Artificial intelligence-based service health check method and service health check plug-in

Abstract

The present invention discloses a service health check method and a service health check plug-in based on artificial intelligence. The method comprises: the health check plug-in uses a health check interface to check the service status of a service to be checked; when the service status is abnormal, the health check plug-in sets an error mark according to the type of the service status; and an application performance management system determines the service status of the service to be checked based on the set error mark. The present invention solves the technical problem of high maintenance cost of service health checks.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to an artificial intelligence-based service health check method and a service health check plug-in. Background Art

[0002] As the science and technology education sector increasingly relies on online service platforms and distributed application systems, ensuring the stability and efficient operation of services has become a vital task. In order to monitor the service status in real time, many applications have introduced health check mechanisms so that problems can be discovered and appropriate measures can be taken in a timely manner. Service health checks usually determine whether the service is running normally by regularly checking the service status. As an important part of service monitoring, the health check interface allows the monitoring system to obtain the health status of the service and generate alarms or take other response measures based on the health status information.

[0003] Currently, many application performance management (APM) systems, such as Skywalking and Prometheus, use health check interfaces to collect service health status data and trigger alarms and other response operations based on this data. Figure 1 As shown in the figure, when the service sets its status to OUT_OF_SERVICE during smooth release, the health check mechanism of the APM system will regard this status as an unhealthy service and trigger an alarm, resulting in a relatively high cost for manual maintenance of the alarm.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a service health check method and a service health check plug-in based on artificial intelligence, so as to at least solve the technical problem of high maintenance cost of service health check.

[0006] According to one aspect of an embodiment of the present invention, a method for service health check for science and technology education is provided, including: a health check plug-in uses a health check interface to check the service status of a service to be checked; when the service status is an abnormal state, the health check plug-in sets an error flag according to the type of the service status; an application performance management system determines the service status of the service to be checked based on the set error flag.

[0007] According to another aspect of an embodiment of the present invention, a service health check plug-in for science and technology education is provided, including: a check module, configured to use a health check interface to check the service status of a service to be checked; a marking module, configured to, when the service status is abnormal, the health check plug-in sets an error mark according to the type of the service status; a determination module, configured to apply a performance management system to determine the service status of the service to be checked based on the set error mark.

[0008] In the embodiment of the present invention, the health check plug-in uses the health check interface to check the service status of the service to be checked; when the service status is abnormal, the health check plug-in sets an error mark according to the type of the service status; the application performance management system determines the service status of the service to be checked based on the set error mark. Through the above solution, the technical problem of high maintenance cost of service health check is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0010] Figure 1 is a flow chart of a method for service health check for science and technology education according to the prior art;

[0011] Figure 2 is a flow chart of an optional service health check method for science and technology education according to an embodiment of the present invention;

[0012] Figure 3 is a flowchart of another optional service health check method for science and technology education according to an embodiment of the present invention;

[0013] Figure 4 is a flowchart of a method for dynamically analyzing the status of a service instance using artificial intelligence and adjusting an alarm strategy according to a status change according to an embodiment of the present invention;

[0014] Figure 5 is a schematic diagram of the structure of an optional service health check plug-in for science and technology education according to an embodiment of the present invention;

[0015] Figure 6 A schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0016] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0018] According to an embodiment of the present invention, an embodiment of a method for service health check for science and technology education is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0019] Skywalking is an open source application performance management (APM) system designed for monitoring, diagnosing and analyzing modern distributed systems (such as microservices, containerization and cloud-native architectures). Skywalking provides powerful performance monitoring, link tracking, log analysis and indicator aggregation functions to help developers and operation and maintenance personnel understand the performance status of applications in real time and quickly locate performance bottlenecks or failures. It supports monitoring across multiple platforms (such as Kubernetes, Docker) and languages ​​(such as Java, Python, Node.js), and can effectively capture and analyze service call links in microservice architectures to discover potential problems and optimization points.

[0020] Rolling Deployment or Canary Release is a modern service update strategy that aims to achieve seamless updates of services to avoid service interruptions. During the smooth release process, the service will gradually and in batches deploy the new version to the production environment instead of updating all at once. In this way, some old and new versions of the service will run in parallel, allowing operations personnel to monitor and test the performance of the new version in the production environment. When a problem occurs, you can quickly roll back to the old version to ensure minimal impact on users. The smooth release strategy is often used to avoid the risks brought by "full release", especially in large-scale distributed systems.

[0021] OUT_OF_SERVICE is a service status indicator, indicating that the service is currently temporarily unavailable. It is usually used when the service is undergoing maintenance, update, or smooth release. When the status of a service is set to OUT_OF_SERVICE, it usually stops accepting new requests until it returns to normal. This status is common in microservice architectures, helping the system avoid receiving traffic when the service is unavailable and avoiding impacts on users. It is also a common status indicator in health check mechanisms, indicating that the service is offline or under maintenance, to avoid triggering errors or unnecessary alarms.

[0022] Alarms are a common notification mechanism in system monitoring. When the system monitors an abnormality or predefined condition (such as service unavailability, performance degradation, resource overload, etc.), an alarm will be triggered. Alarms usually notify operation and maintenance personnel or developers in the form of emails, text messages, real-time notifications, or console pop-ups to remind them that there are potential problems that need to be solved. Alarms can help the team detect and respond to system failures, performance bottlenecks, or potential security threats in a timely manner. The accuracy of alarms and the timeliness of responses are crucial to ensuring system stability and improving operation and maintenance efficiency.

[0023] Figure 2 is a method for service health check for science and technology education according to an embodiment of the present invention, such as Figure 2 As shown, the method comprises the following steps:

[0024] Step S202: The health check plug-in uses the health check interface to check the service status of the service to be checked.

[0025] The health check interface sends a heartbeat message to the service to be checked, and the service to be checked returns the service status; the health check plug-in intercepts and obtains the service status returned by the service to be checked.

[0026] Step S204: When the service status is abnormal, the health check plug-in sets an error flag according to the type of the service status.

[0027] When the type of the service status returned by the service to be checked is unavailable, the health check plug-in sets the error flag to false; when the type of the service status returned by the service to be checked is other than the unavailable and normal states, the health check plug-in sets the error flag to true.

[0028] Step S206: The application performance management system determines the service status of the service to be checked based on the set error flag.

[0029] When the error flag is set to false, the application performance management system determines that the service to be checked is in a healthy state; when the error flag is set to true, the application performance management system determines that the service to be checked is in an unhealthy state, and the application performance management system triggers an alarm.

[0030] By setting adaptive error markers, the present invention effectively avoids false alarm triggering when the service is smoothly released, ensuring that the Skywalking health check accurately judges the service status. During the smooth release process, the system avoids triggering unnecessary alarms because the service status is set to OUT_OF_SERVICE, thereby reducing the interference of developers. In addition, the present application reduces the troubleshooting work performed due to false alarms, reduces the workload of operation and maintenance personnel and developers, and saves a lot of time and cost. Finally, through a more intelligent health check mechanism, the present invention enhances the detection capability of the health status of services in the microservice architecture, ensuring that the system maintains efficient and stable operation during upgrades and maintenance.

[0031] Skywalking is an open source APM system for monitoring and diagnosing system performance issues in a microservice architecture. In a microservice architecture, in order to achieve smooth release, services usually set their own status to OUT_OF_SERVICE to avoid receiving new requests. However, Skywalking's health check mechanism will regard this status as an unhealthy service and trigger an alarm.

[0032] The above existing solutions have shortcomings. First, when the service is smoothly released, the status of the service will be set to OUT_OF_SERVICE, but Skywalking will mistakenly regard this status as an unhealthy service, resulting in frequent triggering of alarms. Secondly, these frequent alarms interfere with developers' judgment of actual problems, making it impossible for them to quickly distinguish between real failures and normal temporary unavailable states, thus affecting the location of problems and response efficiency. Finally, since developers need to investigate each alarm, it leads to unnecessary workload and increases maintenance costs.

[0033] The present application also provides a service health check method for science and technology education. The method is particularly targeted at service health status detection in a microservice architecture, and intelligently determines the status of the service through a health check plug-in, avoids false alarms, and reduces the maintenance burden on developers. Figure 3 As shown, the method comprises the following steps:

[0034] Step S302: Service is released smoothly.

[0035] During the service update and deployment process, a rolling deployment strategy is used to ensure system stability and user experience. Rolling deployment is a progressive update strategy that can gradually apply new code versions to the production environment without interrupting existing services. In this way, the new version only replaces some service instances, while other service instances remain running, ensuring high availability of the service.

[0036] In the early stages of a smooth release, the service will maintain its original state and continue to respond to requests. At this point, the status of the service to be checked is set to normal to allow the new version to be gradually deployed to the production environment. The key point of a smooth release is that the various steps in the release process will not directly affect the availability of the service, and the service can still continue to process requests, ensuring that the system is not interrupted by the update.

[0037] Step S304, setting the service status to OUT_OF_SERVICE

[0038] To ensure that the health status of the new version of the service can be accurately judged during the service update process, the service will be set to OUT_OF_SERVICE to indicate that the service is currently temporarily unavailable and will not receive new requests. This setting occurs during the service upgrade or maintenance process. The setting of OUT_OF_SERVICE status is a common practice in smooth releases, and its purpose is to prevent new services from receiving too many requests and prevent request conflicts or data inconsistencies.

[0039] In this state, the health check interface will not respond to new requests, and the service will be removed from the load balancing pool to prevent excessive traffic and requests from entering the service instance currently being updated. After setting it to OUT_OF_SERVICE state, the health check plug-in needs to make intelligent judgments on the alarm logic based on this state to avoid false alarms and excessive intervention.

[0040] Step S306, set up a health check plug-in in Skywalking and check the / health interface.

[0041] As an APM (application performance management) system, Skywalking relies on a health check interface to assess the health of a service. This interface is usually an HTTP endpoint called / health, through which a health check plug-in can access the current health status of a service. However, in the embodiment of the present application, a self-developed health check plug-in is provided in Skywalking.

[0042] The health check plug-in will periodically initiate health check requests for each service instance, access the service's / health interface, and obtain the service's health status data. The request will return the status of the current service instance, including whether it is normal (UP) or abnormal. Usually, abnormal conditions include: unavailable (OUT_OF_SERVICE), DOWN, UNKNOWN and other key indicators.

[0043] Step S308, determine if the interface returns OUT_OF_SERVICE.

[0044] After the health check plug-in receives the response from the / health interface, it will make a judgment based on the returned data. If the returned status is OUT_OF_SERVICE, it means that the service instance is undergoing smooth release or maintenance and should not receive new requests. Therefore, the plug-in needs to further judge this status and handle it accordingly.

[0045] The plug-in determines the subsequent alarm processing method based on the service status returned by the / health interface. Since it is expected that the service is in an abnormal state during the smooth release process, the plug-in will decide whether to trigger the alarm based on the status. Specifically, if OUT_OF_SERVICE is returned, step S310 is executed, otherwise step S312 is executed.

[0046] Step S310, set the error flag to false, Skywalking does not consider the service unhealthy.

[0047] If the health check plugin finds that the / health interface returns an OUT_OF_SERVICE status (which indicates that it is caused by the service being in the process of smooth release rather than an actual failure), the plugin will automatically set the error flag to false. In this way, the plugin can distinguish between normal service downtime (such as status updates during smooth release) and real service failures, thereby avoiding misjudging the service as unhealthy.

[0048] At this time, the Skywalking system will not mark the service as unhealthy, so no alarm will be triggered. Through this mechanism, the plug-in avoids triggering alarms due to state changes when the service is smoothly released, ensuring the accuracy and effectiveness of the alarm. In this way, it can effectively reduce the maintenance pressure on developers caused by false alarms and reduce unnecessary workload.

[0049] Step S312: trigger an alarm, the developer receives the alarm and conducts troubleshooting

[0050] If the health check plug-in finds that the status returned by the / health interface is not OUT_OF_SERVICE, or returns other abnormal status (such as DOWN or UNKNOWN), it means that the service is abnormal and there may be a fault. In this case, the plug-in will set the error flag to true, indicating that the service is unhealthy, and Skywalking will trigger an alarm.

[0051] At this point, the system will notify the developer or operation and maintenance team, prompting them that there is a problem with the service and further investigation is required. The alarm notification usually includes detailed exception information, the time of occurrence, the affected service and the specific error message to help developers quickly locate and solve the problem. In this way, Skywalking can effectively help operation and maintenance personnel discover and respond to actual fault problems and ensure the stable operation of the system.

[0052] In addition, after analyzing the actual effect of each alarm trigger (such as whether the problem was successfully solved or whether a false alarm was generated), the plug-in can self-optimize through an online learning algorithm. The plug-in adjusts its decision-making process based on real-time feedback (for example, whether the developer responded to an alarm). For example, if an alarm is frequently misjudged, the plug-in will adjust the prediction model or change the alarm threshold to improve accuracy. When the plug-in detects an abnormal service status, in addition to triggering an alarm, the plug-in can also combine historical data, fault characteristics, and trend analysis to provide intelligent troubleshooting suggestions. These suggestions can be operational steps based on AI analysis to help developers locate problems more quickly, such as: "The request response time of service instance A is abnormal. You can check the database connection of the instance." "The load of instance B suddenly increased. It may be due to an abnormal external API call. It is recommended to check the API dependency."

[0053] The present invention adopts an adaptive error mark setting mechanism, which avoids false alarms by dynamically judging the health status of the service, and reduces unnecessary troubleshooting work for operation and maintenance personnel. Traditional health check methods often cannot distinguish between temporary downtime and service failures, which easily leads to false alarms. The present invention intercepts the returned status through a plug-in and intelligently sets error marks, so that the system can intelligently distinguish between normal service status changes and abnormal situations, thereby improving the accuracy of alarms and reducing the interference of alarms on developers in judging real problems.

[0054] In addition, the technical solution of the present invention fully considers the service update strategy under the microservice architecture. In particular, when smooth release is performed, the service status may become OUT_OF_SERVICE, but this status does not mean that there is a problem with the health status of the service. By optimizing the judgment logic of the health check plug-in in a targeted manner, the present invention can dynamically adjust the error mark according to the service status type, avoiding false alarms caused by normal status changes such as smooth release.

[0055] The health status and alarm strategy of service instances are not only affected by instantaneous data, but also involve complex dynamic changes and timing correlations. Although the service status can be directly read and a simple alarm judgment can be made, when faced with a large number of service instances and a frequently changing system environment, traditional alarm methods often produce too many false alarms or missed alarms, resulting in waste of resources or potential problems being ignored. Therefore, an embodiment of the present application also provides a method for using artificial intelligence to dynamically analyze the status of service instances and adjust alarm strategies based on status changes. The AI ​​model can identify and understand the complex patterns, dynamic trends, and interdependencies of service status by learning historical data, thereby making more accurate and intelligent judgments. Figure 4 As shown, the method comprises the following steps:

[0056] Step S402: Use the data set to train the AI ​​model.

[0057] The plugin first obtains the health status data of the service instance by regularly accessing the / health interface of the service. The data returned by each health check request includes the following key information: the current service status (such as UP, OUT_OF_SERVICE, DOWN, UNKNOWN, etc.); the load of the service instance (CPU, memory, network and other indicators); historical alarm data (whether an alarm is triggered, the historical frequency of alarms, etc.); service request response time and error rate; external system dependencies (such as database connections, third-party API calls, etc.)

[0058] Model the features of the health check data of each service instance. By constructing multiple decision surfaces, the data is divided layer by layer. In each division step, a feature and its corresponding split point are randomly selected to split the data. Since the abnormal points are isolated earlier in the plane structure, they are quickly divided into shallower nodes, so the path length of these points is shorter, and a higher abnormality score is obtained. Next, the data in the normal state is further separated from the data in the abnormal state, and the interval between the two types of data is maximized. In the process of optimizing the decision surface, the position and normal vector of the decision surface are adjusted to minimize the classification error, so that the normal and abnormal states of the service instance can be effectively distinguished in the high-dimensional space, providing an accurate judgment basis for the subsequent state classification. During the training process, the separation distance between the normal state and the abnormal state is maximized. In this way, in the feature space, the samples in the normal state will be as far away from the samples in the abnormal state as possible, and vice versa. At the same time, the decision surface is also optimized by minimizing the classification error. By adjusting the position and normal vector of the decision surface, these misclassified samples can be reduced, thereby ensuring that more data points are correctly classified. Finally, by solving the optimization problem, the optimal normal vector and intercept are obtained, so that the normal and abnormal samples can be separated to the greatest extent in the feature space.

[0059] After that, data analysis is performed based on the classification results. In order to further improve the alarm accuracy of the system, multi-dimensional time series data analysis is introduced. The time series correlation between multiple abnormal service instances is analyzed, the dependencies between different service instances are identified, and the potential connections of service status changes are determined. In addition, based on the interaction relationship between services, the chain effect of failures between services is captured, thereby improving the accuracy of fault prediction and anomaly detection.

[0060] Finally, decide whether to trigger an alarm. If the current service status is determined to be abnormal (such as a service instance in the "DOWN" state or the load is too high), an alarm is triggered; if the service status is normal (such as the OUT_OF_SERVICE state is within the expected fluctuation range), avoid triggering the alarm. By training machine learning models based on historical alarm data (such as the error rate of service instances, request response time, etc.), it is also possible to predict whether a service is about to enter a faulty state. For example, if the error rate of a service instance has an upward trend over the past period of time, the plug-in can issue an early warning to the operation and maintenance personnel. The model uses a time series prediction algorithm to predict the health status of the service in the next few hours based on historical time series data such as system load and request response. If it is predicted that a service may enter an unavailable state, the plug-in will lower the alarm threshold in advance or set a high-priority alarm to ensure that developers and operation and maintenance personnel have enough time to deal with potential problems.

[0061] Step S404: Automatically issue an alarm using an AI model.

[0062] AI enables adaptive alarm strategies, automatically adjusting alarm levels or thresholds based on state changes. The plug-in can dynamically adjust the trigger threshold of the alarm by learning data such as historical alarm frequency, type, and time period. For example, if certain states (such as OUT_OF_SERVICE) are identified as normal and do not require intervention during a smooth release, the system can increase the threshold time of the alarm to avoid frequent alarm triggering in a short period of time. Depending on the type of fault and the predicted results, the plug-in can set different alarm levels. For example, for minor performance fluctuations or load increases, the plug-in triggers a low-level alarm (such as "Warning"), and for faults that may cause service interruptions, the plug-in triggers a high-level alarm (such as "Critical" or "Emergency"). In this way, developers can prioritize more urgent issues based on the severity of the alarm.

[0063] The embodiment of the present application uses an AI model to make adaptive adjustments in real time based on the operating status and historical data of the service instance, thereby optimizing the accuracy of the alarm and reducing false positives and missed positives. Through deep learning and machine learning algorithms, the plug-in can predict potential system failures.

[0064] This application also provides a service health check plug-in for technology education, such as Figure 5 As shown, it includes: a checking module 52, which is configured to use a health check interface to check the service status of the service to be checked; a marking module 54, which is configured to set an error mark according to the type of the service status when the service status is abnormal; and a determination module 54, which is configured to determine the service status of the service to be checked based on the set error mark by the application performance management system.

[0065] It should be noted that the service health check plug-in for science and technology education provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the service health check plug-in for science and technology education provided in the above embodiment belongs to the same concept as the service health check method embodiment for science and technology education. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0066] This embodiment uses an intelligent error mark setting mechanism to ensure that APM systems such as Skywalking can accurately determine the health status of services, thereby avoiding false alarms and reducing maintenance costs. Through the present invention, the system can improve development and operation efficiency while ensuring service stability and reduce overall costs.

[0067] Figure 6 FIG. 1 shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiment of the present disclosure. It should be noted that: Figure 6 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0068] like Figure 6 As shown, the electronic device includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage part 1008 into a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for system operation are also stored. The CPU 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0069] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1010 as needed, so that a computer program read therefrom is installed into the storage section 1008 as needed.

[0070] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for service health check based on artificial intelligence, characterized in that: include: The health check plug-in uses the health check interface to check the service status of the service to be checked; In the case where the service status is an abnormal state, the health check plug-in sets an error flag according to the type of the service status; The application performance management system determines the service status of the service to be checked based on the set error flag.

2. The method according to claim 1, characterized in that The health check plug-in uses the health check interface to check the service status of the service to be checked, including: The health check interface sends a heartbeat message to the service to be checked, and the service to be checked returns the service status; The health check plug-in intercepts and obtains the service status returned by the service to be checked.

3. The method according to claim 2, characterized in that When the service status is abnormal, the health check plug-in sets an error flag according to the type of the service status, including: When the type of the service status returned by the service to be checked is unavailable, the health check plug-in sets the error flag to false; When the type of the service status returned by the service to be checked is other than the unavailable and normal states, the health check plug-in sets the error flag to true.

4. The method according to claim 3, characterized in that The application performance management system determines the service status of the service to be checked based on the set error flag, including: In the case where the error flag is set to false, the application performance management system determines that the service to be checked is in a healthy state; In the case where the error flag is set to true, the application performance management system determines that the service to be checked is in an unhealthy state.

5. The method according to claim 4, characterized in that After the application performance management system determines that the service to be checked is in an unhealthy state, the method further includes: the application performance management system triggering an alarm.

6. The method according to claim 1, characterized in that Before the health check plug-in uses the health check interface to check the service status of the service to be checked, the method also includes: the service to be checked is smoothly released, and the service status of the service to be checked is set to an unavailable state, wherein the abnormal state includes the unavailable state.

7. An artificial intelligence-based service health check plug-in, characterized in that: include: A check module is configured to check the service status of the service to be checked by using a health check interface; a marking module, configured to, when the service status is an abnormal status, cause the health check plug-in to set an error mark according to the type of the service status; The determination module is configured to determine the service status of the to-be-checked service based on the set error mark by applying the performance management system.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 6.

9. A computer device, characterized in that: include: Memory and processor, The memory stores a computer program; The processor is used to execute the computer program stored in the memory, and when the computer program is run, the processor is enabled to execute the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images