CN112860508B

CN112860508B - Abnormal positioning method, device and equipment based on knowledge graph

Info

Publication number: CN112860508B
Application number: CN202110045016.4A
Authority: CN
Inventors: 陈景珂; 张建伟; 盛闯; 朱冠胤; 范炳赟; 舒昌衡
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2021-01-13
Filing date: 2021-01-13
Publication date: 2023-02-28
Anticipated expiration: 2041-01-13
Also published as: CN112860508A

Abstract

The embodiment of the specification discloses a method, a device and equipment for abnormal positioning based on a knowledge graph, and the scheme comprises the following steps: determining a knowledge graph containing monitored nodes and relations among the nodes; acquiring monitoring data corresponding to the monitored nodes, and determining a plurality of monitored nodes as transaction nodes according to transaction conditions reflected by the monitoring data; determining the association between the abnormal conditions corresponding to the abnormal nodes according to the topology in the knowledge graph; screening at least part of abnormal conditions from the abnormal conditions according to the correlation; and determining abnormal nodes in the abnormal nodes according to the abnormal conditions left after screening.

Description

Abnormal positioning method, device and equipment based on knowledge graph

Technical Field

The present disclosure relates to the field of computer software technologies, and in particular, to a method, an apparatus, and a device for locating an anomaly based on a knowledge graph.

Background

With the rapid development of computer technology and the internet, more and more businesses are carried out on the internet, which brings great convenience to the life of people, but also brings more and more risks. In order to protect the risk, the service is usually monitored by a wind control system.

In the existing wind control system, each monitor is relatively independent, and an alarm can be given when a certain monitor monitors abnormal conditions.

Based on this, there is also a need for a solution that can more efficiently locate anomalies in a wind-controlled scene.

Disclosure of Invention

One or more embodiments of the present disclosure provide a method, an apparatus, a device, and a storage medium for knowledge-graph-based anomaly location, so as to solve the following technical problems: the method can be used for more effectively positioning the abnormity in the wind control scene.

To solve the above technical problems, one or more embodiments of the present specification are implemented as follows:

one or more embodiments of the present specification provide a method for anomaly location based on a knowledge-graph, including:

determining a knowledge graph containing monitored nodes and relations among the nodes;

acquiring monitoring data corresponding to the monitored nodes, and determining a plurality of monitored nodes as transaction nodes according to transaction conditions reflected by the monitoring data;

determining the association between the abnormal conditions corresponding to the abnormal nodes according to the topology in the knowledge graph;

screening at least part of the abnormal situations from the abnormal situations according to the association;

and determining abnormal nodes in the abnormal nodes according to the abnormal conditions left after screening.

One or more embodiments of the present specification provide a knowledge-graph-based anomaly locating apparatus, including:

the knowledge graph determining module is used for determining a knowledge graph containing monitored nodes and relations among the nodes;

the abnormal node determining module is used for acquiring monitoring data corresponding to the monitored nodes and determining the monitored nodes as abnormal nodes according to abnormal conditions reflected by the monitoring data;

the association determining module is used for determining the association between the abnormal conditions corresponding to the abnormal nodes according to the topology in the knowledge graph;

a screening module for screening at least part of the abnormal situations from the abnormal situations according to the association;

and the abnormal node determining module is used for determining abnormal nodes in the abnormal nodes according to the abnormal conditions left after the screening.

at least one processor; and (c) a second step of,

a memory communicatively coupled to the at least one processor; wherein, the first and the second end of the pipe are connected with each other,

the memory stores instructions executable by the at least one processor to enable the at least one processor to:

screening at least part of the abnormal conditions from the abnormal conditions according to the association;

One or more embodiments of the present specification provide a non-transitory computer storage medium storing computer-executable instructions configured to:

At least one technical scheme adopted by one or more embodiments of the specification can achieve the following beneficial effects: through the knowledge map, normal abnormal movement can be screened out when abnormal positioning is carried out, so that the range of abnormal positioning is reduced, the accuracy of abnormal positioning is improved, the disturbance rate of alarming can be reduced, and the working efficiency is improved.

Drawings

In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative effort.

FIG. 1 is a flow diagram of a method for knowledge-graph based anomaly location according to one or more embodiments of the present disclosure;

FIG. 2 is a schematic illustration of relationships in a knowledge-graph provided in one or more embodiments of the present disclosure;

FIG. 3 is a schematic diagram of a topology formed by some abnormal nodes and non-abnormal nodes in a knowledge graph according to one or more embodiments of the present disclosure;

FIG. 4 is a schematic diagram of resolving an abnormal situation in a knowledge-graph provided in one or more embodiments of the present disclosure;

FIG. 5 is a schematic diagram of a knowledge-graph based anomaly locating device according to one or more embodiments of the present disclosure;

fig. 6 is a schematic structural diagram of an abnormality location apparatus based on a knowledge-graph according to one or more embodiments of the present disclosure.

Detailed Description

The embodiment of the specification provides a knowledge graph-based anomaly positioning method, a knowledge graph-based anomaly positioning device, knowledge graph-based anomaly positioning equipment and a storage medium.

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any inventive step based on the embodiments of the present disclosure, shall fall within the scope of protection of the present application.

In one or more embodiments of the present disclosure, a knowledge map is generated in advance, and at least a portion of the abnormal situations are screened out based on the knowledge map after the abnormal situations are monitored. Therefore, normal abnormal movement is filtered during abnormal positioning, the range of the abnormal positioning is narrowed, the disturbance rate of alarming is reduced, and the working efficiency is improved. The association between the abnormal conditions is obtained through the relationship between the nodes or the preset service scene, and the abnormal conditions are screened out through the association, which is beneficial to increasing the accuracy in screening out.

The following is a detailed description based on such a concept.

Fig. 1 is a schematic flowchart of a method for locating an anomaly based on a knowledge-graph according to one or more embodiments of the present disclosure. The method can be applied to different business fields, such as the field of internet financial business, the field of electric business, the field of instant messaging business, the field of game business, the field of official business and the like. The process can be executed by a computing device (e.g., a wind control server or an intelligent mobile terminal corresponding to the payment service) in the corresponding field, and some input parameters or intermediate results in the process allow manual intervention and adjustment to help improve accuracy. The process in fig. 1 may include the following steps:

s102, determining a knowledge graph containing the monitored nodes and the relations among the nodes.

The knowledge map is a series of graphs for displaying the relation between knowledge development process and structure, and is called knowledge domain visualization or knowledge domain mapping map in the book intelligence world. In the graph, knowledge resources and carriers thereof are described through a visualization technology, and knowledge and mutual relations among the knowledge resources and the carriers are mined, analyzed, constructed, drawn and displayed. Colloquially, a knowledge graph may be a relational network that is generated by joining various pieces of information together. In the knowledge graph, nodes are used for representing corresponding main bodies, and edges are used for connecting each node to represent the relation between the main bodies.

The monitored node refers to an object which is deployed in a corresponding service system and monitored by a wind control system, and in the process of generating the knowledge graph, the monitored object is used as a node in the knowledge graph. The type of the monitored node may be various, for example, an object related to software such as a policy, an interface, and a program used in the operation process of the business system, an object related to business such as an account and a commodity, an object having an entity such as a server and a terminal, and the like.

The monitored node typically has one or more attributes that reflect the state of the monitored node while it is operating. For example, when the monitored node is an interface, the corresponding attribute may include a call amount, a failure amount, a success amount, a response time, and the like, and when the monitored node is a commodity, the corresponding attribute may include a transaction amount, a success amount, a failure amount, an inventory, a price, and the like.

The inter-node relationship refers to the relationship between monitored nodes, and is embodied in the knowledge graph as an edge connecting the monitored nodes. The relationship between the nodes comprises a strong association rule and a weak association rule, when the support degree and the reliability of the association rule between the monitored nodes are higher than a preset threshold value, the relationship between the nodes can be regarded as the strong association rule, otherwise, the relationship between the nodes can be regarded as the weak association rule. Of course, the relationships between nodes may vary based on the service, device, environment, etc. For example, when the flow of a certain service is changed, the relationship between the nodes of the corresponding monitored node is also changed correspondingly; or, in different operating systems, the relationship between the nodes of the same monitored node is changed correspondingly.

And S104, acquiring monitoring data corresponding to the monitored nodes, and determining the monitored nodes as the abnormal nodes according to abnormal conditions reflected by the monitoring data.

The monitoring data come from the data obtained by the monitored nodes monitored by the wind control system. If the abnormal situation is found through the monitoring data in the monitoring process, the monitored node corresponding to the monitoring data is determined as the abnormal node. When only one abnormal node exists, abnormal positioning is directly carried out according to the abnormal node, so that corresponding treatment can be conveniently carried out through software or manual work. The following discussion is mainly directed to a case where a plurality of monitored nodes are determined as the transaction nodes.

The assumption of a transaction is determined by a separate monitoring data. If the monitoring data is data that can be quantified, such as the call volume, response time, transaction price, weight, ambient temperature, etc., a corresponding threshold range is preset. And if the monitoring data exceeds the threshold range, judging that the monitoring data corresponds to the abnormal situation at present. If the monitoring data is data which is difficult to quantify, such as account names, equipment models and the like, abnormal conditions are judged based on corresponding models, algorithms and the like. For example, the text content corresponding to the monitoring data is identified through a pre-trained text identification model to determine whether the text content contains sensitive keywords, whether the text content meets preset model requirements, and the like.

The assumed abnormal situation is obtained by judging a plurality of monitoring data, and is specifically obtained by comprehensively judging the association among the plurality of monitoring data. For example, if more than a certain amount of monitoring data in the plurality of monitoring data are in an abnormal state, that is, the monitoring data exceed the threshold range or do not meet the preset requirement, the situations corresponding to the plurality of monitoring data may all be determined as abnormal situations, or the situation corresponding to the monitoring data in the abnormal state may be determined as abnormal situations.

S106, determining the association between the abnormal conditions corresponding to the abnormal nodes according to the topology in the knowledge graph.

The topology refers to the mutual relationship among the spatial data satisfying the topological geometry principle, and when the topology is embodied in the knowledge graph, the topology includes the relationship among the monitored nodes and the attributes corresponding to the monitored nodes. The transaction situation is a situation on a monitored node (in this case, a transaction node) represented by the monitoring data, and therefore, the association between the transaction situations can be determined based on the relationship between the transaction nodes represented by the topology of the knowledge graph.

Specifically, the relationships between nodes include multiple types, such as mutual exclusion relationships, dependency relationships, and the like. When the relationship between the nodes corresponding to the abnormal nodes is a dependency relationship, a change in one of the plurality of abnormal nodes easily causes a change in the other, and therefore, before the association between the abnormal conditions corresponding to the abnormal nodes is determined, the relationship between the nodes may be determined to be a dependency relationship.

For convenience of description, the relationship between nodes is explained by taking a monitored node a and a monitored node B as an example, where a and B are used to distinguish different monitored nodes. If the relationship between the nodes is a dependency relationship, when the attribute of A changes, the attribute of B is likely to change accordingly. If the relationship between the nodes is an exclusive relationship, when the attribute of a changes (assuming that the current service selects a but not B, and the call volume to a is reduced), B will not change under the influence of a, because in this case, the call volume of B is temporarily and continuously 0 without change.

Further, the dependency includes a variety of forms, for example, as follows.

The relationship between a and B (hereinafter referred to as "between a and B") is between a whole and a part, and the part is difficult to be separated from the whole and exists alone. For example, if a is an enterprise node and B is a department node, the department corresponding to B has employees who leave, the attribute "employee number" of B will change, and the attribute "employee number" of a will also change.

The two are in a whole-to-part relationship, but the parts can easily be separated from the whole and exist separately. For example, if a is a server cluster node and B is a server node in the server cluster, when the attribute "log generation amount" of B changes during the process of processing data and generating a log by B, the attribute "log generation amount" of a also changes.

There is an association in the business process between the two, such as a business pre-relation, a business post-relation, an optional or necessary resource configuration relation, etc. For example, if a is called during the service execution, the corresponding attribute of the service front-end node where a is B may be changed, and if no other abnormal condition occurs, B may also be called during the service execution, and the attribute may be changed accordingly. For another example, a is a policy node, B is a corresponding interface node, and assuming that the interface must use the policy, a and B are a necessary resource configuration relationship, and in the process of invoking B to execute a service, a must be used, and at this time, if the attribute of a changes, the attribute of B is likely to change. Similarly, assuming that the interface optionally uses the policy, A and B configure the relationship for optional resources, a change in the A attribute may also cause a change in the B attribute.

In addition, to a lesser extent, the correlation between transaction scenarios is typically tighter. Based on this, when determining the association between the abnormal situations, it is first determined whether the abnormal situations belong to the same small range. If the abnormal conditions belong to the same small range, the association between the abnormal conditions is judged, so that the abnormal positioning efficiency is improved.

For example, the positions of the abnormal nodes corresponding to the abnormal conditions in the knowledge graph are determined, and then whether the abnormal conditions belong to the same small range is judged according to the distance between the abnormal nodes. If the distance between any two different moving nodes is smaller than the preset distance, or the distance between the two different moving nodes with the farthest distance is smaller than the preset distance, the different moving conditions can be considered to belong to the same small range. If there are multiple paths between two different nodes, the distance corresponding to the shortest path may be used as the distance between the two different nodes.

And S108, screening at least part of abnormal conditions from the abnormal conditions according to the association.

If at least part of the abnormal conditions exist in the abnormal conditions through the association between the abnormal conditions, and the root cause of the abnormal conditions is not caused by the abnormality of the corresponding monitored node, but is indirectly caused by the abnormality of other monitored nodes or caused by some normal services, the at least part of the abnormal conditions can be screened out.

Specifically, when the relationship between the nodes is a dependency relationship, the abnormal condition of the monitored node is more easily affected by the abnormal conditions of other monitored nodes. Based on this, in determining the association between the abnormal situations, if it is determined that there is a causal relationship between the abnormal situations according to the dependency relationship, at least part of the resulting abnormal situations are screened out in screening out the abnormal situations.

For example, fig. 2 is a schematic diagram of a relationship in a knowledge graph provided in one or more embodiments of the present disclosure. In fig. 2, the knowledge graph spectrum includes a plurality of monitored nodes, the first policy node is a selectable or mandatory policy of the first interface node, and there is a dependency relationship therebetween; the first interface node is a service preposition node of the second interface node, and service preposition relation is formed among the first interface node and the second interface node. In the service execution process, the corresponding failure amount of the first policy node rises by 50% due to objective reasons. Due to the rising of the failure amount of the first strategy node, the calling amount of the first interface node when the first strategy node is called drops. And because the first interface node is the service preposition node of the second interface node, the calling amount of the first interface node drops, so that the next step is difficult to be executed in the service execution process, and the calling amount of the second interface node drops by 20%.

At this time, the failure amount of the first policy node, the call amount of the first interface node, and the call amount of the second interface node all generate a transaction condition, and the first policy node, the first interface node, and the second interface node are all regarded as transaction nodes. However, since there is a causal relationship between the transaction statuses of the three, that is, the transaction statuses of the call quantity of the first interface node and the call quantity of the second interface node, there is a high possibility that the transaction statuses are caused by the transaction statuses of the failure quantity of the first policy node, and based on this, the transaction statuses corresponding to the call quantity of the first interface node and the call quantity of the second interface node, which are the results, can be eliminated.

And S110, determining abnormal nodes in the abnormal nodes according to the abnormal conditions left after screening.

Assuming that all abnormal conditions are screened out, the multiple abnormal conditions monitored and determined by the wind control system are noise, and the noise belongs to the abnormal conditions, but the noise is not caused by abnormal conditions in the service system. And determining abnormal nodes in the abnormal nodes to realize abnormal positioning under the assumption that residual abnormal conditions still exist after screening. And after the abnormal positioning is finished, alarming is carried out based on the position of the abnormal node so as to facilitate the processing of a manual or related program.

In order to improve efficiency, for example, the abnormal node is determined in the abnormal nodes corresponding to the remaining abnormal situations. Of course, if the screening action is already executed, the remaining abnormal node may be determined as the abnormal node. If the screening action is not completed, the screening action is repeated in the rest abnormal nodes until the abnormal node is determined.

When abnormal positioning is carried out, if only one abnormal node exists, the abnormal positioning is directly carried out according to the position of the abnormal node. If a plurality of abnormal nodes exist, the plurality of abnormal nodes can be connected to obtain corresponding abnormal links, and abnormal positioning is carried out according to the positions of the abnormal links. The abnormal link can more clearly reflect the conduction direction of the abnormal condition, and the subsequent processing of the abnormal condition is facilitated.

When the abnormal nodes are connected to form the abnormal link, the abnormal nodes can be connected in various ways. For example, the abnormal node is connected according to the shortest connection mode to form an abnormal link, or the abnormal node is connected according to the traffic flow to form an abnormal link. More intuitively, see fig. 3.

In fig. 3, there are four monitored nodes in the knowledge graph, which are all abnormal nodes, two of which are abnormal nodes and two of which are non-abnormal nodes.

If the shortest connection mode is adopted, the obtained abnormal link is as follows: first abnormal node-second non-abnormal node-second abnormal node. Although the abnormal link in the mode can only approximately reflect the conduction direction of the abnormal condition, people can position the abnormal condition more quickly, and the method is suitable for a service scene that the topology of the monitored node is simpler and the abnormality needs to be positioned quickly.

Assuming that the abnormal link is obtained according to the connection mode of the service flow direction: first abnormal node-first non-abnormal node-second abnormal node. The abnormal link in the mode can reflect the conduction direction of the abnormal condition in more detail, and is suitable for a service scene that the topological relation of the monitored node is complex and the abnormality needs to be positioned in detail.

In addition, some monitored nodes are abnormal, but the monitored nodes do not have abnormal conditions, but cause abnormal conditions of other monitored nodes. Based on this, when determining the abnormal node, the abnormal node may be determined not only in the abnormal node but also in the monitored node. By performing the abnormality localization in a wider range, the accuracy of the abnormality localization can be improved.

In one or more embodiments of the present specification, in the knowledge graph, besides the monitored node, some other objects may be used as nodes, for example, an object that monitors the monitored node is used as a monitoring node, and the like.

The monitoring node is connected with the corresponding monitored node, and the specified attribute is monitored specifically. As mentioned above, the monitored node has one or more attributes, the attributes are used as monitored indexes, the monitoring node collects monitoring data for the indexes, the indexes include, for example, a call volume, response time, transaction price, etc., and the indexes are usually in a numerical form, and then a judgment standard of the abnormal situation is predefined according to the numerical value, for example, the abnormal situation is judged when the predefined call volume falls to a certain threshold.

When the monitoring node collects the monitoring data, the monitoring data corresponding to part or all indexes of the monitored node is collected for judging the abnormal node. For convenience of description, the index collected by the monitoring node and used for judging the transaction node is referred to as a designated index, and the designated index corresponding to each monitored node may be preset. And when the collected first index of the first monitored node and the second index of the second monitored node have abnormal conditions, determining the first monitored node and the second monitored node as abnormal nodes.

It has been mentioned above that certain abnormal situations caused by normal traffic are screened out, which is further explained. The normal service refers to a service meeting a corresponding rule or plan of the service system, for example, a service executed in the service system according to a preset plan, or a measure adopted when the service system deals with some problems.

When a service system executes some normal services, a transaction situation may be caused, the reason is not in other transaction nodes, and if only through the association between the transaction situations, it is difficult to determine the root cause of the transaction situation. Based on this, whether the current service scenario (considered as a normal service in the service scenario) is determined according to the association between the transaction scenarios, the transaction scenarios themselves, and the service definition of the specified index. If the service scene is successfully identified, the current abnormal situation is probably caused by the service scene and is not abnormal actually, so that at least part of abnormal situations are screened out from the abnormal situation according to the service scene.

For different service scenarios, corresponding specific indexes are defined in advance, for example, the specific indexes and the degree to which the specific indexes are to be reached included in the service scenarios are defined. And when the specified indexes in the collected monitoring data accord with the service definition, determining that the service scene is successfully identified.

When the service system executes different services, the corresponding designated indexes may be different. For example, if the service scenario is a commodity transaction service, the designated index is a transaction success amount, a transaction failure amount, a transaction total amount, and the like of the commodity node. If the service scene is an account registration service, the specified indexes are account number, account authority and the like.

Since normal traffic generally conforms to the corresponding rules or plans, there is a high probability that all abnormal situations caused by the normal traffic can be regarded as noise, and all abnormal situations caused by the normal traffic are screened out.

In particular, the normal traffic may be a traffic surge scenario in response to a stimulus. And if the increment corresponding to the specified index exceeds a preset threshold within a certain time, the service scene is considered as a service surge scene. For example, in an e-commerce platform, a business scenario may be a large promotional campaign. The merchant or the e-commerce platform stimulates the user to shop by means of price reduction, coupon selling and the like. At the moment, indexes such as business volume, business failure volume and the like in the E-commerce platform are increased rapidly, and the wind control system monitors abnormal conditions of the specified indexes. However, in the service scenario, the service system executes the service according to the preset plan, and an abnormal situation may not occur, so that the abnormal situation caused by the service scenario can be completely eliminated. For example, there is a membership platform that normally has a strict member registration limit such as the number of registered members or registration time limit, so that a large number of potential users cannot register members. But on some specific days, the affiliate platform incentivizes users to enroll for affiliates by lowering or removing the enrolment limit. At the moment, indexes such as member registration success amount, member registration failure amount and the like are increased rapidly, the wind control system monitors that the specified indexes are abnormal, and abnormal conditions caused by the service scene can be screened out completely.

The normal traffic may also be related traffic coping with an attack scenario. In actual operation, a business system may be subject to external attacks. In order to deal with the external attack, the service system usually takes corresponding measures, such as bulk authority control on the attack account, changing the protection policy of the firewall or closing a specific channel, and the like. At this point, there may be a discrepancy in metrics relating to account permissions, firewalls, or particular channels. However, in the service scenario, the measures adopted by the service system belong to normal countermeasures, so that all abnormal situations caused by the service scenario can be eliminated.

Further, when the e-commerce platform is attacked by black-out, for example, an attacker performs malicious bill-swiping and malicious bad comment on a certain commodity in the e-commerce platform through a plurality of accounts. In order to cope with the blackout attack, the business system restricts the permissions of a plurality of accounts in a batch manner, for example, the plurality of accounts are subjected to processes such as restricted shopping and seal numbers in a batch manner. At the moment, the wind control system monitors that the corresponding monitoring data such as account authority and the like generate abnormal changes in batches. Based on the above, the current service scene is judged to be the batch authority control scene corresponding to the attack, and abnormal conditions caused by the service scene can be completely screened out.

Further, when the e-commerce platform executes services in different operating systems (such as an android system, an iOS system, and the like), interfaces, policies, and the like to be invoked are usually different. Compared with the iOS system, the android system has an open source characteristic, so that attackers generally adopt the android system to carry out black-producing attack more frequently. Therefore, if the abnormal situation occurs to the monitoring data corresponding to the android system and the abnormal situation does not occur to the monitoring data corresponding to the iOS system, it can be further determined that the current service scenario is the batch authority control scenario corresponding to the attack.

In one or more embodiments of the present description, the knowledge-graph is generated based on online business data. The online service data is also called online traffic, and refers to service data generated during an actual service execution process after the service system is online. The on-line means that the service system has the necessary condition for formal operation and completes the publishing work. The online business system can be a formal version, and can also be other versions such as a test version.

In the process of performing a service after the service system is online, actual service data is generated, and is referred to as online service data. Acquiring online service data, deducing according to the online service data, determining the service flow direction between monitored nodes, and finally generating a knowledge graph based on the service flow direction.

Specifically, in the process of executing a certain service, the online service data sequentially passes through a plurality of monitored nodes, which is obtained through logs in a service system. The monitored nodes are sequentially plotted as a path tree. After a plurality of path trees are obtained through drawing for a plurality of times, the plurality of path trees are aggregated together according to the service flow direction in each path tree and the repeated nodes in the path trees to form graph association. Based on preset rules or artificial expert experience precipitation, editing the attributes, relationships and the like of part or all monitored nodes in the graph association to enable the attributes, relationships and the like to be strong rule association, and therefore the generation of the knowledge graph is completed. The human expert experience means judgment or prediction made by an expert, a consultant, or the like based on the accumulation of long-term experience.

Compared with the traditional mode that the system architecture of the knowledge map is established through an established knowledge base, and then the system architecture is completed through the steps of knowledge extraction, knowledge fusion, knowledge reasoning and the like, the knowledge map is generated through the on-line flow data, the knowledge map can better accord with the actual business working process, the modification times in the knowledge map generation process are reduced, and the accuracy of the knowledge map is improved.

In one or more embodiments of the present specification, after an alarm is given to an abnormal situation, the abnormal situation is processed manually or through a corresponding program, and based on a processing result, the knowledge graph is modified and perfected correspondingly, so that the content contained in the knowledge graph is more accurate, and the next use is facilitated.

In one or more embodiments of the present description, after the exception node is determined, the exception location may be considered to be completed. Although the alarm can be sent to manual processing, the alarm is sent to the personnel for processing, and a long time is consumed in the process until the processing is finished, so that the normal operation of the service system is easily affected.

Based on this, after the abnormal node is determined. Determining an abnormal reason of the abnormal node when the abnormal node is abnormal; and replacing the abnormal node with a first other node in a service system corresponding to the abnormal node according to the abnormal reason and the topology of the abnormal node in the knowledge graph, or assisting the abnormal node by a second other node in the service system to solve the abnormal condition.

For example, fig. 4 is a schematic diagram of solving an abnormal situation in a knowledge graph according to one or more embodiments of the present disclosure. In fig. 4, the knowledge graph includes five monitored nodes, and the first policy node causes an increase in the corresponding failure amount due to objective reasons, and is finally determined as an abnormal node.

If the objective reason is that the first policy node itself is abnormal, for example, a code corresponding to the first policy node is defective. At this time, in the service system, the first policy node may be replaced by the second policy node, and the second policy node may replace the first policy node to continue to execute the relevant policy, thereby solving the abnormal situation. The relationship between the second policy node and the first policy node may be an alternative relationship, where the alternative relationship indicates that when one of the two is in an abnormal condition, the other policy node can be used to temporarily replace the abnormal condition and continue to execute the service. For example, the second policy node is a backup node of the first policy node, or the second policy node is a policy used by the first policy node in an old version of the business system.

If the objective reason is that the external resource corresponding to the first policy node is abnormal, for example, the processing device corresponding to the first policy node is abnormal, it is difficult to call the first policy node in time, and the first policy node is abnormal. At this time, other server nodes may be invoked to assist in invoking the first policy node to resolve the current abnormal situation. And a service implementation relation is formed between the server node and the first strategy node, and the service implementation relation indicates that the first strategy node can be called to execute corresponding services through the server node when necessary.

Furthermore, the abnormal condition is solved by replacing the abnormal node by the first other node, assisting the abnormal node by the second other node or manually adjusting the abnormal node. Updating the topology in the knowledge graph to form a local attention topology, wherein the local attention topology comprises the connection relation between the monitoring nodes with abnormal conditions; and determining a topology monitoring node, and monitoring the local attention topology through the topology monitoring node to determine whether the abnormal condition is solved.

After the abnormal condition is solved, the monitoring node with the abnormal condition generates a new local attention topology in the knowledge graph, and the topology monitoring node monitors the monitoring node in the topology in a future period of time, so that whether the abnormal condition is really solved or not can be determined. The topology monitoring node can be a monitoring node in a local attention topology or a new node. The monitoring nodes are directly connected, so that the association between abnormal conditions can be more intuitively reflected, and the judgment efficiency of abnormal positioning and the judgment efficiency of the solution result of the abnormal conditions can be improved. The connection relationship between the monitoring nodes is similar to the relationship between the monitored nodes, and may include a mutual exclusion relationship, a dependency relationship, and the like.

In one or more embodiments of the present specification, when there is a dependency relationship between two corresponding to the monitored node, the corresponding dependency degree is different based on different forms of the dependency relationship. The degree of dependence indicates the possibility that one of the two changes when the other changes, and the higher the degree of dependence, the higher the possibility that the other changes.

If the two are in service pre-relation, service post-relation, optional or necessary resource configuration relation, the service can be executed only by the mutual cooperation of the two, and the dependence degree between the two is higher in the form. Of course, the dependency of the mandatory resource configuration relationship is higher than that of the optional resource configuration relationship.

If the relationship between the whole and the part is between the two, when the monitored node as the part changes, the monitored node as the whole changes with a high probability. However, when the monitored node as a whole changes, the possibility that part of the monitored nodes change is not so high. Thus the degree of dependence is lower in this form. Of course, the part can be separated from the whole easily and exist in the form, and compared with the part which is difficult to separate from the whole and exists in the form, the monitored node as the part can be separated from the whole easily and exists in the form, so the dependence degree is lower.

The difference of the dependence degree can be embodied in the relation between the nodes contained in the knowledge graph. When the association between the abnormal situations is determined by the dependency relationship, the higher the degree of dependency corresponding to the association, the higher the possibility of having the association between the abnormal situations.

In one or more embodiments of the present disclosure, after determining the association between the abnormal situations corresponding to the abnormal situations, the abnormal nodes related to the association and causing the abnormal situations to conduct may be merged to obtain a merged node, edges between the abnormal nodes are included in the merged node, and edges between the abnormal nodes and other nodes are adjusted to connect with the merged node as a representative. Certainly, more reliably, the original knowledge graph can be retained, one copy of the knowledge graph is generated, node combination is carried out on the copy, a plurality of different copies can be obtained through updating along with service progress, and then when abnormal conditions or alarms occur on the original knowledge graph or one copy, rapid searching and matching can be carried out in each copy to help to screen out the abnormal conditions more efficiently and determine abnormal nodes.

In one or more embodiments of the present disclosure, some nodes (referred to as risk nodes) have an association with a current transaction node, which may cause mobile conduction, but may not be conducted due to a transaction condition that has occurred yet to reach a conduction threshold, or due to a time delay, etc., or even the transaction reason itself is a risk node itself, and it is not known whether a transaction has occurred or not currently; in this case, in order to ensure the risk nodes, the risk nodes may be brought into the transaction node set temporarily, and then the risk nodes may be positioned as abnormal nodes according to specific situations, so that the reliability of the abnormal positioning scheme is improved with partial efficiency, and the subsequent maintenance and the resolution of the service hidden troubles are facilitated more comprehensively.

Based on the same idea, one or more embodiments of the present specification further provide apparatuses and devices corresponding to the above-described method, as shown in fig. 5 and fig. 6.

Fig. 5 is a schematic structural diagram of an anomaly locating apparatus based on a knowledge-graph according to one or more embodiments of the present specification, in which a dashed box represents an optional module, and the apparatus includes:

a knowledge graph determining module 502 for determining a knowledge graph containing monitored nodes and relationships between the nodes;

the abnormal node determining module 504 is configured to obtain monitoring data corresponding to the monitored nodes, and determine a plurality of the monitored nodes as abnormal nodes according to abnormal conditions reflected by the monitoring data;

an association determining module 506, configured to determine, according to the topology in the knowledge graph, an association between the transaction conditions corresponding to the transaction nodes;

a screening module 508 for screening at least part of the transaction conditions from the transaction conditions according to the association;

the abnormal node determining module 510 determines an abnormal node in the abnormal nodes according to the abnormal conditions remaining after the screening.

Optionally, the association determining module 506 includes a dependency relationship determining sub-module 5062 and an association determining sub-module 5064;

the dependency relationship determining submodule 5062 determines that the relationship between the nodes corresponding to the transaction nodes is a dependency relationship according to the topology in the knowledge graph;

the association determining submodule 5064 determines, according to the dependency relationship, an association between transaction conditions corresponding to the transaction nodes.

Optionally, the association determining submodule 5064 determines, according to the dependency relationship, that a causal relationship exists among a plurality of transaction conditions corresponding to the transaction node;

the culling module 508 culls at least a portion of the resulting transaction cases from the plurality of transaction cases based on the causal relationship.

Optionally, the monitored node represents an interface or a policy usable by the interface, and the relationship between nodes includes a dependency relationship between interfaces or a dependency relationship between an interface and a policy.

Optionally, the knowledge-graph further includes a monitoring node that connects and monitors the monitored nodes; the transaction node determining module 504 includes: a monitoring data acquisition sub-module 5042 and a transaction node determination sub-module 5044;

the monitoring data acquisition sub-module 5042 is configured to acquire monitoring data acquired by the monitoring node for a specified index of the monitored node;

the abnormal node determining submodule 5044 is configured to determine that abnormal conditions exist in a first index of a first monitored node and a second index of a second monitored node according to the monitoring data;

and determining the first monitored node and the second monitored node as the abnormal node.

Optionally, the screening module 508 includes an identification sub-module 5082, a screening sub-module 5084;

the identifying sub-module 5082, identifying whether the service belongs to a predetermined service scenario according to the association, the transaction condition and the service definition of the corresponding specified index;

the screening submodule 5084 screens out at least part of the transaction situation from the transaction situation according to the service scenario if the service scenario is successfully identified.

Optionally, the service scenario includes: responding to an excited service surge scene and a batch authority control scene corresponding to an attack;

the screening submodule 5084 screens all the abnormal situations corresponding to the specified indexes associated with the service scenario from the abnormal situations if the service scenario is successfully identified.

Optionally, the apparatus further comprises: an abnormal link determination module 512;

the abnormal link determining module 512 determines an abnormal link for reflecting the abnormal condition conducting direction in the knowledge graph according to the abnormal node and the dependency corresponding to the abnormal node.

Optionally, the apparatus further comprises: a knowledge graph generation module 514;

the knowledge graph generating module 514 determines a traffic flow direction between monitored nodes according to online traffic data of the monitored nodes;

and generating the knowledge graph according to the service flow direction.

Fig. 6 is a schematic structural diagram of an anomaly locating apparatus based on a knowledge-graph according to one or more embodiments of the present specification, where the apparatus includes:

at least one processor; and the number of the first and second groups,

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

Based on the same idea, one or more embodiments of the present specification further provide a non-volatile computer storage medium corresponding to the above method, and storing computer-executable instructions configured to:

determining a knowledge graph containing monitored nodes and relationships among the nodes;

In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as ABEL (Advanced Boolean Expression Language), AHDL (alternate Hardware Description Language), traffic, CUPL (core universal Programming Language), HDCal, jhddl (Java Hardware Description Language), lava, lola, HDL, PALASM, rhyd (Hardware Description Language), and vhigh-Language (Hardware Description Language), which is currently used in most popular applications. It will also be apparent to those skilled in the art that hardware circuitry for implementing the logical method flows can be readily obtained by a mere need to program the method flows with some of the hardware description languages described above and into an integrated circuit.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, atmel AT91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller in purely computer readable program code means, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functions of the various elements may be implemented in the same one or more software and/or hardware implementations of the present description.

As will be appreciated by one skilled in the art, the present specification embodiments may be provided as a method, system, or computer program product. Accordingly, embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.

The description has been presented with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may 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 means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.

This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the embodiments of the apparatus, the device, and the nonvolatile computer storage medium, since they are substantially similar to the embodiments of the method, the description is simple, and for the relevant points, reference may be made to the partial description of the embodiments of the method.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The above description is merely one or more embodiments of the present disclosure and is not intended to limit the present disclosure. Various modifications and alterations to one or more embodiments of the present description will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of one or more embodiments of the present specification should be included in the scope of the claims of the present specification.

Claims

1. A knowledge graph-based anomaly positioning method comprises the following steps:

determining the association between the transaction conditions corresponding to the transaction nodes according to the topology in the knowledge graph, which specifically includes: determining the relationship among the nodes corresponding to the abnormal nodes as a dependency relationship according to the topology in the knowledge graph, and determining that causal relationships exist among a plurality of abnormal conditions corresponding to the abnormal nodes according to the dependency relationship;

according to the association, screening at least part of the abnormal situations from the abnormal situations, specifically comprising: screening at least part of the resulting transaction events from the plurality of transaction events according to the causal relationship;

2. The method according to claim 1, wherein the determining, according to the topology in the knowledge-graph, the association between the transaction conditions corresponding to the transaction nodes specifically comprises:

and determining the association between the abnormal conditions corresponding to the abnormal nodes according to the dependency relationship.

3. The method of claim 1 or 2, the monitored nodes represent interfaces or policies usable by the interfaces, and the inter-node relationships include dependencies between the interfaces or dependencies between the interfaces and the policies.

4. The method of claim 1, the knowledge-graph further comprising a monitoring node connecting and monitoring the monitored node;

the acquiring of the monitoring data corresponding to the monitored node and the determining of the plurality of monitored nodes as the abnormal node according to the abnormal situation reflected by the monitoring data specifically include:

acquiring monitoring data acquired by the monitoring node aiming at the specified index of the monitored node;

determining that the first index of the first monitored node and the second index of the second monitored node have abnormal conditions according to the monitoring data;

5. The method according to claim 4, wherein the screening out at least part of the abnormal situations according to the association specifically comprises:

identifying whether the business belongs to a preset business scene according to the association, the transaction condition and the corresponding business definition of the designated index;

and if the service scene is successfully identified, screening at least part of the abnormal conditions from the abnormal conditions according to the service scene.

6. The method of claim 5, the traffic scenario comprising: responding to an excited service surge scene and a batch authority control scene corresponding to an attack;

if the service scenario is successfully identified, screening at least part of the transaction situations from the transaction situations according to the service scenario, specifically including:

if the service scene is successfully identified, screening all the transaction conditions corresponding to the specified indexes associated with the service scene from the transaction conditions.

7. The method of claim 1, said method further comprising, after determining an anomalous node in said transaction node:

and determining an abnormal link for reflecting the conduction direction of the abnormal condition in the knowledge graph according to the abnormal node and the dependency relationship corresponding to the abnormal node.

8. The method of claim 1, prior to determining the knowledge-graph containing the monitored nodes and relationships between the nodes, the method further comprising:

determining the service flow direction between the monitored nodes according to the on-line service data of the monitored nodes;

and generating the knowledge graph according to the service flow direction.

9. A knowledge-graph-based anomaly locating apparatus comprising:

an association determining module, configured to determine, according to the topology in the knowledge graph, an association between transaction conditions corresponding to the transaction nodes, where the association determining module specifically includes: the dependency relationship determining submodule included in the association determining module determines the inter-node relationship corresponding to the abnormal node to be the dependency relationship according to the topology in the knowledge graph, and the association determining submodule included in the association determining module determines that causal relationships exist among a plurality of abnormal conditions corresponding to the abnormal node according to the dependency relationship;

a screening module for screening at least part of the abnormal situations from the abnormal situations according to the association, specifically comprising: screening at least part of the resulting transaction events from the plurality of transaction events according to the causal relationship;

10. The apparatus according to claim 9, wherein the association determining sub-module determines, according to the dependency relationship, an association between transaction conditions corresponding to the transaction nodes.

11. The apparatus according to claim 9 or 10, the monitored node represents an interface or a policy usable by the interface, and the inter-node relationship includes a dependency relationship between interfaces or a dependency relationship between an interface and a policy.

12. The apparatus of claim 9, the knowledge-graph further comprising a monitoring node that connects to and monitors the monitored node; the transaction node determination module comprises: a monitoring data acquisition submodule and a transaction node determination submodule;

the monitoring data acquisition submodule acquires monitoring data acquired by the monitoring node aiming at the specified index of the monitored node;

the abnormal node determining submodule determines that abnormal conditions exist in a first index of a first monitored node and a second index of a second monitored node according to the monitoring data;

13. The apparatus of claim 12, the screening module comprising an identification submodule, a screening submodule;

the identification submodule identifies whether the business scene belongs to a preset business scene or not according to the association, the transaction condition and the corresponding business definition of the specified index;

and the screening submodule screens at least part of the abnormal conditions from the abnormal conditions according to the service scene if the service scene is successfully identified.

14. The apparatus of claim 13, the traffic scenario comprising: responding to an excited service surge scene and responding to an attack batch authority control scene;

and if the service scene is successfully identified, the screening submodule screens all the abnormal conditions corresponding to the specified indexes associated with the service scene from the abnormal conditions.

15. The apparatus of claim 9, the apparatus further comprising: an abnormal link determination module;

and the abnormal link determining module is used for determining an abnormal link for reflecting the conduction direction of an abnormal condition in the knowledge graph according to the abnormal node and the dependency corresponding to the abnormal node.

16. The apparatus of claim 9, the apparatus further comprising: a knowledge graph generation module;

the knowledge graph generation module is used for determining the service flow direction between the monitored nodes according to the online service data of the monitored nodes;

and generating the knowledge graph according to the service flow direction.

17. A knowledge-graph based anomaly locating apparatus comprising:

at least one processor; and the number of the first and second groups,

determining the association between the transaction conditions corresponding to the transaction nodes according to the topology in the knowledge graph, which specifically includes: determining the relationship among the nodes corresponding to the abnormal nodes as a dependency relationship according to the topology in the knowledge graph, and determining the causal relationship among a plurality of abnormal conditions corresponding to the abnormal nodes according to the dependency relationship;