Intelligent campus operation and maintenance data management system based on Internet

Abstract

The invention discloses a smart campus operation and maintenance data management system based on the Internet, and relates to the technical field of campus operation and maintenance management, and the system comprises a topology modeling unit which is used for recognizing each target object in a campus and constructing a service dependence graph, the service dependency graph is constructed based on a calling dependency relationship between target objects and is persistently stored in a graph database; according to the method, fault positioning is accurate and rapid, the system can identify all upstream and downstream services and complete service links which are directly and intermittently affected when the services are abnormal by constructing the dynamic service dependency map, and the time of fault root cause positioning and influence range analysis is greatly shortened.

Description

A smart campus operation and maintenance data management system based on the Internet Technical Field

[0001] This invention belongs to the field of campus operation and maintenance management technology, specifically a smart campus operation and maintenance data management system based on the Internet. Background Technology

[0002] As the construction of smart campuses progresses, campus information systems are becoming increasingly complex, with various business applications, microservices, and middleware forming a vast and interconnected service network. In this environment, traditional operation and maintenance management methods face severe challenges.

[0003] First, fault location is difficult. When a service node malfunctions, due to the lack of a global, visual view of the service call dependencies, operations and maintenance personnel find it difficult to quickly and accurately locate the root cause of the fault and its scope of impact. They often rely on experience to troubleshoot one by one, resulting in a long mean time to repair (MTTR) and amplifying the impact of business interruption.

[0004] Secondly, the response and handling are passive and inefficient. Although the existing monitoring system can report abnormal indicators, alarms and actual actions are disconnected, heavily relying on manual judgment and operation, resulting in slow response times and a high risk of errors in emergency situations. Furthermore, it lacks predictive and preventative capabilities. The system cannot learn patterns from historical fault data to identify error-prone "weak" services and their peak periods, thus failing to implement targeted preventative measures, such as advance resource allocation or establishing emergency channels, leading to recurring similar faults.

[0005] Therefore, there is an urgent need for an operation and maintenance data management system that can integrate service topology, realize intelligent monitoring, automatically analyze and make decisions, and perform predictive maintenance, so as to improve the stability, availability and operation and maintenance efficiency of smart campus business systems. Summary of the Invention

[0006] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an Internet-based smart campus operation and maintenance data management system.

[0007] A smart campus operation and maintenance data management system based on the Internet includes: a topology modeling unit, used to identify target objects within the campus and construct a service dependency graph, wherein the target objects are software units with independent business functions, and the service dependency graph is constructed based on the call dependency relationships between target objects and persistently stored in a graph database; a status monitoring unit, used to monitor each target object in real time, collect its key performance indicators, and generate abnormal events when the indicators meet abnormal conditions; an analysis and decision-making module, used to, after receiving an abnormal event, take the abnormal object as the starting point, analyze the upstream services directly affected and the indirectly affected downstream services based on the service dependency graph, determine the set of affected service chains and the scope of business impact, and generate an impact analysis report; and a decision execution module, used to select and execute corresponding service link adjustment strategies from a pre-set strategy library according to the impact analysis report and the severity of the abnormality, wherein the strategies include at least one of service isolation restart, service degradation bypass, and traffic switching to a backup link.

[0008] Furthermore, the topology modeling unit is specifically used to: connect two target objects with a direct call relationship through directed edges, construct a service chain based on the call order, and store all target objects and service chain relationships as a service dependency graph.

[0009] Furthermore, the key performance indicators collected by the status monitoring unit include at least one of the following: request response time, request error rate, service availability status, and resource utilization.

[0010] Furthermore, the status monitoring unit is also used to set dynamic thresholds for each indicator and determine that the target object has entered an abnormal state when multiple indicators are continuously abnormal within a continuous time window.

[0011] Furthermore, the analysis and decision-making module is also used to determine the set of services that directly or indirectly depend on the abnormal object by traversing the incoming edges of the abnormal object in the service dependency graph in reverse, as the direct impact domain of the fault.

[0012] Furthermore, the analysis and decision-making module is also used to assess the indirect impact of anomalous objects on their downstream services, including observing the affected downstream services by simulating anomalous situations.

[0013] Furthermore, it also includes a fault-prone object identification unit, which is used to count the frequency of abnormal events occurring in each target object within a set time window, and mark it as a fault-prone object if it exceeds a preset threshold.

[0014] Furthermore, it also includes a fault-prone time period analysis unit, which is used to count the number of abnormal events for each fault-prone object according to a time period. If the number of abnormal events in a certain period reaches the periodic abnormality threshold, then the period is marked as the fault-prone time period of that object.

[0015] Furthermore, it also includes a backup and temporary channel management unit, which is used to back up faulty objects before the start of the faulty period and switch traffic to the backup instance when an anomaly is detected during that period.

[0016] Compared with the prior art, the beneficial effects of the present invention are: the fault location of this application is accurate and fast. By constructing a dynamic service dependency graph, the system can identify all directly and indirectly affected upstream and downstream services and the complete business link when a service is abnormal, which greatly shortens the time for fault root cause location and impact range analysis.

[0017] Simultaneously, it enables intelligent and automated decision-making in handling situations. The system integrates analysis, decision-making, and execution modules, and can automatically match or generate handling strategies, such as isolation, downgrading, and switching, based on the fault impact analysis report. These strategies are then automatically executed via API calls, achieving a closed loop from perception to recovery, reducing manual intervention, and improving handling efficiency and accuracy. Attached Figure Description

[0018] Figure 1 is a system block diagram of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please refer to Figure 1. This application provides an Internet-based smart campus operation and maintenance data management system. As an embodiment of this application, it specifically includes: a topology modeling unit, used to identify various target objects in the campus and construct a service dependency graph, wherein the target object is a software unit with independent business functions, and the service dependency graph is constructed based on the call dependency relationship between target objects and persistently stored in a graph database; a status monitoring unit, used to monitor each target object in real time, collect its key performance indicators, and generate abnormal events when the indicators meet abnormal conditions; an analysis and decision module, used to, after receiving an abnormal event, take the abnormal object as the starting point, analyze the upstream services directly affected by it and the indirectly affected downstream services based on the service dependency graph, determine the set of affected service chains and the scope of business impact, and generate an impact analysis report; and a decision execution module, used to select and execute corresponding service link adjustment strategies from a preset strategy library according to the impact analysis report and the severity of the abnormality, wherein the strategies include at least one of service isolation restart, service degradation bypass, and traffic switching to a backup link.

[0021] As a second embodiment of this application, this embodiment is implemented based on the first embodiment, with the difference being that this embodiment specifically includes: a topology modeling unit, used to obtain all target objects. Here, a target object refers to each independently accessible software unit with a clear business function. For ease of description later, the target objects are marked as Di, i=1,...,n, indicating that there are n target objects. For example, D1 is a unified identity authentication interface, D2 is a book borrowing business logic service, D3 is a book inventory query database service, etc. Each Di is assigned a globally unique identifier and metadata, such as service type, business department, and physical / logical cluster. The topology modeling unit is also used to analyze historical call logs, real-time traffic sniffing, or configuration declarations to determine the call dependencies between target objects and construct a service chain. Specifically, the construction method is as follows: two objects Di and Dj with a direct call relationship are connected by a directed edge Eij, where Di calls Dj, representing a basic service dependency. Multiple such dependency edges are connected end-to-end in the call order to form a complete service chain SC. k It represents the service path traversed by an end-to-end business process; a service chain can be represented as SC. k =(D k1 →D k2 →…→D kn ), where D k1 Typically, this is a user-facing entry point service, such as a book borrowing interface, D kn It is the terminal service of the link, such as a database; all target objects D and service chain relationships E are persistently stored in a graph database, forming a dynamic service dependency graph G=(D,E); this graph is the basis for all subsequent analysis and decisions; a status monitoring unit is used for each target object D. i Deploy lightweight monitoring probes or integrate them with existing monitoring systems to continuously collect key performance indicators (M). i (t), including request response time RT i (t), Request Error Rate ER i (t), Service Availability Status S i (t) (such as HTTP status codes), resource utilization U i (t) (e.g., CPU, memory); the system sets dynamic thresholds for each indicator; when D is detected within a continuous time window Δt... i When one or more indicators consistently meet abnormal conditions, for example, ER i (t)>θ e And RT i (t)>θ y That is, to determine the target object D. iEntering an abnormal state and generating an abnormal event, which includes the target object of the abnormal state and the corresponding abnormal indicator; θ e and θ y All are preset dynamic thresholds; the analysis and decision module, when the status monitoring unit detects information about target object D... i Upon the occurrence of an abnormal event, the impact analysis and decision-making module is immediately triggered; firstly, the target object corresponding to the abnormal event is marked as abnormal object D. f ; with exception object D f Starting from the role of the callee in graph G, traverse all incoming edges in reverse order, that is, find all calls to D. f These services directly or indirectly depend on D f The upstream services that function normally constitute the direct impact domain Du of the fault. For example, if D f If it is a book inventory query service, then the book borrowing business logic service D2 that directly calls it will be immediately incorporated into Du.

[0022] Simultaneously check D f As the caller, assess the potential indirect impact of its anomaly on its downstream services (i.e., the services it calls), such as timeout accumulation. Here, we can directly simulate the anomaly of the corresponding anomaly object to observe the downstream services that will be affected. Finally, by combining upstream and downstream information, we can determine the complete set of affected service chains Sc, estimate the scope of affected business and user scale, and generate an impact analysis report.

[0023] The decision execution module is used to match or dynamically generate reorganization plans from a pre-set strategy library based on the impact analysis report and the severity of the anomaly. Specific plans are as follows: Strategy P1, rapid isolation and restart, suitable for temporary faults; Action: Isolate and restart the abnormal object D. f Isolate the instance from the load balancer pool, restart it, and reconnect it after the health check passes.

[0024] Strategy P2, service degradation and bypass, is applicable to non-core dependency failures; Action: Modify the calling logic of upstream service D2 to skip the abnormal D. f Instead, it calls a pre-defined simplified interface or returns cached data to ensure the main process is available; Strategy P3, traffic switching and backup connection management, for core links with backup service chains; Action: Transfer traffic from the link containing D f Original service chain SC k The entire system switches to a pre-configured, functionally equivalent backup service chain SCb; Strategy P4 (automatic scaling): Addresses performance anomalies caused by excessive load. Action: Automatically triggers D. f Horizontal scaling of the service cluster involves adding instances to distribute the load.

[0025] According to Df Based on the business importance, fault type, and criticality of Sc, select one or more strategy combinations, such as first executing P1, and if it fails, upgrading to P3 after X seconds; after determining the solution, the reorganization execution module automatically executes the selected strategy by calling platform APIs (such as Kubernetes API, service mesh control plane API, application configuration center API) to complete the real-time adjustment of service links.

[0026] As a third embodiment of this application, this embodiment is implemented based on embodiment two, except that it further includes: a fault-prone object identification unit, used to continuously monitor each target object D. i The frequency of abnormal events; if within the statistical time window T, the target object D i The number of abnormal events exceeds the preset frequency threshold F t If it is an exception, it is marked as an error-prone object, and the timestamp sequence of its exception events is recorded.

[0027] The error-prone time period analysis unit is used to analyze the occurrence pattern of abnormal events for each error-prone object according to the time series. The specific analysis process includes: 1. Dividing the time into months, with each half-month as a cycle, i.e., the first cycle is from the 1st to the 15th of each month, and the second cycle is from the 16th to the last day of the month; 2. Counting the number of abnormal events Cc of the error-prone object in each cycle; 3. Setting a cycle abnormality threshold Ct. If the number of abnormal events Cc ≥ Ct in a certain cycle, then the cycle is marked as the error-prone time period of that error-prone object; 4. Recording the set of error-prone time periods for each error-prone object and storing it in the error-prone time period knowledge base.

[0028] The management unit is used to back up error-prone objects before the error-prone period arrives and to establish a temporary channel to connect with the backup content if a problem occurs during the error-prone period. Specifically, it includes: 1. Based on the error-prone period knowledge base, triggering a backup operation on the error-prone object at a preset time ΔT before the start of the error-prone period, generating a backup instance Db and maintaining its ready state; 2. When an error-prone object D is detected during the error-prone period... i When an abnormal event occurs, the following actions will be performed automatically: Establish a temporary channel to redirect the original flow to D. i Some or all of the traffic is switched to backup instance D. b ; By calling platform APIs (such as load balancer API, service mesh control plane API), redirect calls to faulty objects in the service chain; record switching events and update the exception mode library; 3. When the faulty period ends or the faulty object recovers and runs stably, automatically switch traffic back to the original target object and retain the backup instance until the next faulty period begins.

[0029] Of course, this application may also include an optimization and learning unit, used to dynamically adjust the identification threshold Ft and periodic anomaly threshold Ct of error-prone objects according to actual operating data, and update the error-prone time period knowledge base to improve the system's adaptability; the specifics depend on the administrator's needs; the above embodiments are only used to illustrate the technical methods of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A smart campus operation and maintenance data management system based on the Internet, characterized in that, include: The topology modeling unit is used to identify target objects within the campus and construct a service dependency graph. The target objects are software units with independent business functions. The service dependency graph is constructed based on the call dependency relationships between target objects and is persistently stored in a graph database. The status monitoring unit is used to monitor each target object in real time, collect its key performance indicators, and generate abnormal events when the indicators meet abnormal conditions. The analysis and decision-making module is used to, upon receiving an abnormal event, take the abnormal object as the starting point, analyze the upstream services directly affected and the indirectly affected downstream services based on the service dependency graph, determine the set of affected service chains and the scope of business impact, and generate an impact analysis report; the decision execution module is used to select and execute the corresponding service link adjustment strategy from the pre-set strategy library according to the impact analysis report and the severity of the abnormality. The strategy includes at least one of the following: service isolation and restart, service degradation and bypass, and traffic switching to a backup link.

2. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, The topology modeling unit is specifically used to: connect two target objects with a direct call relationship through directed edges, build a service chain based on the call order, and store all target objects and service chain relationships as a service dependency graph.

3. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, The key performance indicators collected by the status monitoring unit include at least one of the following: request response time, request error rate, service availability status, and resource utilization.

4. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, The status monitoring unit is also used to set dynamic thresholds for each indicator and determine that the target object has entered an abnormal state when multiple indicators are continuously abnormal within a continuous time window.

5. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, The analysis and decision-making module is also used to determine the set of services that directly or indirectly depend on the abnormal object by traversing the incoming edges of the abnormal object in the service dependency graph in reverse, which serves as the direct impact domain of the fault.

6. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, The analysis and decision-making module is also used to assess the indirect impact of anomalous objects on their downstream services, including observing the affected downstream services by simulating anomalous situations.

7. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, It also includes a fault-prone object identification unit, which is used to count the frequency of abnormal events occurring in each target object within a set time window, and mark it as a fault-prone object if it exceeds a preset threshold.

8. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, It also includes a fault-prone time period analysis unit, which is used to count the number of abnormal events for each fault-prone object by time period. If the number of abnormal events in a certain period reaches the period abnormal threshold, then the period is marked as the fault-prone time period of that object.

9. The Internet-based smart campus operation and maintenance data management system according to claim 1, characterized in that, It also includes a backup and temporary channel management unit, which is used to back up error-prone objects before the start of an error-prone period and switch traffic to the backup instance when an anomaly is detected during that period.

Images