[0090] S2, knowledge base maintenance. Based on the service information deployed on the asset management platform, the process data of problem handling for each project is maintained using the fault troubleshooting knowledge base;

[0091] When similar problems occur in other projects, quickly retrieve relevant case handling solutions through the knowledge base to improve the efficiency of fault handling; start from the fault scenarios of each project and maintain the case handling processes corresponding to each project version.

[0092] In step S2, the problems of each project include: memory leak, CPU overload, exhaustion of database and middleware connection pools, performance degradation, abnormal crash, upstream and downstream system data problems, upstream and downstream system link problems, resource leak or cache invalidation.

[0093] Classify according to project characteristics and technical architectures, record various fault scenarios and handling processes encountered in each project respectively, and add corresponding tags according to project characteristics (integration, mobile, esb, reports, etc.) and technical architectures (java background CS version, BS version, fusion version,.net version, etc.).

[0094] Abstract the fault scenarios into two major categories: general class faults and troubleshooting and handling processes for each system. The general class faults are mainly divided into troubleshooting processes for servers, storage, networks, jvms, performance, databases, etc. and the usage processes of corresponding tools. The system class is first divided into integration, mobile, esb, reports, etc. according to product characteristics. Each system is classified internally according to memory leakage, CPU overload, database connection problems, performance problems, abnormal crashes, etc. According to different classifications and industry fault handling experiences, summarize the targeted scenario handling processes. For the airport project specifically, information such as specific servers and operation directories can be maintained, and relevant pages can be conveniently drilled down for general class fault troubleshooting and tool usage.

[0095] For example: The integration system of a certain airport is stuck. After retrieval, first retrieve the cases related to the stuck integration system of the corresponding airport preferentially, and arrange the other cases with the keyword "stuck" according to the number of hit keywords and tags. After querying the detailed information, the case process can be referred to for handling to reduce the fault handling efficiency. After regularly improving and adjusting the above knowledge base handling process, on this basis, the present invention can train machine learning models (such as KNN, SVM, deep learning models) in the future to predict possible solutions according to historical data.

[0096] S3. Abstract each fault scenario into a corresponding technical handling process.

[0097] As can be seen from the above embodiments, the present invention realizes centralized management of information such as airports, projects, servers, and deployment information, establishes a perfect information management system, uniformly standardizes the discrete information of each project, and helps to improve the risk prevention awareness and emergency response ability of all employees through the case knowledge base solution. Through the asset management platform, the present invention realizes the centralized management of projects, servers, and deployment information, solves the problems of scattered information and lack of standardization in management. Through the functions of one-key import, automatic acquisition, and scheduled synchronization, the update speed and accuracy of operation and maintenance information are improved, and the operation and maintenance cost is reduced.

[0098] Embodiment 2. Due to the particularity of airport security, the deployment information of each airport's production operation system is maintained in the airport intranet server, and the systems of each airport cannot be centrally managed and maintained;

[0099] In step S2, each project problem includes each problem in service and link problems. Service and link problems include memory leakage, CPU overload, exhaustion of database and middleware connection pools, performance degradation, abnormal crashes, upstream and downstream system data problems, upstream and downstream system link problems, resource leakage, or cache invalidation, etc., which cause service anomalies. Monitoring and troubleshooting these common problems are important measures to ensure the stability and performance of the Jaxa service.

[0100] The fault troubleshooting knowledge base includes each project problem and the troubleshooting information using the corresponding troubleshooting tools.

[0101] Specifically include:

[0102] Regarding memory leakage, the memory consumed by the application continuously increases, GC is frequent, and finally the IVM crashes or the application becomes very slow. Use Jprofler MAT, JConsole Arthus troubleshooting tools for troubleshooting;

[0103] Regarding the CPU overload problem, the application consumes a large amount of CPU resources, resulting in an increase in system load and even making the system unresponsive; use the Jprofler troubleshooting tool for troubleshooting;

[0104] Regarding the exhaustion of the database or middleware connection pool, the application cannot obtain a database or middleware connection and throws an exception of connection pool exhaustion; use the JConsole Arthus troubleshooting tool for troubleshooting;

[0105] Regarding performance degradation, the response time of the application becomes longer and the throughput of processing requests decreases. Use Jprofiler Arthas troubleshooting tools for troubleshooting;

[0106] Regarding abnormal crashes, the application suddenly crashes or throws an uninvestigated exception. Use the Arthas troubleshooting tool for troubleshooting;

[0107] For resource leaks, the application fails to properly close resources, such as files and database connections, which causes system resources to be exhausted or unavailable. Use the Jprofiler Arthas troubleshooting tool to troubleshoot;

[0108] For cache failure, inconsistent or invalid cache data in the application, which leads to frequent database queries or calculations, use Arthas troubleshooting tool to troubleshoot;

[0109] For problems with upstream and downstream system links, the online system link call error and data error lead to system function abnormalities, and Arthas SkyWalking troubleshooting tool is used for troubleshooting;

[0110] Exemplarily, in step S3, each fault scenario is further abstracted into a corresponding technical processing flow.

[0111] This includes the process of troubleshooting abstracted as memory leaks, CPU overload, database and middleware connection pool exhaustion, performance degradation, abnormal crashes, upstream and downstream system data problems, upstream and downstream system link problems, resource leaks, or cache failures;

[0112] Exemplary memory leak troubleshooting process:

[0113] Analyze memory usage;

[0114] Monitor GC recycling frequency;

[0115] Check for large objects occupying memory;

[0116] Check the reference relationship of large objects in memory;

[0117] Find related codes based on reference relationships;

[0118] Troubleshoot code analysis reasons;

[0119] Use memory leak detection and analysis tools (MAT, Jprofler, etc.);

[0120] Another exemplary process for troubleshooting CPU overload is as follows: monitor CPU utilization, analyze thread occupancy, find resource contention and blocking, analyze memory usage, use tools to troubleshoot time-consuming code, and review code logic (complex calculations).

[0121] Finding resource contention and blocking includes:

[0122] First step, resource competition detection: For the competition results obtained by the resource competition mechanism, use the method of quadratic Fourier transform accumulation. Perform Fourier transform by sliding backward at intervals that are not integer multiples of the length of the non-link transmission frequency domain code; then set a judgment threshold for screening and judgment. If the maximum value of the Fourier transform output is greater than the threshold, the screening is successful;

[0123] Second step, resource blockage: Adopt the method of sliding window to find the maximum frequency domain wave peak value. Start sliding at intervals at two phase points before and after the screening position to perform Fourier transform, and perform 1 Fourier transform each time of sliding; then find the maximum value of the frequency domain peak values output by these 8 Fourier transforms and its corresponding position to determine the resource blockage position; The starting points of the 8 slides of the Fourier transform window are respectively at two phase points before and after the screening position. The interval of the first slide of the Fourier transform window is Q×H - 2, the interval of the second slide is Q×H + 1, the interval of the third slide is Q×H + 2,..., the interval of the eighth slide is Q×H + 6, where Q is the length of the link transmission frequency domain code and H is the upsampling multiple.

[0124] In the first step, the resource competition mechanism includes:

[0125] (A) Competition constraint: Establish the best function, and the expression is:

[0126]

[0127] In the formula, l is the vector expression of the project information obtained from the new scenario of the system, Z is the matrix composed of the training samples used by the model, is the model parameter, ξ i is the defined competition weight corresponding to the i-th project sample, Z -i is the matrix composed of the samples excluding the i-th faulty project in the training data, O -i is the decoding coefficient, E is the total number of faulty projects included in the entire training sample;

[0128] (B) Set the competition weight, and the expression is:

[0129]

[0130] In the formula, g max is the maximum value of the distance between the link transmission resource sample and the project sample, g i is the Euclidean distance between the link transmission resource sample and the i-th project sample, and θ is the weight coefficient;

[0131] (C) Establish the optimal link transmission model to obtain the optimal decoding result of the link transmission resource sample.

[0132] In the second step, the method of using a sliding window to find the maximum frequency domain wave peak value for resource blockage includes:

[0133] (1) The CPU first performs 8-phase quantization on the received link transmission frequency-domain signal, maps the signals in different transmission path ranges to 8 phases respectively, and then identifies the quantized signal to obtain the correlation value; the correlation value is obtained by the following formula:

[0134]

[0135] In the formula, corr(j) is the correlation value after identification, F is the link transmission frequency-domain signal after 8-phase quantization, H PN is the link transmission frequency-domain code, k is the sampling point position of the received link transmission frequency-domain signal F, and j is the identification sliding position;

[0136] (2) For the identified correlation value, the quadratic Fourier transform accumulation method is adopted; the Fourier transform window slides backward at intervals of Q×H - 2 sampling points, and one Fourier transform is performed each time it slides;

[0137] (3) Set the decision threshold σ 0 according to the peak value of the Fourier transform output sequence, and perform troubleshooting and judgment; when the accumulated peak value is greater than the decision threshold σ 0 , it is considered that the troubleshooting is successful, and the troubleshooting position is recorded at the same time; otherwise, the Fourier transform window slides backward once and repeats the above steps.

[0138] For the troubleshooting process of database and middleware connection pool exhaustion, performance degradation, abnormal crashes, upstream and downstream system data problems, upstream and downstream system link problems, resource leaks, or cache invalidation, corresponding troubleshooting tools can be used for troubleshooting.

[0139] Example 2, as Figure 2 shown, the offline management system for airport production operation system software service operation and maintenance information provided by the embodiment of the present invention includes:

[0140] Asset management module 1, which constructs an asset management platform based on Java, uses an improved information anomaly identification method to maintain project information and server information. For different middleware covered by the production operation system, through PowerShell scripts or Bash scripts, the installation directory, port number, and version number information of the corresponding scripts of different middleware are obtained by finding environment variables and services, and the script files of different middleware are automatically generated and transmitted to the server side, and the data is uniformly sent back to the asset management platform;

[0141] Knowledge base maintenance module 2, which maintains the data of the problem handling process of each project by using the fault troubleshooting knowledge base based on the service information deployed on the asset management platform;

[0142] Abstraction module 3, which is used to abstract each fault scenario into a corresponding technical processing process.

[0143] In the above embodiments, the descriptions of the respective embodiments have their own focuses. For parts not detailed or recorded in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0144] As described above, only the relatively preferable specific implementation manners of the present invention are provided, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field of the present invention within the technical scope disclosed by the present invention, any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be covered by the protection scope of the present invention.

Claims

1. An offline management method for software service operation and maintenance information of an airport production and operation system, characterized in that: The method includes: S1, asset management, builds an asset management platform based on Java, uses an improved information anomaly identification method to maintain project information and server information, and uses PowerShell scripts or Bash scripts to find environment variables and services to obtain the installation directory, port number, and version number information of the corresponding scripts of different middlewares covered by the production operation system, automatically generates script files of different middlewares, transmits them to the server, and uniformly sends data back to the asset management platform; S2, knowledge base maintenance, based on the service information deployed by the asset management platform, using the troubleshooting knowledge base to maintain the problem handling process data of each project; S3, abstracts each fault scenario into a corresponding technical processing flow.

2. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 1, characterized in that: In step S1, the project information includes: project name, related fault project, and maintenance status; Server information includes IP, virtual IP, operating system version, and system uptime.

3. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 1, characterized in that: In step S1, an asset management platform is constructed based on Java, and the project information and server information are maintained using an improved information anomaly identification method, including: The improved information anomaly identification method includes: an information anomaly identification method that first identifies the normal project server information signal and then identifies the abnormal project server information signal, and an information anomaly identification method that first identifies the abnormal project server information signal and then identifies the normal project server information signal; (1) When as well as When both are established, consider identifying abnormal project server information, among which, When both cannot be established at the same time, the normal project server information separate identification mode is adopted; In the formula, is the maximum transmission power of the RF transmission module of the normal project server information, Z The total amount of target rate superposition for normal project server information, is the maximum transmission power of the project server information identification node, k is the order of derivation, H Z is the target rate of normal project server information, f Z,H is the link power gain between the normal project server information RF transmission module and the project server information identification node, is the power spectral density of Gaussian white noise, f H,Z The link power gain between the project server information identification node and the normal project server information receiving module; (2) When as well as Established, and as well as When both are established, the transmission powers used by the normal project server information radio frequency transmission module and the abnormal project server information radio frequency transmission module in the first identification frequency band are: The project server information identification node first identifies the normal project server information signal and then identifies the abnormal project server information signal; In the formula, is the maximum transmission power of the radio frequency transmission module of the abnormal project server information, Y is the total amount of target rate superposition of abnormal project server information, f Y,H is the link power gain between the abnormal project server information radio frequency transmission module and the project server information identification node; The power allocated by the project server information identification node in the second identification frequency band to the abnormal project server information and the normal project server information signal is: The normal project server information receiving module regards the abnormal project server information signal as noise and directly identifies its own signal. At the abnormal project server information receiving module, the abnormal project server information receiving module first identifies the normal project server information signal, deletes it from the aliasing signal, and then identifies its own signal.

4. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 1, characterized in that: In step S1, the data is uniformly sent back to the asset management platform, including: for airports with a turnover of tens of millions, the script file is regularly sent back to the asset management platform via the websocket interface through the opened secure link; for airports with a turnover of less than tens of millions, the installation directory, port number, and version number information of different middleware received by the server are generated into an Excel file according to the standard template, and the Excel file is regularly downloaded to the local computer and then imported into the asset management platform; the standard template includes the running project, IP, service name, version, port, installation directory, startup method, and process name; Different middleware include Mysql, Oracle, Jboss, ActiveMq, Tomcat, Redis, Nginx, Kettle, FTP, JDK, HTTPD; Jboss uses PowerShell scripts or Bash scripts to search for environment variables and service acquisition methods to obtain the installation directory, port number, and version number information of the corresponding script.

5. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 4 is characterized in that: Jboss uses PowerShell scripts or Bash scripts to search for environment variables and service acquisition methods to obtain the installation directory, port number, and version number of the corresponding script, including: Use the following formula to search for environment variables and services and generate candidate solutions: s = rand * peri Where b is the current search iteration number for searching environment variables and services. is the dth environment variable and service dimension element of the ith solution at iteration b+1, is the dth environment variable and service dimension element of the r1th solution during iteration b, It is the dth environment variable and service dimension element of the s2th solution during iteration b; s is a random number in the environment variable and service search, peri is set to 2-20, s1 is a random number representing the search for a random individual in environment variable and service set domain 1, and s2 is a random number representing the search for a random individual in environment variable and service set domain 2; the candidate solution includes the accurate solution of the installation directory, port number, and version number information of the corresponding script, and the environment variables and services include change information of project information and server information.

6. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 1, characterized in that: In step S2, the deployed service information includes the service name, deployment directory, version, startup method, backup information, and log information; Project issues include: memory leaks, CPU overload, database and middleware connection pool exhaustion, performance degradation, abnormal crashes, upstream and downstream system data issues, upstream and downstream system link issues, resource leaks or cache failures; CPU overload includes: monitoring CPU utilization, analyzing thread occupancy, finding resource competition and blocking, analyzing memory usage, using tools to troubleshoot time-consuming code, and reviewing code logic.

7. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 6 is characterized in that: Finding resource contention and blocking includes: The first step is resource contention detection. The competition results obtained by the resource contention mechanism are accumulated by quadratic Fourier transform, and Fourier transform is performed by sliding backward at intervals of integer multiples of the non-link transmission frequency domain code length. Then a judgment threshold is set for investigation and judgment. If the maximum value of the Fourier transform output is greater than the threshold, the investigation is successful. In the second step, the resource blocking method uses a sliding window to find the maximum frequency domain peak value. The sliding starts at two phase points before and after the troubleshooting position to perform Fourier transform, and one Fourier transform is performed each time. Then the maximum value of the frequency domain peak value output by these eight Fourier transforms and its corresponding position are calculated to determine the resource blocking position. The starting points of the eight sliding of the Fourier transform window are the two phase points before and after the troubleshooting position. The interval of the first sliding of the Fourier transform window is Q×H-2, the interval of the second sliding is Q×H+1, the interval of the third sliding is Q×H+2, ..., and the interval of the eighth sliding is Q×H+6, where Q is the length of the link transmission frequency domain code and H is the upsampling multiple.

8. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 7, characterized in that: In the first step, the resource competition mechanism includes: (A) Competition constraint: Establish the optimal function, expressed as: Where l is the vector expression of the project information obtained by the new scene of the system, Z is the matrix composed of the training samples used by the model, is the model parameter, ξ i is the competition weight corresponding to the i-th project sample, Z -i is the matrix consisting of samples used to remove the i-th fault item in the training data, O -i is the decoding coefficient, E is the total number of fault items contained in the entire training sample; (B) Set the competition weight, the expression is: In the formula, g max is the maximum distance between the link transmission resource sample and the project sample, g i is the Euclidean distance between the link transmission resource sample and the i-th project sample, and θ is the weight coefficient; (C) Establish an optimized link transmission model to obtain the optimal decoding result of the link transmission resource sample.

9. The offline management method for software service operation and maintenance information of an airport production and operation system according to claim 7, characterized in that: In the second step, resource blocking uses a sliding window method to find the maximum frequency domain peak value, including: (1) The CPU first performs 8-phase quantization on the received link transmission frequency domain signal, maps the signals within different transmission path ranges to 8 phases respectively, and then identifies the quantized signal to obtain the correlation value; the correlation value is obtained by the following formula: Where corr(j) is the correlation value after identification, F is the link transmission frequency domain signal after 8-phase quantization, and H PN is the link transmission frequency domain code, k is the sampling point position of the received link transmission frequency domain signal F, and j is the identification sliding position; (2) For the identified correlation values, a quadratic Fourier transform accumulation method is used; the Fourier transform window is spaced by Q×H-2 sampling points and slides backward in sequence, with one Fourier transform performed each time; (3) Set the judgment threshold σ0 according to the peak value of the Fourier transform output sequence, and perform the screening judgment; when the accumulated peak value is greater than the judgment threshold σ0, the screening is considered successful, and the screening position is recorded; otherwise, the Fourier transform window slides back once and repeats the above steps.

10. An airport production and operation system software service operation and maintenance information offline management system, characterized in that: The system is implemented by the offline management method for software service operation and maintenance information of an airport production and operation system according to any one of claims 1 to 9, and the system comprises: The asset management module (1) builds an asset management platform based on Java, uses an improved information anomaly identification method to maintain project information and server information, and uses PowerShell scripts or Bash scripts to obtain the installation directory, port number, and version number information of the corresponding scripts of different middlewares covered by the production operation system by searching environment variables and services, automatically generates script files of different middlewares, transmits them to the server, and uniformly transmits the data back to the asset management platform; The knowledge base maintenance module (2) maintains the problem handling process data of each project based on the service information deployed by the asset management platform and using the troubleshooting knowledge base; The abstraction module (3) is used to abstract each fault scenario into a corresponding technical processing flow.

Images