A method and system for panoramic monitoring of industrial infrastructure resource service operation
By establishing a generational genealogy of resource instances and a temporal set of resource dependencies, the actual resource instances and dependency links to which monitoring events belong can be identified and restored, thus solving the problem of incorrect attribution of monitoring events and achieving accuracy and traceability of the overall operation of the industrial cloud platform.
Patent Information
- Application Number
- CN202611635959.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-22
- Publication Date
- 2026-11-17
AI Technical Summary
In industrial cloud platforms, the occurrence time of monitoring events is earlier than the startup time of the currently associated resource instance and earlier than the establishment time of the target dependency relationship. This causes the monitoring events to be incorrectly associated with descendant resource instances, forming a reverse-generational anomaly chain before birth, which affects the accuracy and traceability of the overall operation.
By establishing a generational genealogy of resource instances and a set of temporal resource dependencies, we can identify temporal logical conflicts in the actual resource instances and dependency connections of monitored events, restore the actual resource instances and dependency links of monitored events, and generate an accurate operational panorama.
This improves the authenticity, completeness, and traceability of the overall operation of basic resource services, providing a reliable basis for anomaly verification and operation and maintenance of industrial cloud platforms.
Smart Images

Figure CN123114901A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of cloud platform operation monitoring technology, specifically to a method and system for panoramic monitoring of industrial infrastructure resource service operation. Background Technology
[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.
[0003] With the centralized deployment of computing resources, container resources, databases, middleware, storage resources, and network resources in industrial cloud platforms, the operation monitoring of basic resource services typically requires the aggregation of operational alarms, error status events, instance lifecycle records, and resource dependency change records, forming a comprehensive operational overview through resource instance nodes and resource dependencies. To adapt to operational modes such as rolling releases, fault reconstruction, elastic scaling, and cross-node migration, the same logical resource may generate multiple resource instances with different resource instance identifiers, host nodes, operating spaces, and effective periods at different operational stages, and previous and subsequent resource instances may run in parallel during switching. Since different monitoring records are generated by monitoring acquisition components, resource management components, and service management components respectively, the time when monitoring events enter the platform may be later than their actual occurrence time. When a monitoring event does not carry a resource instance identifier that can uniquely point to the original resource instance, the platform usually determines its initial affiliation based on the mapping relationship between the logical resource identifier and the currently running resource instance.
[0004] When monitoring events generated by previous generation resource instances are delayed in entering the platform, while subsequent generation resource instances have already started and established new resource dependencies, the monitoring events are easily associated with the subsequent generation resource instances and displayed along the current resource dependency topology of the subsequent generation resource instances. This may result in situations where the occurrence time of the monitoring event is earlier than the instance startup time of the currently associated resource instance and earlier than the establishment time of the corresponding resource dependency relationship. Historical anomalies of previous generation resource instances may be incorrectly configured in subsequent generation resource instances, and the valid resource dependencies at the time of the event may be replaced by the current resource dependencies, thus forming a reverse-flowing anomaly chain between instance generations. This anomaly chain mixes the current operating state with the resource state at the time of the anomaly, causing the overall operating view to fail to accurately reflect the actual resource instance to which the monitoring event belongs and the objectively existing resource dependency chain at the time of the event, affecting the verification of basic resource anomaly information and the accurate monitoring of the operating status of the industrial cloud platform. Summary of the Invention
[0005] To address the aforementioned issues, this disclosure proposes a panoramic monitoring method and system for industrial infrastructure resource services. By establishing a generational genealogy of resource instances and a temporal resource dependency set, it identifies temporal logical conflicts in the current instance affiliation and current dependency connections of monitored events. Simultaneously, it restores the actual resource instance to which the monitored event belongs and the dependency links at the time of the event, accurately displaying the actual resource instance to which the monitored event belongs and the objectively existing resource dependencies at the time of the event. This improves the authenticity, completeness, and traceability of the panoramic view of infrastructure resource service operation, providing a reliable basis for anomaly verification and operation and maintenance of industrial cloud platforms.
[0006] According to some embodiments, the present disclosure adopts the following technical solutions: A method for panoramic monitoring of the operation of industrial infrastructure resource services includes: Acquire monitoring events, instance lifecycle records, and dependency change records from the industrial cloud platform; Establish the generational genealogy of resource instances for the same logical resource and the effective period of the instance based on the instance lifecycle record; establish the temporal resource dependency set and the current resource dependency topology based on the dependency change record. Based on the generational genealogy of resource instances, the validity period of instances, and the current resource dependency topology, identify the pre-birth abnormal chain of instance generational backflow; For instance generational backflow type pre-birth anomaly chain, search along the generational genealogy of resource instances to find the previous generation resource instance whose event occurred during the instance's valid period, and determine the actual resource instance to which it belongs; The event occurrence time is selected from the temporal resource dependency set and the dependency relationship is connected to the actual resource instance, forming the dependency link when the event occurs; The system generates a panoramic monitoring result based on the actual resource instance, the dependency link at the time of the event, and the current resource dependency topology.
[0007] As one embodiment, the acquisition of monitoring events, instance lifecycle records, and dependency change records of the industrial cloud platform includes: Record the logical resource identifier, currently associated resource instance identifier, instance location information, and event occurrence time of the monitored event; Record the logical resource identifier, resource instance identifier, instance location information, instance start time, and instance termination time of the instance lifecycle. Record the dependency initiating instance identifier, dependency target instance identifier, dependency establishment time, and dependency termination time of dependency relationship changes; The instance location information includes the host node identifier and the running space identifier, and the instance lifecycle record includes the instance startup record, instance stop record, instance destruction record, and instance switchover record.
[0008] As one embodiment, establishing the resource instance generational lineage and instance validity period for the same logical resource based on the instance lifecycle record includes: Group resource instances according to their logical resource identifiers; Based on the instance switching record, the resource instance identifier of the resource instance to be replaced is recorded as the previous generation resource instance identifier of the resource instance that replaces the resource instance, and the resource instances in the same group are arranged according to the instance startup time. The effective period of an instance is defined as the time from instance startup to instance termination. For resource instances that do not record the instance termination time but record the instance switchover completion time, the instance switchover completion time shall be determined as the instance termination time. For resource instances that are still running, extend the instance's validity period to the current moment when the method is executed; For resource instances running in parallel, the corresponding instance validity period is retained.
[0009] As one embodiment, establishing a temporal resource dependency set and the current resource dependency topology based on dependency change records includes: Combine the dependency initiating instance identifier, the dependency target instance identifier, the dependency establishment time, and the dependency termination time into a temporal resource dependency relationship; The effective period of the temporal resource dependency is defined as the time from when the dependency is established to when the dependency is terminated. For temporal resource dependencies for which the release time is not recorded, extend their validity period to the current moment; Select the temporal resource dependencies that are within the valid period at the current time to form the current resource dependency topology.
[0010] As one embodiment, identifying instance-intergenerational backflow-type prenatal abnormality chains includes: Based on the current associated resource instance identifier, read the instance start time of the current associated resource instance from the generational genealogy of the resource instances; If the event occurs before the instance startup time, it is determined that the monitoring event has a time conflict before the instance was created. Read the dependencies connected to the current associated resource instance from the current resource dependency topology. If the event occurs earlier than the dependency establishment time of at least one of the aforementioned dependencies, determine that there is a pre-establishment propagation conflict in the corresponding dependency. When a previous generation of a resource instance with the same logical resource identifier exists in the current associated resource instance, and the monitoring event simultaneously exhibits both pre-instance birth time conflict and pre-establishment propagation conflict, the monitoring event, the current associated resource instance, and the dependency relationship with the pre-establishment propagation conflict are identified as an instance generation backflow type pre-instance birth anomaly chain.
[0011] As one embodiment, for an instance generational backflow type pre-birth anomaly chain, the process involves tracing the generational genealogy of resource instances to find previous generation resource instances whose event occurrence time falls within the instance's valid period, thereby determining the actual resource instance to which it belongs, including: Read the previous generation resource instances sequentially along the previous generation resource instance identifiers of the current associated resource instance; Select a previous generation resource instance that has the same logical resource identifier as the currently associated resource instance, whose event occurrence time falls within the corresponding instance's valid period, and whose instance location information is the same as the instance location information of the monitoring event; When only one predecessor resource instance meets the selection criteria, the predecessor resource instance is determined as the actual resource instance to which the monitoring event belongs. The monitoring event is then reassigned from the current associated resource instance to the actual resource instance, and the connection of the monitoring event in the current resource dependency topology along the dependency relationship that has a pre-propagation conflict is canceled.
[0012] As one embodiment, the step of filtering dependencies from the temporal resource dependency set whose event occurrence time falls within the valid period and is connected to the actual resource instance to form the dependency chain at the time of event occurrence includes: Using the event occurrence time of the monitored event as the filtering time, select temporal resource dependencies from the temporal resource dependency set that are connected to the actual resource instance, whose event occurrence time falls within the corresponding valid period, and whose dependent target instance is within the instance's valid period during the event occurrence time; The selected temporal resource dependencies are connected to form a dependency chain when the event occurs.
[0013] According to some embodiments, the present disclosure adopts the following technical solutions: A panoramic monitoring system for the operation of industrial infrastructure resource services includes: The monitoring record acquisition module is used to acquire monitoring events, instance lifecycle records, and dependency change records of the industrial cloud platform. The instance generation construction module is used to establish the resource instance generational lineage and instance validity period of the same logical resource based on the instance lifecycle record. The temporal dependency building module is used to establish a temporal resource dependency set and the current resource dependency topology based on dependency change records. The anomaly chain identification module is used to identify pre-birth anomaly chains with generational backflow based on the generational lineage of resource instances, the validity period of instances, and the current resource dependency topology. The synchronous backtracking module is used to find the previous generation resource instance whose event occurrence time falls within the instance's valid period and determine the actual resource instance it belongs to by searching the generational backflow anomaly chain of instances. It also filters the dependencies whose event occurrence time falls within the valid period and are connected to the actual resource instance it belongs to from the temporal resource dependency set, forming the dependency chain at the time of the event occurrence. The panoramic generation module is used to generate panoramic monitoring results based on the actual resource instance, the dependent links at the time of the event, and the current resource dependency topology.
[0014] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned method for panoramic monitoring of the operation of industrial infrastructure resource services.
[0015] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform a panoramic monitoring method for the operation of industrial infrastructure resource services.
[0016] Compared with the prior art, the beneficial effects of this disclosure are as follows: This disclosure discloses a panoramic monitoring method for industrial infrastructure resource services. It establishes a generational genealogy of resource instances and their effective periods based on instance lifecycle records, and establishes a temporal resource dependency set and current resource dependency topology based on dependency change records. This allows different generations of resource instances and their corresponding dependencies to be recorded separately according to their actual effective time. When the occurrence time of a monitoring event is earlier than the instance startup time of the currently associated resource instance and earlier than the dependency establishment time of the target dependency, the monitoring event, the currently associated resource instance, and the target dependency are identified as an instance generational backflow-type pre-birth anomaly chain. This simultaneously detects time conflicts between the monitoring event instance affiliation and the resource instance lifecycle, as well as cross-generational connection conflicts between the monitoring event and the current resource dependency, avoiding incorrect affiliation of delayed monitoring events based solely on logical resource identifiers or current operating status.
[0017] This disclosure discloses a panoramic monitoring method for industrial infrastructure resource services. For identified instance-generational backflow anomaly chains before birth, the method traces back along the generational genealogy of resource instances to the previous generation resource instance corresponding to the event occurrence time. Combining the instance's validity period and instance location information, it determines the actual resource instance to which the monitoring event belongs, cancels erroneous connections formed along the target dependency relationship of the monitoring event, and recovers the actually valid dependency links at the time of the event from the temporal resource dependency relationship set. Based on this, it generates a current operational panoramic view and a panoramic view at the time of the anomaly, respectively. This method can distinguish between the current resource status and the historical resource status at the time of the monitoring event, accurately displaying the actual resource instance to which the monitoring event belongs and the objectively existing resource dependencies at the time of the event. This improves the authenticity, completeness, and traceability of the panoramic view of infrastructure resource service operation, providing a reliable basis for anomaly verification and operation maintenance of industrial cloud platforms. Attached Figure Description
[0018] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.
[0019] Figure 1 This is a flowchart of a method for panoramic monitoring of the operation of industrial infrastructure resource services according to an embodiment of the present disclosure; Figure 2 This is a structural diagram of a panoramic monitoring system for the operation of industrial infrastructure resource services, according to an embodiment of this disclosure. Detailed Implementation
[0020] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.
[0021] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.
[0022] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0023] Example 1 One embodiment of this disclosure provides a method for panoramic monitoring of the operation of industrial basic resource services. This method is executed by the management node of an industrial cloud platform or by a monitoring server connected to the industrial cloud platform. It is applicable to computing resource services, virtual machine services, container services, database services, middleware services, storage services, and network services in the industrial cloud platform. During operation, the basic resource services can form specific resource instances, and the industrial cloud platform records the start-up, stop, destruction, switching, and changes in dependencies of the resource instances.
[0024] This disclosure specifically addresses the following operational scenario: During the rolling release, fault reconstruction, elastic scaling, or cross-node migration of the same logical resource, a predecessor resource instance and a successor resource instance are formed successively; the monitoring event generated by the predecessor resource instance does not carry a resource instance identifier that can uniquely point to the predecessor resource instance, and enters the industrial cloud platform after the successor resource instance is established; the industrial cloud platform initially associates the monitoring event with the successor resource instance based on the mapping relationship between the logical resource identifier and the currently running resource instance identifier, and displays the monitoring event according to the current resource dependency topology corresponding to the successor resource instance, thereby forming an instance generation backflow type pre-birth anomaly chain where the event occurrence time is earlier than the successor resource instance startup time and earlier than the establishment time of at least one current dependency relationship.
[0025] As one embodiment, the present disclosure of a panoramic monitoring method for industrial infrastructure resource services establishes a generational genealogy of resource instances and a set of temporal resource dependencies, identifies temporal logical conflicts in the current instance affiliation and current dependency connections of a monitored event, and synchronously restores the actual resource instance to which the monitored event belongs and the dependency links at the time the event occurred.
[0026] In this embodiment, the event occurrence time of the monitored event, the instance start time and instance termination time of the resource instance, and the dependency establishment time and dependency release time of the dependency relationship all adopt the unified time base of the industrial cloud platform; the monitoring acquisition component, resource management component and dependency acquisition component keep time synchronized with the unified time source provided by the industrial cloud platform control node; when each component generates a record, the corresponding time field is converted into a timestamp that is continuously counted from the same time starting point; Before comparing execution times, the event occurrence time, instance start time, instance termination time, dependency establishment time, and dependency dissolution time are converted to the same time zone and time precision. For cases where the original records have different time precisions, the time fields involved in the comparison are unified to the time precision supported by all the original time fields, without changing the original time order by adding low-order time data that is not present in the original record. For records that lack valid time fields, whose time fields cannot be converted to the unified time base, or whose components that generated the record have not completed time synchronization, the record is not used to determine time conflicts before the instance is born or propagation conflicts before the dependency is established. In this embodiment, the "current moment" refers to the panoramic snapshot moment corresponding to when the industrial cloud platform generates the current resource dependency topology according to the unified time base; whether resource instances and temporal resource dependencies are valid at the current moment is determined based on this panoramic snapshot moment. Specifically, the detailed implementation of the method disclosed herein is as follows: Step 1: Obtain monitoring events, instance lifecycle records, and dependency change records from the industrial cloud platform; Specifically, the industrial cloud platform obtains monitoring events, instance lifecycle records, and dependency change records through monitoring and acquisition components, resource management components, and dependency acquisition components, respectively. The monitoring and acquisition components are deployed on host nodes, virtual machines, containers, or basic resource service processes to collect the operation records generated by the basic resource services during operation.
[0027] The monitoring events described in this embodiment are operation records that have been assigned abnormal identifiers by the monitoring and acquisition component during the acquisition phase, including operation alarms, error status events, and operation logs with error level identifiers or alarm level identifiers.
[0028] This embodiment does not re-determine whether the running records are abnormal. Instead, it takes the monitoring events that the monitoring and acquisition components have already identified as abnormal as the processing objects, performs time logic checks on the resource instances and dependencies currently associated with the monitoring events, and corrects the resource instance affiliation and dependency links at the time of the event after identifying the instance generation backflow type pre-birth abnormal chain. General running logs, normal start-stop events, and normal state change records without abnormal identifiers are not considered as monitoring events in this embodiment. However, normal start-stop events can be used as instance lifecycle records, and normal dependency activation events and dependency failure events can be used as dependency change records.
[0029] For each monitoring event, at least the following information should be recorded: logical resource identifier, currently associated resource instance identifier, instance location information, and event occurrence time. The logical resource identifier represents a logical object that continuously provides the same basic resource service across different operational phases; for example, the same database service may form two resource instances before and after instance reconstruction. These two resource instances have the same logical resource identifier but different resource instance identifiers. The event occurrence time represents the actual time the monitoring event occurred on the underlying resource side that generated the event; the event occurrence time is obtained from the source time field, source log time field, or source alarm time field of the monitoring event, and the time when the monitoring event enters the industrial cloud platform is not used as a substitute for the event occurrence time.
[0030] Furthermore, the formation process of the current associated resource instance identifier is as follows: the current associated resource instance identifier represents the initial resource instance identifier assigned to the monitoring event by the industrial cloud platform before the monitoring event enters this method; the current associated resource instance identifier only indicates the initial attribution of the monitoring event, and does not indicate that the initial attribution is necessarily correct; when the monitoring event enters the industrial cloud platform, the event access component reads the logical resource identifier in the monitoring event, queries the mapping relationship between the logical resource identifier and the currently running resource instance identifier maintained by the industrial cloud platform when the event is accessed, and writes the queried currently running resource instance identifier into the monitoring event as the current associated resource instance identifier; When a monitoring event already carries a resource instance identifier that uniquely points to a specific resource instance, the event access component retains the resource instance identifier carried by the monitoring event and no longer maps the monitoring event to other resource instances based on the logical resource identifier. If a monitoring event generated by a previous generation resource instance does not carry a resource instance identifier that uniquely points to the previous generation resource instance, and the monitoring event enters the industrial cloud platform after the establishment of a descendant resource instance, the currently running resource instance queried by the event access component based on the logical resource identifier is the descendant resource instance. Therefore, the monitoring event generated by the previous generation resource instance may be initially associated with the descendant resource instance.
[0031] Further, the process of obtaining instance location information is as follows: The instance location information includes a host node identifier and a runtime space identifier; the host node identifier is used to represent the physical server, virtual machine, or compute node where the resource instance resides; the runtime space identifier is used to represent the cluster space, tenant space, namespace, or container runtime space where the resource instance resides; the host node identifier and runtime space identifier of the monitoring event are read from the resource context of the resource instance that generated the monitoring event and written into the monitoring event by the monitoring acquisition component when the monitoring event is generated; when tracing back to the actual resource instance to which the monitoring event belongs in the subsequent process, the instance location information of the monitoring event is determined to be the same as the instance location information of the previous generation resource instance only when both the host node identifier and the runtime space identifier of the monitoring event exist and are the same as the corresponding fields of the queried previous generation resource instance; When a monitoring event lacks a host node identifier or a runtime space identifier, or when any identifier differs from the corresponding field of the queried predecessor resource instance, a deterministic relationship between the monitoring event and the resource instance is not formed based on the predecessor resource instance.
[0032] The process of obtaining the instance lifecycle record in this embodiment includes: Resource management components are used to record the creation, startup, shutdown, destruction, and switching processes of resource instances; resource management components can be resource orchestration components, container management components, virtual machine management components, or service instance management components of industrial cloud platforms.
[0033] For each instance lifecycle record, its data attributes include at least the following: logical resource identifier, resource instance identifier, instance location information, instance start time, and instance termination time; the resource instance identifier is a unique identifier assigned to a specific running instance by the industrial cloud platform; the predecessor and successor resource instances corresponding to the same logical resource have the same logical resource identifier and different resource instance identifiers; the host node identifier and runtime space identifier in the instance lifecycle record are recorded by the resource management component when creating or starting the resource instance; the instance start time indicates the time when the resource instance begins to be in a runnable state; the instance termination time indicates the time when the resource instance stops running, is destroyed, or completes service switchout.
[0034] The types of instance lifecycle records include instance startup records, instance shutdown records, instance destruction records, and instance switching records; the instance switching record includes at least the identifier of the replaced resource instance, the identifier of the replaced resource instance, and the instance switching completion time, which are used to determine the previous generation relationship between resource instances and to determine the instance termination time of the replaced resource instance when there is no explicit instance termination time.
[0035] The process of obtaining dependency change records in this implementation includes: The dependency acquisition component obtains dependency establishment and termination records between resource instances from the resource management or service management components of the industrial cloud platform. In this embodiment, only the relationship change records explicitly generated by the industrial cloud platform are used to establish temporal resource dependencies between resource instances. The dependency establishment record includes at least the dependency initiating instance identifier, the dependency target instance identifier, and the relationship effective time. The dependency termination record includes at least the dependency initiating instance identifier, the dependency target instance identifier, and the relationship expiration time corresponding to the dependency establishment record. The dependency establishment time is determined by the relationship effectiveness time in the dependency establishment record, and the dependency termination time is determined by the relationship expiration time in the dependency termination record. Individual records with similar names, records with only time proximity, and single runtime log entries that cannot simultaneously determine the dependency initiating instance and the dependency target instance are not used to establish temporal resource dependencies. The dependencies include resource hosting relationships and service binding relationships. For resource hosting relationships, the time when the resource instance completes deployment binding with the host resource is determined by the dependency establishment time, and the time when the resource instance is unbound from deployment binding or terminates operation is determined by the dependency termination time. For service binding relationships, the effectiveness time in the service binding effectiveness record is determined by the dependency establishment time, and the termination time in the service binding termination record is determined by the dependency termination time.
[0036] For each dependency change record, at least the following information should be recorded: dependency initiating instance identifier, dependency target instance identifier, dependency establishment time, and dependency termination time.
[0037] Finally, the acquired monitoring events, instance lifecycle records, and dependency change records are combined to form a basic resource operation record set. The basic resource operation record set can be stored in the same data storage device, or it can be stored separately according to monitoring events, instance lifecycle records, and dependency change records, as long as it can be queried according to logical resource identifier, resource instance identifier, and time field.
[0038] Step 2: Establish the generational lineage of resource instances for the same logical resource and the effective period of the instances based on the instance lifecycle records; Specifically, instance lifecycle records are read from the basic resource operation record set, and resource instances are grouped according to logical resource identifiers; resource instances with the same logical resource identifier are grouped into the same logical resource group; different logical resource identifiers correspond to different logical resource groups, and no generational relationship between resource instances is established between different logical resource groups; for the same logical resource group, instance switching records are read from the instance lifecycle records. When the instance switching record indicates that the second resource instance takes over the corresponding basic resource service from the first resource instance, the first resource instance is identified as the predecessor resource instance of the second resource instance, and the resource instance identifier of the first resource instance is recorded as the predecessor resource instance identifier of the second resource instance. When the same logical resource is replaced multiple times, the resource instances are connected according to the succession order determined by the instance switching record. For example, if the first resource instance is replaced by the second resource instance, and the second resource instance is replaced by the third resource instance, a generational genealogy of resource instances is formed, in which the first resource instance, the second resource instance, and the third resource instance are connected in sequence.
[0039] Instance startup time is used to arrange resource instances within the same logical resource group, but does not determine the predecessor relationship between resource instances solely based on instance startup time; for multiple resource instances running simultaneously due to horizontal scaling, if there is no instance switching record where one resource instance replaces another, the multiple resource instances are recorded as parallel resource instances under the same logical resource, and no predecessor relationship is established between parallel resource instances.
[0040] For each resource instance in the resource instance generational genealogy, the instance validity period is determined based on the instance start time and instance end time. When the instance lifecycle record has a clearly defined instance end time, the instance start time is used as the start time of the instance validity period, and the instance end time is used as the end time of the instance validity period. When the instance lifecycle record does not record an instance end time, but the instance switchover record records an instance switchover completion time, and the instance switchover completion record indicates that the previous generation resource instance has stopped carrying the corresponding basic resource service, the instance switchover completion time is determined as the instance end time of the previous generation resource instance. When a descendant resource instance has started, but the instance switchover record does not indicate that the previous generation resource instance has stopped carrying the corresponding basic resource service, the instance start time of the descendant resource instance is not used as the instance end time of the previous generation resource instance. When new and old resource instances run in parallel during the rolling release process, the validity period of each instance is retained. For example, if the later resource instance starts at 10:12 and the earlier resource instance terminates at 10:15, then the period from 10:12 to 10:15 is the overlapping period of the validity periods of the two resource instances. When a resource instance is still running and there is no instance termination record, the validity period of the resource instance is extended to the current time. Each resource instance in the resource instance generational genealogy corresponds to at least the following: resource instance identifier, logical resource identifier, instance location information, instance start time, instance end time, and the identifier of the previous generation resource instance.
[0041] Step 3: Establish a temporal resource dependency set and the current resource dependency topology based on dependency change records; Specifically, dependency change records are read from the basic resource operation record set; the dependency initiating instance identifier, dependency target instance identifier, dependency establishment time and dependency termination time in the same dependency change record are combined to form a temporal resource dependency; a temporal resource dependency corresponds to two specific resource instances, rather than just two logical resources; When a previous generation resource instance is replaced by a subsequent generation resource instance, the dependency relationship formed between the subsequent generation resource instance and the dependent target instance is recorded as a new temporal resource dependency relationship, which does not cover the temporal resource dependency relationship between the previous generation resource instance and the original dependent target instance; the time from the establishment of the dependency relationship to the time of the termination of the dependency relationship is determined as the effective period of the temporal resource dependency relationship. When the dependency change record does not record the dependency termination time, and both the dependency initiating instance and the dependency target instance are currently within their instance validity period, the validity period of the temporal resource dependency is extended to the current time. When the dependency change record does not record the dependency termination time, but the dependency initiating instance or the dependency target instance has terminated, the earlier of the dependency initiating instance's instance termination time and the dependency target instance's instance termination time is determined as the end time of the validity period of the temporal resource dependency. When the dependency termination time is later than the dependency initiating instance's instance termination time or the dependency target instance's instance termination time, the earlier of the dependency termination time, the dependency initiating instance's instance termination time, and the dependency target instance's instance termination time is taken as the end time of the validity period of the temporal resource dependency. Thus, the validity period of the temporal resource dependency is constrained by both the dependency change record and the instance validity period of the connected resource instance.
[0042] When the dependency relationship is established earlier than the instance startup time of the dependency initiating instance or the instance startup time of the dependency target instance, the dependency relationship is not valid in terms of time logic. The dependency relationship will not be added to the temporal resource dependency relationship set, nor will it be used for subsequent instance generation backflow type pre-birth anomaly chain identification. Filter the temporal resource dependencies in the current time frame; If the current moment falls within the valid period of a temporal resource dependency, and both the dependency initiating instance and the dependency target instance connected by the temporal resource dependency are within their instance validity periods at the current moment, then the temporal resource dependency is determined as a currently valid dependency. Each currently valid dependency is connected according to the dependency initiating instance identifier and the dependency target instance identifier to form the current resource dependency topology. The current resource dependency topology reflects the resource instances and resource dependencies in the industrial cloud platform at the current moment. The set of temporal resource dependencies retains valid resource dependencies from different historical time periods.
[0043] Step 4: Identify the pre-birth anomaly chain of instance generational backflow based on the generational lineage of resource instances, the validity period of instances, and the current resource dependency topology; Specifically, monitoring events are first read from the basic resource operation log set, including: For the currently read monitoring event, first check whether the event occurrence time, the instance startup time of the currently associated resource instance, and the dependency establishment time of the corresponding dependency relationship in the current resource dependency topology have been converted to a unified time base. When all the above times have valid values and can be compared with the same time precision, continue to perform the identification of pre-instance birth time conflicts and pre-dependency propagation conflicts. When any time is missing or cannot be converted to a unified time base, do not form a deterministic pre-instance birth anomaly chain of intergenerational backflow based on the monitoring event.
[0044] Secondly, identify prenatal time conflicts in the instances: Based on the current associated resource instance identifier in the monitoring event, the current associated resource instance is queried in the resource instance generational genealogy, and its logical resource identifier, instance startup time, and predecessor resource instance identifier are read. The event occurrence time of the monitoring event is compared with the instance startup time of the current associated resource instance. If the event occurrence time is not earlier than the instance startup time of the current associated resource instance, a pre-birth time conflict is not determined based on this comparison. If the event occurrence time is earlier than the instance startup time of the current associated resource instance, a pre-birth time conflict is determined to exist in the monitoring event. The pre-birth time conflict means that, according to the current association relationship of the monitoring event, the monitoring event belongs to the current associated resource instance; however, when the monitoring event actually occurs, the current associated resource instance has not yet started, so the current associated resource instance could not have actually generated the monitoring event.
[0045] Then, identify pre-propagation conflicts in dependency relationships, including: After determining that a monitoring event has a time conflict before the instance is created, the dependencies connected to the currently associated resource instance are read from the current resource dependency topology. For each read dependency, the dependency establishment time is read from the temporal resource dependency set, and the event occurrence time is compared with the dependency establishment time. When the event occurrence time is not earlier than the dependency establishment time of the corresponding dependency, a pre-establishment propagation conflict is not determined based on the dependency. When the event occurrence time is earlier than the dependency establishment time of the corresponding dependency, a pre-establishment propagation conflict is determined for the dependency. The pre-establishment propagation conflict means that the current running panorama connects the monitoring event to the currently associated resource instance and displays it along the current dependencies of the currently associated resource instance; however, when the monitoring event actually occurs, the dependency has not yet been established, so the dependency connection cannot be established at the time of the event.
[0046] Finally, the specific formation of the intergenerational reverse-flow prenatal abnormality chain is as follows: Read the identifier of the previous generation resource instance of the current associated resource instance; if the current associated resource instance does not have a previous generation resource instance, the monitoring event is not identified as an instance generation backflow type pre-birth anomaly chain; if the current associated resource instance has a previous generation resource instance, and the previous generation resource instance has the same logical resource identifier as the current associated resource instance, and the monitoring event has an instance birth time conflict, and at least one dependency relationship in the current resource dependency topology connected to the current associated resource instance has a dependency relationship establishment pre-propagation conflict, the monitoring event, the current associated resource instance, and the dependency relationship with the dependency relationship establishment pre-propagation conflict are jointly identified as an instance generation backflow type pre-birth anomaly chain; This step does not simply correct the attribution based on the fact that the monitored event occurred before the startup time of the current associated resource instance. Instead, it simultaneously confirms that the monitored event has formed a dependency connection in the current resource dependency topology that did not exist at the time of the event. This combines the identification of instance generational errors and topological generational errors. The instance generational backflow type pre-birth anomaly chain identified in this step indicates that the current resource instance attribution and the current dependency connection of the monitored event cannot be established in terms of time logic. It does not determine that the monitored event has already propagated causally along the dependency relationship based solely on the resource dependency relationship. The identified instance generational backflow type pre-birth anomaly chain is formed into an instance generational backflow type pre-birth anomaly chain candidate set; the instance generational backflow type pre-birth anomaly chain candidate set is used in step five to perform the actual resource instance backtracking and the dependency link recovery when the event occurs.
[0047] Step 5: For the pre-birth anomaly chain of the generational backflow of instances, search for the previous generation resource instances along the generational genealogy of resource instances whose event occurrence time falls within the valid period of the instance to determine the actual resource instance to which it belongs; filter the dependencies that fall within the valid period of the event occurrence time and are connected to the actual resource instance to which it belongs from the temporal resource dependency set to form the dependency chain at the time of the event occurrence. Specifically, for each monitoring event in the candidate set of pre-birth anomaly chains with intergenerational backflow, the logical resource identifier, the currently associated resource instance identifier, the instance location information, and the event occurrence time of the monitoring event are read.
[0048] First, based on the above information, a backtracking process is performed to determine the actual resource instance to which it belongs, including: Based on the identifier of the predecessor resource instance of the current associated resource instance, read the direct predecessor resource instance of the current associated resource instance; if the event time does not fall within the instance validity period of the direct predecessor resource instance, continue to read the predecessor resource instance of the direct predecessor resource instance until the resource instances before the current associated resource instance are traversed, or a predecessor resource instance that meets the certain conditions is found. For each read predecessor resource instance, the following conditions are checked in sequence: whether the predecessor resource instance and the currently associated resource instance have the same logical resource identifier; whether the occurrence time of the monitoring event falls within the instance validity period of the predecessor resource instance; and whether the instance location information of the monitoring event is the same as the instance location information of the predecessor resource instance. The same instance location information means that the host node identifier of the monitoring event is the same as the host node identifier of the previous generation resource instance, and the running space identifier of the monitoring event is the same as the running space identifier of the previous generation resource instance. Instance location information comparison is performed only when both the host node identifier and the runtime space identifier of the monitored event exist. When only one predecessor resource instance meets the above conditions, the predecessor resource instance is determined as the actual resource instance to which the monitored event belongs, and the monitored event is reassigned from the current associated resource instance to the actual resource instance, forming a corrected monitoring event and resource instance affiliation relationship. When no predecessor resource instance meets the above conditions, no deterministic resource instance affiliation correction is performed on the monitored event. When the monitored event lacks a host node identifier or a runtime space identifier, no deterministic resource instance affiliation correction is performed on the monitored event. When the instance validity periods of multiple resource instances cover the event occurrence time, but only one resource instance belongs to the predecessor resource instance of the current associated resource instance, and the instance location information of the predecessor resource instance is the same as the instance location information of the monitored event, the predecessor resource instance is determined as the actual resource instance to which the monitored event belongs. Monitoring events that cannot uniquely identify the actual resource instance to which they belong retain their current association and candidate status of the pre-birth anomaly chain of the instance generational backflow type, and are not forcibly assigned to any previous generation resource instance.
[0049] Secondly, the dependency chain is constructed when the event occurs, specifically as follows: After determining the actual resource instance to which the monitoring event belongs, the event occurrence time is used as the filtering time for temporal resource dependencies; temporal resource dependencies that are directly connected to the actual resource instance are selected from the set of temporal resource dependencies. For each selected temporal resource dependency, the following conditions are determined: First, whether the event occurrence time falls within the valid period of the temporal resource dependency; Second, whether the target resource instance connected to the temporal resource dependency is within the instance's valid period at the event occurrence time; Only temporal resource dependencies that simultaneously meet the following conditions are retained: the temporal resource dependency is connected to the actual resource instance to which it belongs; the event occurrence time falls within the valid period of the temporal resource dependency; and the target resource instance connected to the temporal resource dependency is within the instance's valid period at the event occurrence time. The retained temporal resource dependencies are combined into an event-time dependency chain according to the connection relationship between the actual resource instance and the target resource instance; dependencies in the current resource dependency topology whose establishment time is later than the event occurrence time will not be included in the event-time dependency chain.
[0050] Furthermore, construct a multi-level dependency chain when events occur, including: When it is necessary to form a multi-level event occurrence dependency chain, the actual resource instance is used as the starting resource instance. The temporal resource dependency relationship that the event occurrence time falls within the valid period, is directly connected to the starting resource instance, and the target resource instance is within the instance's valid period when the event occurs is selected from the temporal resource dependency relationship set. The corresponding target resource instance is then determined as the next-level resource instance. For each next-level resource instance, continue to select temporal resource dependencies that meet the above conditions according to the same event occurrence time, and add the newly selected target resource instance to the dependency chain at the event occurrence time; Resource instances already added to the event-time dependency chain are not traversed again to avoid repeatedly forming the same resource instance node when there is a closed loop in the resource dependency relationship; when there is no temporal resource dependency relationship that has not yet been selected and is within the valid period of the event occurrence time, the formation of the event-time dependency chain is stopped; the event-time dependency chain is used to represent the resource dependency relationship that objectively exists when the monitoring event occurs, and the event-time dependency chain is not directly determined as the abnormal causal propagation chain; this step outputs the corrected monitoring event and resource instance attribution relationship and the event-time dependency chain.
[0051] Step Six: Generate a comprehensive monitoring result based on the actual resource instance, the dependency link at the time of the event, and the current resource dependency topology; Specifically, based on the generational lineage of resource instances, the current resource dependency topology, the corrected relationship between monitoring events and resource instance affiliation, and the dependency links at the time of the event, a panoramic monitoring result of the basic resource service operation is generated; the panoramic monitoring result of the basic resource service operation includes the current operation panorama and the panorama at the time of the anomaly.
[0052] First, the process of generating the currently running panorama is as follows: Select the resource instance that is currently in the valid period from the generational genealogy of resource instances, and set the selected resource instance as the current resource instance node; connect each current resource instance node with the currently valid dependency relationship in the current resource dependency topology to form the current running panorama; the current running panorama is used to represent the resource instances and resource dependencies that are currently in a valid state.
[0053] Secondly, the process of generating the panorama when an anomaly occurs is as follows: The actual resource instance to which the monitored event belongs is set as the resource instance node to which the monitored event belongs in the panorama when the anomaly occurs; a dependency link is used to connect the actual resource instance and the target resource instance that is in a valid state when the event occurs, forming the panorama when the anomaly occurs; for resource instances with the same logical resource identifier but belonging to different generations, independent resource instance nodes are established for each, and the previous generation resource instances and the next generation resource instances are not merged into the same node according to the logical resource identifier; for monitoring events that have been traced back through resource instance affiliation, the monitoring event is removed from the resource instance node corresponding to the currently associated resource instance and set to the resource instance node corresponding to the actual resource instance; for dependencies with pre-propagation conflicts in the current resource dependency topology, the connection formed by the monitoring event along the dependency relationship is not displayed in the panorama when the anomaly occurs; When an anomaly occurs, the resource instance nodes and resource dependencies in the panorama are all judged based on the time of the event. Therefore, the panorama when an anomaly occurs reflects the resource instances and resource dependencies that are in a valid state when the monitoring event actually occurs, rather than the resource instances and current resource dependencies that are in a valid state when the panorama is generated. When an anomaly occurs, the dependency link in the panorama only indicates that there is a valid dependency relationship between the corresponding resource instances when the monitoring event occurs. It does not take this as a basis to conclude that the monitoring event has propagated causally along the dependency relationship. By restoring the monitoring event to the actual resource instance to which it belongs and excluding the dependency relationship that was not established when the event occurred, the cross-instance generational error connection in the current running panorama is eliminated. In this way, the panorama monitoring of the basic resource service operation is completed.
[0054] Example 2 One embodiment of this disclosure takes logical resource service A in an industrial cloud platform as an example to specifically illustrate the identification of pre-birth anomaly chains of intergenerational backflow of instances, the backtracking of resource instance ownership, and the recovery process of dependent links when an event occurs.
[0055] Logical resource service A is a basic middleware service in the industrial cloud platform, and its logical resource identifier is "SRV-A"; Logical resource service A successively formed a first resource instance and a second resource instance; the resource instance identifier of the first resource instance is "A-001", the host node identifier is "NODE-01", and the runtime space identifier is "SPACE-01"; the instance start time of the first resource instance is 9:00 on June 18, 2026, and the instance termination time is 10:15 on the same day; the resource instance identifier of the second resource instance is "A-002", the host node identifier is "NODE-02", and the runtime space identifier is "SPACE-02"; the instance start time of the second resource instance is 10:12 on the same day; The instance switching record shows that the second resource instance took over the logical resource service A from the first resource instance and completed the instance switching at 10:15; therefore, the first resource instance "A-001" is identified as the predecessor resource instance of the second resource instance "A-002". The first resource instance and the second resource instance run in parallel between 10:12 and 10:15; the generational genealogy of the resource instances retains the instance validity period of the first resource instance from 9:00 to 10:15, and the instance validity period of the second resource instance starting from 10:12.
[0056] The industrial cloud platform also runs a first database instance and a second database instance. The resource instance of the first database instance is identified as "D-001", and its validity period covers the period from 9:00 to 10:15. The resource instance of the second database instance is identified as "D-002", and its instance started at 10:11 and is still running at the current time. The dependency establishment record indicates that the service binding relationship between the first resource instance "A-001" and the first database instance "D-001" took effect at 9:05 AM; the dependency termination record indicates that the service binding relationship between the first resource instance "A-001" and the first database instance "D-001" was terminated at 10:15 AM; thus, a first temporal resource dependency relationship is established; the dependency initiating instance of the first temporal resource dependency relationship is identified as "A-001", the dependency target instance is identified as "D-001", and the effective period is from 9:05 AM to 10:15 AM; The dependency establishment record also indicates that the service binding relationship between the second resource instance "A-002" and the second database instance "D-002" took effect at 10:13 and the release time has not yet been recorded; thus, a second temporal resource dependency relationship is established; the dependency initiating instance of the second temporal resource dependency relationship is identified as "A-002", the dependency target instance is identified as "D-002", and the effective period begins at 10:13.
[0057] Furthermore, the first resource instance generated a monitoring event with an error level identifier at 10:08; The logical resource identifier of this monitoring event is "SRV-A", the host node identifier is "NODE-01", the runtime space identifier is "SPACE-01", and the event occurred at 10:08. This monitoring event does not carry a resource instance identifier that uniquely points to the first resource instance "A-001". Because the monitoring event is cached on the basic resource side before entering the industrial cloud platform, when the event access component processes this monitoring event, the currently running resource instance corresponding to logical resource service A is already the second resource instance "A-002". The event access component queries the mapping relationship between the logical resource identifier "SRV-A" and the currently running resource instance identifier, and sets the current associated resource instance identifier of this monitoring event to "A-002". At the current moment, the second temporal resource dependency relationship between the second resource instance "A-002" and the second database instance "D-002" is valid. Therefore, the current resource dependency topology connects the second resource instance "A-002" and the second database instance "D-002". During the anomaly chain identification process, based on the current associated resource instance identifier "A-002" of the monitoring event, the instance start time of the second resource instance, 10:12, is read from the generational genealogy of the resource instances. The event occurrence time of the monitoring event, 10:08, is earlier than the instance start time of the second resource instance, 10:12. Therefore, it is determined that there is a time conflict before the instance's birth in this monitoring event. Further analysis of the dependency relationship between the second resource instance "A-002" and the second database instance "D-002" in the current resource dependency topology reveals that the dependency relationship was established at 10:13 AM. The monitoring event occurred at 10:08 AM, which is earlier than the dependency relationship establishment time of 10:13 AM, thus indicating a pre-dependency propagation conflict. The second resource instance "A-002" has a predecessor resource instance "A-001", and the first resource instance "A-001" and the second resource instance "A-002" have the same logical resource identifier "SRV-A". Therefore, the monitoring event, the second resource instance "A-002", and the dependency relationship between the second resource instance "A-002" and the second database instance "D-002" are identified as an instance generational backflow type pre-birth anomaly chain.
[0058] During the backtracking process of the actual resource instance, the first resource instance "A-001" was found by tracing the predecessor resource instance identifier of the second resource instance "A-002". The first resource instance "A-001" and the second resource instance "A-002" have the same logical resource identifier "SRV-A". The monitoring event occurred at 10:08, which falls within the instance validity period of the first resource instance from 9:00 to 10:15. The host node identifier "NODE-01" and the runtime space identifier "SPACE-01" of the monitoring event are the same as the host node identifier and runtime space identifier of the first resource instance "A-001", respectively. Therefore, the first resource instance "A-001" is determined to be the actual resource instance to which the monitoring event belongs, and a corrected resource instance ownership relationship is formed between the monitoring event and the first resource instance "A-001".
[0059] Using the event occurrence time of 10:08 as the filtering time, temporal resource dependencies are filtered from the temporal resource dependency set. The first temporal resource dependency from the first resource instance "A-001" to the first database instance "D-001" is valid at 10:08, and the first database instance "D-001" is also valid at 10:08, therefore the first temporal resource dependency is retained. The second temporal resource dependency from the second resource instance "A-002" to the second database instance "D-002" has not yet been established at 10:08, therefore the second temporal resource dependency is not included in the event occurrence dependency chain. Finally, the event occurrence dependency chain from the first resource instance "A-001" to the first database instance "D-001" is formed. When generating the current running panorama, the second resource instance "A-002" and the second database instance "D-002" that are still valid are displayed, along with the currently valid second temporal resource dependencies between them. When generating the panorama when an anomaly occurs, the first resource instance "A-001" and the first database instance "D-001" are displayed. The monitoring event that occurred at 10:08 is set on the resource instance node corresponding to the first resource instance "A-001", and the event dependency link from the first resource instance "A-001" to the first database instance "D-001" is displayed. When an anomaly occurs, the panoramic view will no longer display the monitoring event in the second resource instance "A-002", nor will it display the connection formed by the dependency relationship between the monitoring event and the second database instance "D-002"; thereby eliminating the cross-generational erroneous connection between the monitoring event of the previous generation resource instance and the current resource dependency topology of the next generation resource instance.
[0060] Furthermore, the verification process for instance location information during overlapping operations is as follows: This embodiment illustrates how to use instance location information to verify the attribution relationship between monitoring events and previous generation resource instances when multiple resource instances under the same logical resource are running in parallel in time. Logical resource service B's logical resource identifier is "SRV-B"; during the expansion period, logical resource service B simultaneously runs the third resource instance and the fourth resource instance; the resource instance identifier of the third resource instance is "B-001", the host node identifier is "NODE-03", the running space identifier is "SPACE-B1", and the instance's validity period is from 13:00 to 14:00; the resource instance identifier of the fourth resource instance is "B-002", the host node identifier is "NODE-04", the running space identifier is "SPACE-B2", and the instance's validity period is from 13:10 to 14:10. Both the third and fourth resource instances belong to logical resource service B, but they are parallel resource instances formed by horizontal scaling. They do not establish a predecessor relationship due to their different instance startup times. During subsequent version switching, the fifth resource instance "B-003" takes over the service shards provided by the third resource instance "B-001" at 2:15 PM. The instance switching record records the third resource instance "B-001" as the predecessor resource instance of the fifth resource instance "B-003". The fourth resource instance "B-002" continues to run as a parallel resource instance and is not a predecessor resource instance of the fifth resource instance "B-003". A monitoring event occurred at 13:30, with a logical resource identifier of "SRV-B", a host node identifier of "NODE-03", and a runtime space identifier of "SPACE-B1". This monitoring event did not carry a resource instance identifier that could uniquely point to the third resource instance "B-001". After the fifth resource instance "B-003" started, it was initially associated with the fifth resource instance "B-003" based on the mapping relationship between the logical resource identifier and the currently running resource instance identifier. The event time of 13:30 was earlier than the instance start time of the fifth resource instance "B-003" at 14:15, thus forming a time conflict before the instance was created. The third resource instance "B-001" is read from the predecessor resource instance identifier of the fifth resource instance "B-003"; the event time of 13:30 falls within the instance validity period of the third resource instance "B-001"; the host node identifier "NODE-03" and the runtime space identifier "SPACE-B1" of the monitoring event are the same as the host node identifier and runtime space identifier of the third resource instance "B-001" respectively; therefore, the third resource instance "B-001" is determined to be the actual resource instance to which the monitoring event belongs. Although the fourth resource instance "B-002" has the same logical resource identifier as the third resource instance "B-001" and its instance validity period also covers 13:30, the fourth resource instance "B-002" is not the predecessor resource instance of the fifth resource instance "B-003", and its instance location information is different from the instance location information of the monitoring event. Therefore, the monitoring event is not attributed to the fourth resource instance "B-002". This embodiment shows that the instance startup time is used to determine the instance validity period and sorting order of resource instances, the previous generation relationship between resource instances is determined by the instance switching record, and the instance location information is used to verify whether the monitoring event was generated by the previous generation resource instance being queried, thereby avoiding incorrectly connecting parallel resource instances as instance generation relationships.
[0061] Furthermore, the process for handling incomplete time fields or instance location information is as follows: This embodiment illustrates the handling method when key fields of monitoring events or resource records are incomplete; A monitoring event has a logical resource identifier and an event occurrence time, but no runtime space identifier. The monitoring event is initially associated with a descendant resource instance, and the event occurrence time is earlier than the instance startup time of the descendant resource instance. Because the monitoring event lacks a runtime space identifier, it cannot be confirmed that the instance location information of the monitoring event is the same as the instance location information of the previous generation resource instance. Therefore, the previous generation resource instance is not identified as the actual resource instance to which the monitoring event belongs, and no deterministic relationship between the monitoring event and the resource instance is formed. Another monitoring event has a logical resource identifier, a currently associated resource instance identifier, a host node identifier, and a runtime space identifier, but the time base used for its event occurrence time cannot be converted to the unified time base of the industrial cloud platform. Since the event occurrence time of this monitoring event cannot be uniformly compared with the instance startup time and the dependency relationship establishment time, this monitoring event is not used to determine time conflicts before the instance is born or propagation conflicts before the dependency relationship is established. Another monitoring event has complete event occurrence time and instance location information, but its current associated resource instance identifier does not have a corresponding resource instance in the resource instance generational genealogy; since the instance startup time and the previous generation resource instance identifier of the current associated resource instance cannot be read, this monitoring event cannot be used to form a deterministic instance generational backflow type pre-birth anomaly chain; Through the above processing, resource instance ownership correction and dependency link recovery at the time of the event are only performed when the event occurrence time, instance startup time, dependency relationship establishment time, predecessor resource instance identifier, and instance location information meet the deterministic judgment conditions, thus avoiding error correction due to missing key fields.
[0062] Furthermore, the handling process for instance termination time and dependency resolution time being missing is as follows: This embodiment illustrates the handling method when an end time is missing in the instance lifecycle record or dependency change record; The predecessor resource instance of Logical Resource Service C is the Sixth Resource Instance, and the successor resource instance is the Seventh Resource Instance. The instance start time of the Sixth Resource Instance is 8:00 AM, and there is no explicit instance termination time in its instance lifecycle record. The instance switchover record shows that the Seventh Resource Instance completed the takeover of the Sixth Resource Instance at 9:30 AM, and after the instance switchover, the Sixth Resource Instance no longer hosts Logical Resource Service C. Therefore, 9:30 AM is determined as the instance termination time of the Sixth Resource Instance, and the instance validity period of the Sixth Resource Instance is from 8:00 AM to 9:30 AM. If the instance switching record only indicates that the seventh resource instance has started, but does not indicate that the sixth resource instance has stopped hosting services, then the instance startup time of the seventh resource instance will not be used as the instance termination time of the sixth resource instance. Instead, the instance validity period of the sixth resource instance will continue to be retained until an instance stop record, instance destruction record, or clear instance switching completion record is obtained. For a temporal resource dependency, if the dependency change record does not record the dependency termination time, and both the dependency initiating instance and the dependency target instance are within their instance validity period at the current time, the validity period of the temporal resource dependency is extended to the current time; if the dependency initiating instance terminated at 9:30, but the dependency change record still does not record the dependency termination time, 9:30 is determined as the end time of the validity period of the temporal resource dependency; if the dependency target instance terminated at 9:20, and the dependency initiating instance terminated at 9:30, the earlier 9:20 is determined as the end time of the validity period of the temporal resource dependency; if the dependency termination record records a dependency termination time of 9:40, but the dependency target instance terminated at 9:20, the earlier 9:20 is taken as the end time of the validity period of the temporal resource dependency. The above processing ensures that both the instance validity period of resource instances and the validity period of temporal resource dependencies have definite boundaries that can be used for time comparison.
[0063] Furthermore, the process of forming dependency links when multi-level events occur is as follows: This embodiment is used to illustrate the formation process when a dependency link contains multiple levels of resource instances when an event occurs; The actual resource instance to which the monitored event belongs is resource instance A, and the event occurred at 11:00 AM. The temporal resource dependency records show the following relationships: the temporal resource dependency between resource instance A and resource instance B is valid from 10:00 AM to 12:00 PM; the temporal resource dependency between resource instance B and resource instance C is valid from 10:30 AM to 11:30 AM; the temporal resource dependency between resource instance C and resource instance A is valid from 10:40 AM to 11:20 AM; and the temporal resource dependency between resource instance B and resource instance D is valid from 11:10 AM onwards. Using resource instance A as the starting resource instance and 11:00 as the filtering time; the temporal resource dependency relationship between resource instance A and resource instance B is valid at 11:00, and resource instance B is within the instance validity period at 11:00, therefore resource instance B is added to the dependency chain when the event occurs; The process continues with resource instance B as the next level resource instance. The temporal resource dependency between resource instance B and resource instance C is valid at 11:00 AM, and resource instance C is within the valid period at 11:00 AM. Therefore, resource instance C is added to the dependency chain at the time of the event. The temporal resource dependency between resource instance B and resource instance D is established at 11:10 AM, and has not been established at the time of the event, 11:00 AM. Therefore, resource instance D is not added to the dependency chain at the time of the event. The process continues by using resource instance C as the next-level resource instance for further filtering. The temporal resource dependency between resource instance C and resource instance A is valid at 11:00 AM, but resource instance A has already been added to the event-time dependency chain, so resource instance A is not traversed again, nor are resource instance nodes formed again. When there are no temporal resource dependencies that have not yet been selected and are valid at 11:00 AM, the formation of the event-time dependency chain stops. The final event-time dependency chain includes resource instance A, resource instance B, and resource instance C, as well as the temporal resource dependencies between resource instance A and resource instance B, and between resource instance B and resource instance C. The dependency link at the time of the event indicates the objectively existing resource dependency relationship at the time the monitoring event occurs. It is not based on this to determine whether the monitoring event has been propagated from resource instance A to resource instance B and resource instance C in sequence.
[0064] Example 3 One embodiment of this disclosure provides a panoramic monitoring system for the operation of industrial infrastructure resource services, including: The monitoring record acquisition module is used to acquire monitoring events, instance lifecycle records, and dependency change records of the industrial cloud platform. The instance generation construction module is used to establish the resource instance generational lineage and instance validity period of the same logical resource based on the instance lifecycle record. The temporal dependency building module is used to establish a temporal resource dependency set and the current resource dependency topology based on dependency change records. The anomaly chain identification module is used to identify pre-birth anomaly chains with generational backflow based on the generational lineage of resource instances, the validity period of instances, and the current resource dependency topology. The synchronous backtracking module is used to find the previous generation resource instance whose event occurrence time falls within the instance's valid period and determine the actual resource instance it belongs to by searching the generational backflow anomaly chain of instances. It also filters the dependencies whose event occurrence time falls within the valid period and are connected to the actual resource instance it belongs to from the temporal resource dependency set, forming the dependency chain at the time of the event occurrence. The panoramic generation module is used to generate panoramic monitoring results based on the actual resource instance, the dependent links at the time of the event, and the current resource dependency topology.
[0065] Specifically, such as Figure 2 As shown, this embodiment provides a comprehensive monitoring system for the operation of basic resource services, including a monitoring record acquisition module, an instance generation construction module, a temporal dependency construction module, an anomaly chain identification module, a synchronization backtracking module, and a comprehensive generation module.
[0066] The monitoring record acquisition module is connected to the monitoring acquisition component, resource management component, service management component, and event access component of the industrial cloud platform, respectively. The monitoring record acquisition module acquires monitoring events, instance lifecycle records, and dependency change records, and forms a basic resource operation record set; The monitoring record acquisition module only identifies the running records that have been assigned an anomaly flag by the monitoring acquisition component as monitoring events; The monitoring record acquisition module acquires the logical resource identifier, the currently associated resource instance identifier, the instance location information, and the event occurrence time of the monitoring event; The current associated resource instance identifier is formed by the event access component based on the mapping relationship between the logical resource identifier and the currently running resource instance identifier; when the monitoring event already carries a resource instance identifier that can uniquely point to a specific resource instance, the resource instance identifier carried by the monitoring event is retained; The monitoring record acquisition module also acquires the logical resource identifier, resource instance identifier, instance location information, instance start time and instance termination time from the instance lifecycle record, and the dependency initiating instance identifier, dependency target instance identifier, dependency establishment time and dependency termination time from the dependency change record; The monitoring record acquisition module converts the event occurrence time, instance start time, instance termination time, dependency establishment time, and dependency removal time to the unified time base of the industrial cloud platform, and excludes records that cannot be converted to the unified time base or lack valid time fields from the time conflict judgment. The instance generation building module reads instance lifecycle records from the basic resource operation record set and groups resource instances according to logical resource identifiers; The instance generation building module determines the previous generation resource instance identifier of each resource instance based on the instance switching record and establishes the resource instance generation genealogy; The instance generation building module determines the instance validity period of each resource instance based on the instance start time, instance termination time, or instance switchover completion time; For resource instances that run in parallel and do not have instance succession relationships, the instance generation building module retains the corresponding instance validity period and does not establish a previous generation relationship between parallel resource instances. The temporal dependency build module reads the dependency relationship establishment and dependency relationship removal records explicitly generated by the industrial cloud platform; The temporal dependency building module determines the dependency initiating instance identifier, dependency target instance identifier, relationship effective time, and relationship expiration time as the dependency initiating instance identifier, dependency target instance identifier, dependency relationship establishment time, and dependency relationship termination time, respectively, and combines them to form a temporal resource dependency relationship set; Temporal dependency builders do not infer temporal resource dependencies based on similar names, single runtime log entries, or simple time proximity. When the dependency resolution time is missing, the temporal dependency building module determines the end time of the validity period of the temporal resource dependency based on the current time and the instance termination time of the connected resource instance. The temporal dependency building module selects temporal resource dependencies from the temporal resource dependency set that are currently valid and the connected resource instances are currently valid, forming the current resource dependency topology; The anomaly chain identification module reads monitoring events, resource instance generational genealogy, temporal resource dependency set, and current resource dependency topology; The anomaly chain identification module compares the occurrence time of the monitored event with the instance startup time of the currently associated resource instance, and determines the pre-birth time conflict when the event occurrence time is earlier than the instance startup time; The anomaly chain identification module also compares the event occurrence time with the dependency establishment time of the connected dependencies in the current resource dependency topology, and determines the pre-dependency propagation conflict when the event occurrence time is earlier than the dependency establishment time. When the current associated resource instance has a predecessor resource instance with the same logical resource identifier, and the same monitoring event corresponds to both the pre-birth time conflict and the pre-propagation conflict of dependency establishment, the anomaly chain identification module determines the monitoring event, the current associated resource instance, and the dependency relationship with the pre-propagation conflict of dependency establishment as an instance generation backflow type pre-birth anomaly chain. The synchronous backtracking module reads the instance's generational backflow-type pre-birth anomaly chain and queries the previous generation resource instance along the previous generation resource instance identifier of the current associated resource instance; The synchronous backtracking module determines the actual resource instance to which the monitored event belongs based on the logical resource identifier, the event occurrence time, the instance validity period, and the instance location information. Only when both the host node identifier and the runtime space identifier of the monitored event exist and are identical to the corresponding fields of the unique predecessor resource instance, can the synchronization backtracking module establish a deterministic relationship between the monitored event and the resource instance. When any instance location field is missing, or the actual resource instance cannot be uniquely identified, the synchronization backtracking module does not perform deterministic resource instance attribution correction. The synchronous backtracking module uses the event occurrence time as the filtering time to select temporal resource dependencies that are connected to the actual resource instance, are valid during the event occurrence time, and the target resource instance is valid during the event occurrence time, thus forming a dependency chain at the time of the event occurrence. When a multi-level event-time dependency chain is formed, the synchronous backtracking module filters the temporal resource dependencies layer by layer starting from the actual resource instance to which it belongs. It does not repeatedly traverse resource instances that have already been added to the event-time dependency chain, and stops when there are no new valid temporal resource dependencies. The panoramic generation module reads the generational lineage of resource instances, the current resource dependency topology, the corrected relationship between monitoring events and resource instance affiliation, and the dependency links when the events occur. The panorama generation module generates the currently running panorama based on the resource instances that are currently in the instance validity period and the current resource dependency topology; The panorama generation module generates a panorama of the time when the anomaly occurs based on the actual resource instance to which the monitored event belongs and the dependency links at the time the event occurs. The panoramic generation module sets resource instances of different generations under the same logical resource as independent resource instance nodes, and sets the monitoring events that have been corrected for resource instance affiliation to the resource instance node corresponding to the actual resource instance. The panorama generation module does not display resource dependencies that were not established at the time of the event in the panorama when an anomaly occurs; The output of the monitoring record acquisition module is passed to the instance generation construction module, the temporal dependency construction module, and the anomaly chain identification module, respectively; the output of the instance generation construction module and the temporal dependency module is passed to the anomaly chain identification module; the output of the anomaly chain identification module is passed to the synchronization backtracking module; and the output of the synchronization backtracking module is passed to the panorama generation module, thus forming a complete data processing chain. The basic resource service operation panoramic monitoring system described in this embodiment can be deployed on the management server, industrial computer, or industrial cloud platform control node of the industrial cloud platform; Each module can be implemented by the processor executing computer programs stored in memory, or by multiple computing nodes working together; the physical deployment location of each module does not affect the corresponding processing performed by each module according to the above data transmission relationship and processing logic.
[0067] Through the above implementation methods, this disclosure can detect time conflicts between the current resource instance affiliation and the lifecycle of the resource instance when resource instance reconstruction, rolling release, elastic scaling, or cross-node migration occurs on the industrial cloud platform, based on the instance validity period of the resource instance and the validity period of the temporal resource dependency relationship. It can also synchronously restore the previous generation resource instance to which the monitoring event actually belongs and the resource dependency relationship that was valid when the event occurred, thereby avoiding the formation of an intergenerational backflow type pre-birth anomaly chain of monitoring events generated by the previous generation resource instance along the current resource dependency topology of the subsequent generation resource instance.
[0068] Example 4 One embodiment of this disclosure provides a non-transitory computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the aforementioned method for panoramic monitoring of industrial infrastructure resource service operation.
[0069] Example 5 One embodiment of this disclosure provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the aforementioned method for panoramic monitoring of industrial infrastructure resource service operation.
[0070] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0071] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0072] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.
Claims
1. A method for panoramic monitoring of the operation of industrial infrastructure resource services, characterized in that, include: Acquire monitoring events, instance lifecycle records, and dependency change records from the industrial cloud platform; Establish the generational genealogy of resource instances for the same logical resource and the effective period of the instance based on the instance lifecycle record; establish the temporal resource dependency set and the current resource dependency topology based on the dependency change record. Based on the generational genealogy of resource instances, the validity period of instances, and the current resource dependency topology, identify the pre-birth abnormal chain of instance generational backflow; For instance generational backflow type pre-birth anomaly chain, search along the generational genealogy of resource instances to find the previous generation resource instance whose event occurred during the instance's valid period, and determine the actual resource instance to which it belongs; The event occurrence time is selected from the temporal resource dependency set and the dependency relationship is connected to the actual resource instance, forming the dependency link when the event occurs; The system generates a panoramic monitoring result based on the actual resource instance, the dependency link at the time of the event, and the current resource dependency topology.
2. The method for panoramic monitoring of industrial infrastructure resource service operation as described in claim 1, characterized in that, The acquisition of monitoring events, instance lifecycle records, and dependency change records from the industrial cloud platform includes: Record the logical resource identifier, currently associated resource instance identifier, instance location information, and event occurrence time of the monitored event; Record the logical resource identifier, resource instance identifier, instance location information, instance start time, and instance termination time of the instance lifecycle. Record the dependency initiating instance identifier, dependency target instance identifier, dependency establishment time, and dependency termination time of dependency relationship changes; The instance location information includes the host node identifier and the running space identifier, and the instance lifecycle record includes the instance startup record, instance stop record, instance destruction record, and instance switchover record.
3. The method for panoramic monitoring of industrial infrastructure resource service operation as described in claim 1, characterized in that, The establishment of the resource instance generational lineage and instance validity period for the same logical resource based on the instance lifecycle record includes: Group resource instances according to their logical resource identifiers; Based on the instance switching record, the resource instance identifier of the resource instance to be replaced is recorded as the previous generation resource instance identifier of the resource instance that replaces the resource instance, and the resource instances in the same group are arranged according to the instance startup time. The effective period of an instance is defined as the time from instance startup to instance termination. For resource instances that do not record the instance termination time but record the instance switchover completion time, the instance switchover completion time shall be determined as the instance termination time. For resource instances that are still running, extend the instance's validity period to the current moment when the method is executed; For resource instances running in parallel, the corresponding instance validity period is retained.
4. The method for panoramic monitoring of industrial infrastructure resource service operation as described in claim 1, characterized in that, Based on dependency change records, establish a temporal resource dependency set and the current resource dependency topology, including: Combine the dependency initiating instance identifier, the dependency target instance identifier, the dependency establishment time, and the dependency termination time into a temporal resource dependency relationship; The effective period of the temporal resource dependency is defined as the time from when the dependency is established to when the dependency is terminated. For temporal resource dependencies for which the release time is not recorded, extend their validity period to the current moment; Select the temporal resource dependencies that are within the valid period at the current time to form the current resource dependency topology.
5. The method for panoramic monitoring of industrial infrastructure resource service operation as described in claim 1, characterized in that, Identify instances of intergenerational backflow-type prenatal abnormal chains, including: Based on the current associated resource instance identifier, read the instance start time of the current associated resource instance from the generational genealogy of the resource instances; If the event occurs before the instance startup time, it is determined that the monitoring event has a time conflict before the instance was created. Read the dependencies connected to the current associated resource instance from the current resource dependency topology. If the event occurs earlier than the dependency establishment time of at least one of the aforementioned dependencies, determine that there is a pre-establishment propagation conflict in the corresponding dependency. When a previous generation of a resource instance with the same logical resource identifier exists in the current associated resource instance, and the monitoring event simultaneously exhibits both pre-instance birth time conflict and pre-establishment propagation conflict, the monitoring event, the current associated resource instance, and the dependency relationship with the pre-establishment propagation conflict are identified as an instance generation backflow type pre-instance birth anomaly chain.
6. The method for panoramic monitoring of industrial infrastructure resource service operation as described in claim 1, characterized in that, For instance-level backflow-type pre-birth anomaly chains, search along the resource instance generational genealogy for previous generation resource instances whose event occurrence time falls within the instance's valid period to determine the actual resource instance to which it belongs, including: Read the previous generation resource instances sequentially along the identifiers of the previous generation resource instances of the current associated resource instance; Select a previous generation resource instance that has the same logical resource identifier as the currently associated resource instance, whose event occurrence time falls within the corresponding instance's valid period, and whose instance location information is the same as the instance location information of the monitoring event; When only one predecessor resource instance meets the selection criteria, the predecessor resource instance is determined as the actual resource instance to which the monitoring event belongs. The monitoring event is then reassigned from the current associated resource instance to the actual resource instance, and the connection of the monitoring event in the current resource dependency topology along the dependency relationship that has a pre-propagation conflict is canceled.
7. The method for panoramic monitoring of industrial infrastructure resource service operation as described in claim 1, characterized in that, The process of filtering dependencies from the temporal resource dependency set that occur within the valid period and are connected to the actual resource instance to which they belong, forming a dependency chain at the time of the event, includes: Using the event occurrence time of the monitored event as the filtering time, select temporal resource dependencies from the temporal resource dependency set that are connected to the actual resource instance, whose event occurrence time falls within the corresponding valid period, and whose dependent target instance is within the instance's valid period during the event occurrence time; The selected temporal resource dependencies are connected to form a dependency chain when the event occurs.
8. A panoramic monitoring system for the operation of industrial infrastructure resource services, characterized in that, include: The monitoring record acquisition module is used to acquire monitoring events, instance lifecycle records, and dependency change records of the industrial cloud platform. The instance generation construction module is used to establish the resource instance generational lineage and instance validity period of the same logical resource based on the instance lifecycle record. The temporal dependency building module is used to establish a temporal resource dependency set and the current resource dependency topology based on dependency change records. The anomaly chain identification module is used to identify pre-birth anomaly chains with generational backflow based on the generational lineage of resource instances, the validity period of instances, and the current resource dependency topology. The synchronous backtracking module is used to find the previous generation resource instance whose event occurrence time falls within the instance's valid period and determine the actual resource instance it belongs to by searching the generational backflow anomaly chain of instances. It also filters the dependencies whose event occurrence time falls within the valid period and are connected to the actual resource instance it belongs to from the temporal resource dependency set, forming the dependency chain at the time of the event occurrence. The panoramic generation module is used to generate panoramic monitoring results based on the actual resource instance, the dependency links at the time of the event, and the current resource dependency topology.
9. The panoramic monitoring system for the operation of industrial infrastructure resource services as described in claim 8, characterized in that, The acquisition of monitoring events, instance lifecycle records, and dependency change records from the industrial cloud platform includes: Record the logical resource identifier, currently associated resource instance identifier, instance location information, and event occurrence time of the monitored event; Record the logical resource identifier, resource instance identifier, instance location information, instance start time, and instance termination time of the instance lifecycle. Record the dependency initiating instance identifier, dependency target instance identifier, dependency establishment time, and dependency termination time of dependency relationship changes; The instance location information includes the host node identifier and the running space identifier, and the instance lifecycle record includes the instance startup record, instance stop record, instance destruction record, and instance switchover record.
10. The panoramic monitoring system for the operation of industrial infrastructure resource services as described in claim 8, characterized in that, The establishment of the resource instance generational lineage and instance validity period for the same logical resource based on the instance lifecycle record includes: Group resource instances according to their logical resource identifiers; Based on the instance switching record, the resource instance identifier of the resource instance to be replaced is recorded as the previous generation resource instance identifier of the resource instance that replaces the resource instance, and the resource instances in the same group are arranged according to the instance startup time. The effective period of an instance is defined as the time from instance startup to instance termination. For resource instances that do not record the instance termination time but record the instance switchover completion time, the instance switchover completion time shall be determined as the instance termination time. For resource instances that are still running, extend the instance's validity period to the current moment when the method is executed; For resource instances running in parallel, the corresponding instance validity period is retained.
