Business request processing method, device, equipment and medium

Abstract

The disclosed embodiment discloses a method, device, equipment and medium for processing business requests, the method comprising: intercepting real-time interaction data, and obtaining at least one real-time business request according to the real-time interaction data; sending a business request rejection instruction in response to determining that the real-time business request includes at least one illegal business parameter; obtaining a target business request model, taking at least one illegal business parameter as input, and inputting the target business request model to obtain a simulated illegal business request; obtaining simulated business response information according to the simulated illegal business request, and obtaining illegal business request countermeasure information according to the simulated business response information; sending illegal business request countermeasure information. This technical solution can eliminate the probability of the business system crashing due to damage to the business system carried by the business server caused by other business requests sent by the client, thereby improving the stability of the business system.

Description

Technical Field

[0001] The present disclosure relates to the field of network technology, and in particular to a method, device, equipment and medium for processing a service request. Background Art

[0002] In recent years, with the development of communication technology, people's dependence on the network in all aspects of daily life has increased. In daily use, the terminal may run sensitive applications such as financial transaction applications, social applications, etc., so that the terminal may interact with the server for sensitive services. In order to ensure the security of sensitive services, it is necessary to detect the interactive data between the terminal and the server to ensure that the service request initiated by the terminal to the server is safe and reliable and will not cause damage to the service system on the server.

[0003] In the related technology, in order to detect the interaction data between the terminal and the server, the interaction data corresponding to the relevant business can be manually analyzed to obtain sample interaction data outside the scope of normal business rules, and the data to be detected can be obtained through front-end code debugging or packet capture and other methods. The data to be detected is sent to the detection end and compared with the above sample interaction data to obtain the detection result. When it is determined according to the detection result that the business request initiated by the terminal to the server may cause damage to the business system on the server, the terminal is prohibited from interacting with the server.

[0004] Although the above scheme can improve the reliability of the business system on the server, considering that the above scheme mainly relies on relevant analysts to fully understand the business functions and obtain sample interaction data outside the scope of normal business rules based on the experience of the analysts, it may not be possible to determine all business requests that may cause damage to the business system based on the above detection results, and thus it is impossible to reasonably process the corresponding business requests, which may cause the business system to crash due to damage, thereby reducing the stability of the business system. Summary of the invention

[0005] In order to solve the problems in the related art, the embodiments of the present disclosure provide a service request processing method, apparatus, device and medium.

[0006] In a first aspect, an embodiment of the present disclosure provides a method for processing a service request, the method comprising:

[0007] Intercepting real-time interaction data between the client and the service server, and obtaining at least one real-time service request sent by the client to the service server according to the real-time interaction data;

[0008] In response to determining that the real-time service request includes at least one illegal service parameter, sending a service request rejection instruction, where the service request rejection instruction is used to instruct the service server to reject data interaction with the client;

[0009] Obtain a pre-trained target service request model, and input at least one illegal service parameter as input into the target service request model to obtain a simulated illegal service request output by the target service request model;

[0010] Acquire simulated service response information according to the simulated illegal service request, and acquire illegal service request countermeasure information according to the simulated service response information, wherein the illegal service request countermeasure information is used to indicate a strategy for the service server to process the service request sent by the client;

[0011] Send illegal business request countermeasure information.

[0012] In an implementation of the present disclosure, before at least one illegal service parameter is input as an input into a target service request model to obtain a simulated illegal service request output by the target service request model, the method further includes:

[0013] Obtaining a target client identifier corresponding to the client sending the real-time service request;

[0014] At least one illegal business parameter is used as input into a target business request model to obtain a simulated illegal business request output by the target business request model, including:

[0015] At least one illegal business parameter and a target client identifier are input as inputs to a target business request model to obtain a simulated illegal business request output by the target business request model.

[0016] In one implementation of the present disclosure, before obtaining the pre-trained target service request model, the method further includes:

[0017] Obtain historical interaction data between the client and the business server and historical logs of the business server;

[0018] In response to the historical log including at least one fault event, a fault time and a fault service parameter corresponding to the fault event are obtained, and at least one associated service request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to a client sending the associated service request are obtained according to the historical interaction data, wherein parameter values ​​of the service parameters in the associated service request are all outside a legal value range of the corresponding service parameters;

[0019] A service request model is obtained, the fault service parameters and the associated client identifier are taken as input, the associated service request is taken as output, and the service request model is trained to obtain a target service request model.

[0020] In an implementation of the present disclosure, the fault service parameter and the associated client identifier are used as input, and the associated service request is used as output, and the service request model is trained to obtain a target service request model, including:

[0021] Obtain a private business request model according to the business request model;

[0022] receiving the updated weight parameter sent by the edge server, and updating the private service request model according to the updated weight parameter;

[0023] Taking the fault service parameters and the associated client identifier as input and the associated service request as output, training the updated service request model;

[0024] When the trained service request model has not converged, a gradient update vector is obtained according to the trained service request model, and the gradient update vector is sent, and the edge server is used to aggregate the gradient update vector, and update the weight parameter of the common service request model of the edge server according to the aggregated gradient update vector to obtain the updated weight parameter;

[0025] When the trained service request model converges, a target service request model is obtained according to the trained service request model.

[0026] In an implementation of the present disclosure, before obtaining the private service request model according to the service request model, the method further includes:

[0027] Receive private data upload instructions;

[0028] In response to the private data upload instruction, send the fault service parameter, the associated client identifier and the associated service request;

[0029] Receiving initial weight parameters sent by the edge server;

[0030] Obtain a private business request model based on the business request model, including:

[0031] The initial service request model is updated according to the initial weight parameters to obtain a private service request model.

[0032] In a second aspect, an embodiment of the present disclosure provides a service request processing device, the device comprising:

[0033] A real-time request acquisition module is configured to intercept the real-time interaction data between the client and the service server, and acquire at least one real-time service request sent by the client to the service server according to the real-time interaction data;

[0034] An illegal request rejection module is configured to send a service request rejection instruction in response to determining that the real-time service request includes at least one illegal service parameter, wherein the service request rejection instruction is used to instruct the service server to reject data interaction with the client;

[0035] A simulated request acquisition module is configured to acquire a pre-trained target service request model, and input at least one illegal service parameter as input into the target service request model to acquire a simulated illegal service request output by the target service request model;

[0036] A request countermeasure acquisition module is configured to acquire simulated service response information according to the simulated illegal service request, and acquire illegal service request countermeasure information according to the simulated service response information, wherein the illegal service request countermeasure information is used to indicate a strategy for the service server to process the service request sent by the client;

[0037] The request countermeasure sending module is configured to send illegal service request countermeasure information.

[0038] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0039] Obtaining a target client identifier corresponding to the client sending the real-time service request;

[0040] At least one illegal business parameter is used as input into a target business request model to obtain a simulated illegal business request output by the target business request model, including:

[0041] At least one illegal business parameter and a target client identifier are input as inputs to a target business request model to obtain a simulated illegal business request output by the target business request model.

[0042] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0043] Obtain historical interaction data between the client and the business server and historical logs of the business server;

[0044] In response to the historical log including at least one fault event, a fault time and a fault service parameter corresponding to the fault event are obtained, and at least one associated service request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to a client sending the associated service request are obtained according to the historical interaction data, wherein parameter values ​​of the service parameters in the associated service request are all outside a legal value range of the corresponding service parameters;

[0045] A service request model is obtained, the fault service parameters and the associated client identifier are taken as input, the associated service request is taken as output, and the service request model is trained to obtain a target service request model.

[0046] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0047] Obtain a private business request model according to the business request model;

[0048] receiving the updated weight parameter sent by the edge server, and updating the private service request model according to the updated weight parameter;

[0049] Taking the fault service parameters and the associated client identifier as input and the associated service request as output, training the updated service request model;

[0050] When the trained service request model has not converged, a gradient update vector is obtained according to the trained service request model, and the gradient update vector is sent, and the edge server is used to aggregate the gradient update vector, and update the weight parameter of the common service request model of the edge server according to the aggregated gradient update vector to obtain the updated weight parameter;

[0051] When the trained service request model converges, a target service request model is obtained according to the trained service request model.

[0052] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0053] Receive private data upload instructions;

[0054] In response to the private data upload instruction, send the fault service parameter, the associated client identifier and the associated service request;

[0055] Receiving initial weight parameters sent by the edge server;

[0056] Obtain a private business request model based on the business request model, including:

[0057] The initial service request model is updated according to the initial weight parameters to obtain a private service request model.

[0058] In a third aspect, an embodiment of the present disclosure provides an electronic device, the electronic device comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement any one of the methods in the first aspect.

[0059] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions, when executed by a processor, implement any one of the methods in the first aspect.

[0060] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0061] In the above technical solution, by intercepting the real-time interaction data between the client and the service server, and obtaining at least one real-time service request sent by the client to the service server according to the real-time interaction data, it is determined whether the real-time service request includes illegal service parameters, that is, whether the real-time service request may cause damage to the service system carried by the service server. When it is determined that the real-time service request includes at least one illegal service parameter, that is, the real-time service request may cause damage to the service system carried by the service server, in response to determining that the real-time service request includes at least one illegal service parameter, a service request rejection instruction is sent to instruct the service server to refuse data interaction with the client, so as to avoid damage to the service system carried by the service server as much as possible. At the same time, a pre-trained target service request model is obtained, and at least one illegal service parameter is used as input and input into the target service request model to obtain a simulated illegal service request output by the target service request model, wherein the simulated illegal service request can be understood as, when it is determined that the client will send a real-time service request including at least one illegal service parameter, before the client sends the real-time service request, the client may have sent other service requests to the service server that may cause damage to the service system carried by the service server. Simulated business response information is obtained based on the simulated illegal business request, and illegal business request countermeasure information is obtained based on the simulated business response information, and the illegal business request countermeasure information is sent to instruct the business server to perform corresponding processing based on the illegal business request countermeasure information, so as to eliminate the possibility of the business system crashing due to damage to the business system carried by the business server caused by other business requests sent by the client, thereby improving the stability of the business system.

[0062] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Other features, objectives and advantages of the present disclosure will become more apparent through the following detailed description of non-limiting embodiments in conjunction with the accompanying drawings. In the accompanying drawings:

[0064] Figure 1 A schematic structural block diagram of a service request processing system according to an embodiment of the present disclosure is shown;

[0065] Figure 2 A schematic flow chart showing a method for processing a service request according to an embodiment of the present disclosure;

[0066] Figure 3 A schematic flow chart showing a method for processing a service request according to an embodiment of the present disclosure;

[0067] Figure 4 A schematic flow chart showing a method for processing a service request according to an embodiment of the present disclosure;

[0068] Figure 5 A schematic flow chart showing a method for processing a service request according to an embodiment of the present disclosure;

[0069] Figure 6 A schematic structural block diagram of a service request processing device according to an embodiment of the present disclosure is shown;

[0070] Figure 7 A schematic structural block diagram of an electronic device according to an embodiment of the present disclosure is shown;

[0071] Figure 8 A schematic diagram showing the structure of a computer system suitable for implementing the method according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0072] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. In addition, for the sake of clarity, parts not related to the description of the exemplary embodiments are omitted in the accompanying drawings.

[0073] In the present disclosure, it should be understood that terms such as "include" or "have" are intended to indicate the presence of features, numbers, steps, behaviors, components, parts, or a combination thereof disclosed in the specification, and do not exclude the possibility that one or more other features, numbers, steps, behaviors, components, parts, or a combination thereof exist or are added.

[0074] It should also be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0075] The details of the embodiments of the present disclosure are described in detail below through specific examples.

[0076] In recent years, with the development of communication technology, people have become increasingly dependent on the Internet in all aspects of their daily lives. Usually, the terminal devices used by people can access the Internet through network devices that are closer to them. Among them, the terminal devices can be smart mobile communication terminals, wearable devices, desktops, laptops, tablets, etc., and the terminal devices can be wireless routers, wired routers, wireless gateways, etc.

[0077] In recent years, with the development of communication technology, people's dependence on the network in all aspects of daily life has increased. In daily use, the terminal may run sensitive applications such as financial transaction applications, social applications, etc., so that the terminal may interact with the server for sensitive services. In order to ensure the security of sensitive services, it is necessary to detect the interactive data between the terminal and the server to ensure that the service request initiated by the terminal to the server is safe and reliable and will not cause damage to the service system on the server.

[0078] In the related technology, in order to detect the interaction data between the terminal and the server, the interaction data corresponding to the relevant business can be manually analyzed to obtain sample interaction data outside the scope of normal business rules, and the data to be detected can be obtained through front-end code debugging or packet capture and other methods. The data to be detected is sent to the detection end and compared with the above sample interaction data to obtain the detection result. When it is determined according to the detection result that the business request initiated by the terminal to the server may cause damage to the business system on the server, the terminal is prohibited from interacting with the server.

[0079] Although the above scheme can improve the reliability of the business system on the server, considering that the above scheme mainly relies on relevant analysts to fully understand the business functions and obtain sample interaction data outside the scope of normal business rules based on the experience of the analysts, it may not be possible to determine all business requests that may cause damage to the business system based on the above detection results, and thus it is impossible to reasonably process the corresponding business requests, which may cause the business system to crash due to damage, thereby reducing the stability of the business system.

[0080] In view of the above defects, in the technical solution provided by the present disclosure, by intercepting the real-time interaction data between the client and the business server, and obtaining at least one real-time business request sent by the client to the business server according to the real-time interaction data, it is determined whether the real-time business request includes illegal business parameters, that is, whether the real-time business request may cause damage to the business system carried by the business server. When it is determined that the real-time business request includes at least one illegal business parameter, that is, the real-time business request may cause damage to the business system carried by the business server, in response to determining that the real-time business request includes at least one illegal business parameter, a business request rejection instruction is sent to instruct the business server to refuse data interaction with the client, so as to avoid damage to the business system carried by the business server as much as possible. At the same time, a pre-trained target business request model is obtained, and at least one illegal business parameter is used as input and input into the target business request model to obtain a simulated illegal business request output by the target business request model, wherein the simulated illegal business request can be understood as, when it is determined that the client will send a real-time business request including the at least one illegal business parameter, before the client sends the real-time business request, the client may have sent other business requests to the business server that may cause damage to the business system carried by the business server. Simulated business response information is obtained based on the simulated illegal business request, and illegal business request countermeasure information is obtained based on the simulated business response information, and the illegal business request countermeasure information is sent to instruct the business server to perform corresponding processing based on the illegal business request countermeasure information, so as to eliminate the possibility of the business system crashing due to damage to the business system carried by the business server caused by other business requests sent by the client, thereby improving the stability of the business system.

[0081] Figure 1 A schematic structural block diagram of a business request processing system according to an embodiment of the present disclosure is shown, wherein the business request processing system includes a client 101, a business server 102 and a network 103, and the network 103 is used to provide a medium for a communication link between the client 101 and the business server 102. The network 103 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0082] The client 101 is a variety of electronic devices with network access function, including but not limited to mobile communication terminals, desktop computers, tablet computers, laptop computers, wearable devices, etc.

[0083] The service server 102 may be a cloud server or a server provided by the operator of the corresponding service. It should be noted that in the embodiment of the present application, one service server 102 may correspond to multiple clients 101. For example, the operator of the service request processing service may divide the area under its jurisdiction into multiple blocks, and multiple clients 101 in each area correspond to one service server 102.

[0084] Figure 2 A schematic flow chart of a service request processing method according to an embodiment of the present disclosure is shown. The service request processing method is applied to a service server management end, which may be a terminal or a server. Figure 2 As shown, the service request processing method includes the following steps:

[0085] In step S101, real-time interaction data between the client and the service server is intercepted, and at least one real-time service request sent by the client to the service server is obtained according to the real-time interaction data.

[0086] In one implementation of the present disclosure, intercepting the real-time interaction data between the client and the business server can be understood as setting a corresponding interception module on the business server, and the interception module samples the interaction data between the client and the business server. When the sampled data meets the preset interception conditions, the interaction data within the preset data range before and after the sampled data that meets the interception conditions is obtained, and the obtained interaction data is forwarded as real-time interaction data. Among them, the sampled data that meets the interception conditions can be understood as sampled data containing parameters related to the business request. Alternatively, the business server management end can directly sample the interaction data between the client and the business server, and send the sampled interaction data as real-time interaction data.

[0087] In an implementation of the present disclosure, obtaining at least one real-time service request sent by a client to a service server according to real-time interaction data can be understood as parsing the real-time interaction data to obtain at least one real-time service request. Alternatively, it can also be understood as forwarding to other devices or systems to obtain corresponding information sent by other devices or systems to indicate at least one real-time service request, and obtaining the at least one real-time service request according to the corresponding information.

[0088] In step S102, in response to determining that the real-time service request includes at least one illegal service parameter, a service request rejection instruction is sent.

[0089] The service request rejection instruction is used to instruct the service server to reject data interaction with the client.

[0090] In an implementation of the present disclosure, determining that a real-time service request includes at least one illegal service parameter can be understood as extracting service parameters from each real-time service request, and querying a pre-acquired service parameter value database based on the extracted service parameters, and when the service parameter value corresponding to the queried service parameter exceeds the reasonable range of the corresponding service parameter value, determining that the real-time service request includes at least one illegal service parameter. Alternatively, the real-time service request can be sent to other devices or systems, and determining whether the real-time service request includes at least one illegal service parameter based on the extraordinary service parameter indication information sent by the other devices or systems.

[0091] In one implementation of the present disclosure, the service request rejection instruction is used to instruct the service server to reject data interaction with the client. It can be understood that after receiving the service request rejection instruction, the service server discards all data sent by the client indicated by the service request rejection instruction.

[0092] Furthermore, the service request rejection instruction may also include a time interval during which the service server refuses to interact with the client.

[0093] In step S103, a pre-trained target service request model is obtained, and at least one illegal service parameter is input as an input into the target service request model to obtain a simulated illegal service request output by the target service request model.

[0094] In one embodiment of the present disclosure, the target service request model may be pre-stored in the terminal device, or may be obtained from other devices or systems.

[0095] The target service request model may be a neural network (NN) model, a convolutional neural network (CNN) model, or a long short-term memory network (LSTM) model, etc.

[0096] The target business request model can be understood as a model that has learned the regularity between illegal business parameters and the client that sends a real-time business request including the illegal business parameters, and other business requests that may have been sent to the business server before sending the real-time business request including the illegal business parameters, which may cause damage to the business system carried by the business server.

[0097] In step S104, simulated service response information is obtained according to the simulated illegal service request, and illegal service request countermeasure information is obtained according to the simulated service response information.

[0098] The illegal service request countermeasure information is used to indicate the strategy of the service server for processing the service request sent by the client.

[0099] In one embodiment of the present disclosure, obtaining simulated business response information according to a simulated illegal business request can be understood as responding to the simulated illegal business request in a virtual business system previously built to simulate the business system to obtain the simulated business response information; it can also be understood as sending the simulated illegal business request to other devices or systems to obtain the simulated business response information. Among them, the simulated business response information can be understood as indicating the system operation status of the business system when the business system responds to the simulated illegal business request. For example, the simulated business response information may include system operation parameters or business system failure events.

[0100] In one embodiment of the present disclosure, obtaining illegal business request countermeasure information based on simulated business response information can be understood as querying the simulated business response information in a previously acquired illegal business request countermeasure database to obtain the illegal business request countermeasure information; or it can be sending the simulated business response information to other devices or systems to obtain the illegal business request countermeasure information sent by other devices or systems.

[0101] In one embodiment of the present disclosure, illegal business request countermeasure information can be understood as being used to instruct the business system mounted on the business server how to process the corresponding business application and the related data generated in response to the corresponding business application, the software module activated in response to the corresponding business application, etc., to avoid any impact of the corresponding business application on the business system.

[0102] In step S105, illegal service request countermeasure information is sent.

[0103] In this embodiment, by intercepting the real-time interaction data between the client and the service server, and obtaining at least one real-time service request sent by the client to the service server according to the real-time interaction data, it is determined whether the real-time service request includes illegal service parameters, that is, whether the real-time service request may cause damage to the service system carried by the service server. When it is determined that the real-time service request includes at least one illegal service parameter, that is, the real-time service request may cause damage to the service system carried by the service server, in response to determining that the real-time service request includes at least one illegal service parameter, a service request rejection instruction is sent to instruct the service server to refuse data interaction with the client, so as to avoid damage to the service system carried by the service server as much as possible. At the same time, a pre-trained target service request model is obtained, and at least one illegal service parameter is used as input and input into the target service request model to obtain a simulated illegal service request output by the target service request model, wherein the simulated illegal service request can be understood as, when it is determined that the client will send a real-time service request including at least one illegal service parameter, before the client sends the real-time service request, the client may have sent other service requests to the service server that may cause damage to the service system carried by the service server. Simulated business response information is obtained based on the simulated illegal business request, and illegal business request countermeasure information is obtained based on the simulated business response information, and the illegal business request countermeasure information is sent to instruct the business server to perform corresponding processing based on the illegal business request countermeasure information, so as to eliminate the possibility of the business system crashing due to damage to the business system carried by the business server caused by other business requests sent by the client, thereby improving the stability of the business system.

[0104] In one implementation of the present disclosure, Figure 3 A schematic flow chart of a method for processing a service request according to an embodiment of the present disclosure is shown as follows: Figure 3 As shown, in step S103, before obtaining a pre-trained target service request model, taking at least one illegal service parameter as input and inputting it into the target service request model to obtain a simulated illegal service request output by the target service request model, the method further includes step S106:

[0105] In step S106, a target client identifier corresponding to the client sending the real-time service request is obtained;

[0106] In step S103, a pre-trained target service request model is obtained, and at least one illegal service parameter is input into the target service request model to obtain a simulated illegal service request output by the target service request model. This can be achieved by step S1031:

[0107] In step S1031, a pre-trained target business request model is obtained, and at least one illegal business parameter and a target client identifier are input into the target business request model to obtain a simulated illegal business request output by the target business request model.

[0108] In an embodiment of the present disclosure, the client identifier may be understood as being used to indicate a client, and the client identifier may correspond to the client one-to-one.

[0109] In this embodiment, by obtaining the target client identifier corresponding to the client sending the real-time business request, and taking at least one illegal business parameter and the target client identifier as input, and inputting the target business request model to obtain a simulated illegal business request output by the target business request model, it can be ensured that the simulated illegal business request is for a specific client that has sent a real-time business request including the at least one illegal business parameter. Before sending the real-time business request, the specific client may have sent other business requests to the business server that may cause damage to the business system carried by the business server, thereby improving the accuracy of the obtained simulated illegal business requests.

[0110] In one implementation of the present disclosure, Figure 4 A schematic flow chart of a method for processing a service request according to an embodiment of the present disclosure is shown as follows: Figure 4 As shown, in step S1031, a pre-trained target service request model is obtained, at least one illegal service parameter and a target client identifier are used as input, and input into the target service request model to obtain a simulated illegal service request output by the target service request model, and the method further includes steps S107-S109:

[0111] In step S107, historical interaction data between the client and the service server and historical logs of the service server are obtained;

[0112] In step S108, in response to the historical log including at least one fault event, the fault time and fault service parameters corresponding to the fault event are obtained, and at least one associated service request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to a client sending the associated service request are obtained according to the historical interaction data;

[0113] The parameter values ​​of the service parameters in the associated service request are all outside the legal value range of the corresponding service parameters;

[0114] In step S109, a service request model is obtained, the fault service parameters and the associated client identifier are taken as input, and the associated service request is taken as output, and the service request model is trained to obtain a target service request model.

[0115] In one embodiment of the present disclosure, historical interaction data between the client and the service server and historical logs of the service server can be obtained by reading pre-stored historical interaction data and historical logs, or by receiving historical interaction data and historical logs sent by other devices or systems.

[0116] In this embodiment, by acquiring historical interaction data between the client and the business server and historical logs of the business server, in response to the historical logs including at least one fault event, the fault time and fault business parameters corresponding to the fault event are acquired, and at least one associated business request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to the client sending the associated business request are acquired according to the historical interaction data, and a business request model is acquired, and the fault business parameters and the associated client identifier are used as inputs and the associated business request is used as output to train the business request model to acquire a target business request model, so as to ensure that the target business request model can learn the regularity between illegal business parameters and the client identifier of the client sending the real-time business request including the illegal business parameters, and other business requests that may have been sent to the business server by the client before sending the real-time business request including the illegal business parameters, which may cause damage to the business system carried by the business server.

[0117] In one implementation of the present disclosure, Figure 5 A schematic flow chart of a method for processing a service request according to an embodiment of the present disclosure is shown as follows: Figure 5 As shown, step S109, obtaining a service request model, taking the fault service parameter and the associated client identifier as input, taking the associated service request as output, training the service request model to obtain a target service request model, can be implemented through steps S1091-S1095:

[0118] In step S1091, a private service request model is obtained according to the service request model;

[0119] In step S1092, the updated weight parameter sent by the edge server is received, and the private service request model is updated according to the updated weight parameter;

[0120] In step S1093, the fault service parameters and the associated client identifier are used as input, and the associated service request is used as output to train the updated service request model;

[0121] In step S1094, when the trained service request model has not converged, a gradient update vector is obtained according to the trained service request model, and the gradient update vector is sent. The edge server is used to aggregate the gradient update vector, and update the weight parameter of the common service request model of the edge server according to the aggregated gradient update vector to obtain the updated weight parameter;

[0122] In step S1095 , when the trained service request model converges, a target service request model is obtained according to the trained service request model.

[0123] In one embodiment of the present disclosure, both the private service request model and the shared service request model can be a neural network model, a convolutional neural network model, a long short-term memory network model, or the like.

[0124] In one embodiment of the present disclosure, obtaining a target business request model based on a trained private business request model can be understood as storing the trained private business request model as the target business request model, or can be understood as directly identifying the trained private business request model as the target business request model.

[0125] In this implementation manner, the updated weight parameter sent by the edge server is received by the business server management end, which is obtained by the edge server aggregating the gradient update vectors sent by multiple business server management ends, and updating the weight parameter of the common business request model of the edge server according to the aggregated gradient update vector. Therefore, the updated common business request model can reflect the common rules learned by the common business request model of the edge server in the previous round of training, the illegal business parameters obtained by the multiple business server management ends, the client identifier of the client that sends the real-time business request including the illegal business parameter, and other business requests that may have been sent to the business server by the client before sending the real-time business request including the illegal business parameter, which may cause damage to the business system carried by the business server; by taking the faulty business parameters and the associated client identifier as input and the associated business request as output, the updated business request model is trained, so that the updated private business request model can be personalized for the illegal business parameters obtained by the business server management end itself and the client identifier of the client that sends the real-time business request including the illegal business parameter on the basis of learning the above-mentioned common rules. The private law between the identification and other business requests that the client may have sent to the business server before sending the real-time business request including the illegal business parameter, which may cause damage to the business system carried by the business server; when the trained private business request model does not converge, it means that the trained private business request model still needs to be trained. By obtaining the gradient update vector according to the trained private business request model and sending the gradient update vector, the edge server can continue to obtain the corresponding update weight parameter based on the gradient update vectors uploaded by multiple business server management terminals without leaking the private data of the business server management terminal, so as to continue to train the private business request model of each business server management terminal; when the trained private business request model converges, the converged private business request model can obtain other business requests that the client may have sent to the business server before sending the real-time business request including the illegal business parameter, which may cause damage to the business system carried by the business server, based on the illegal business parameter and the client identifier of the client that sends the real-time business request including the illegal business parameter, so the converged private business request model is used as the target business request model.

[0126] In the technical solution provided by the present disclosure, a private business request model is obtained according to a business request model; an updated weight parameter sent by an edge server is received, and the private business request model is updated according to the updated weight parameter; the fault business parameter and the associated client identifier are taken as input, and the associated business request is taken as output to train the updated business request model; when the trained business request model has not converged, a gradient update vector is obtained according to the trained business request model, and the gradient update vector is sent, and the edge server is used to aggregate the gradient update vector, and update the weight parameter of the common business request model of the edge server according to the aggregated gradient update vector to obtain the updated weight parameter; when the trained business request model converges, a target business request model is obtained according to the trained business request model, so as to ensure that the training process can be jointly executed by the edge server and the terminal, and the business request model can not only learn the common rules but also learn the private rules, thereby improving the training efficiency and the user experience.

[0127] In an implementation of the present disclosure, before obtaining the private service request model according to the service request model, the method further includes:

[0128] Receive private data upload instructions;

[0129] In response to the private data upload instruction, send the fault service parameter, the associated client identifier and the associated service request;

[0130] Receiving initial weight parameters sent by the edge server;

[0131] Obtain a private business request model based on the business request model, including:

[0132] The initial service request model is updated according to the initial weight parameters to obtain a private service request model.

[0133] Among them, the initial business request model can be a neural network model, a convolutional neural network model or a long short-term memory network model, etc. The initial business request model can be understood as an untrained model.

[0134] In this implementation, by receiving a private data upload instruction; responding to the private data upload instruction, sending faulty service parameters, associated client identifiers, and associated service requests, the edge server can perform preliminary training on its own initial service request model based on the faulty service parameters, associated client identifiers, and associated service requests, and obtain a common service request model on the edge server. At this time, the common service request model can be understood as a model that has been preliminarily learned about the rules between the faulty service parameters, associated client identifiers, and associated service requests obtained by multiple terminals. Afterwards, by receiving the initial weight parameters sent by the edge server, the initial service request model is updated according to the initial weight parameters to obtain a private service request model. The private service request model can understand the model that has learned the rules learned by the common service request model, thereby facilitating multiple rounds of training of the private service request model, without the need to train based on the initial service request model, thereby reducing the difficulty of training.

[0135] The following are embodiments of the apparatus of the present disclosure, which can be used to execute embodiments of the method of the present disclosure.

[0136] According to a service request processing device according to an embodiment of the present disclosure, the service request processing device can be implemented as part or all of an electronic device through software, hardware, or a combination of both. Figure 6 A schematic structural block diagram of a service request processing device according to an embodiment of the present disclosure is shown as follows: Figure 6 As shown, the service request processing device 200 includes:

[0137] The real-time request acquisition module 201 is configured to intercept the real-time interaction data between the client and the service server, and acquire at least one real-time service request sent by the client to the service server according to the real-time interaction data;

[0138] The illegal request rejection module 202 is configured to send a service request rejection instruction in response to determining that the real-time service request includes at least one illegal service parameter, wherein the service request rejection instruction is used to instruct the service server to reject data interaction with the client;

[0139] The simulated request acquisition module 203 is configured to acquire a pre-trained target service request model, and input at least one illegal service parameter as input into the target service request model to acquire a simulated illegal service request output by the target service request model;

[0140] The request countermeasure acquisition module 204 is configured to acquire simulated service response information according to the simulated illegal service request, and acquire illegal service request countermeasure information according to the simulated service response information, wherein the illegal service request countermeasure information is used to indicate a strategy for the service server to process the service request sent by the client;

[0141] The request countermeasure sending module 205 is configured to send illegal service request countermeasure information.

[0142] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0143] Obtaining a target client identifier corresponding to the client sending the real-time service request;

[0144] At least one illegal business parameter is used as input into a target business request model to obtain a simulated illegal business request output by the target business request model, including:

[0145] At least one illegal business parameter and a target client identifier are input as inputs to a target business request model to obtain a simulated illegal business request output by the target business request model.

[0146] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0147] Obtain historical interaction data between the client and the business server and historical logs of the business server;

[0148] In response to the historical log including at least one fault event, a fault time and a fault service parameter corresponding to the fault event are obtained, and at least one associated service request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to a client sending the associated service request are obtained according to the historical interaction data, wherein parameter values ​​of the service parameters in the associated service request are all outside a legal value range of the corresponding service parameters;

[0149] A service request model is obtained, the fault service parameters and the associated client identifier are taken as input, the associated service request is taken as output, and the service request model is trained to obtain a target service request model.

[0150] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0151] Obtain a private business request model according to the business request model;

[0152] receiving the updated weight parameter sent by the edge server, and updating the private service request model according to the updated weight parameter;

[0153] Taking the fault service parameters and the associated client identifier as input and the associated service request as output, training the updated service request model;

[0154] When the trained service request model has not converged, a gradient update vector is obtained according to the trained service request model, and the gradient update vector is sent, and the edge server is used to aggregate the gradient update vector, and update the weight parameter of the common service request model of the edge server according to the aggregated gradient update vector to obtain the updated weight parameter;

[0155] When the trained service request model converges, a target service request model is obtained according to the trained service request model.

[0156] In one implementation of the present disclosure, the simulation request acquisition module is further configured to:

[0157] Receive private data upload instructions;

[0158] In response to the private data upload instruction, send the fault service parameter, the associated client identifier and the associated service request;

[0159] Receiving initial weight parameters sent by the edge server;

[0160] Obtain a private business request model based on the business request model, including:

[0161] The initial service request model is updated according to the initial weight parameters to obtain a private service request model.

[0162] In the above technical solution, by intercepting the real-time interaction data between the client and the service server, and obtaining at least one real-time service request sent by the client to the service server according to the real-time interaction data, it is determined whether the real-time service request includes illegal service parameters, that is, whether the real-time service request may cause damage to the service system carried by the service server. When it is determined that the real-time service request includes at least one illegal service parameter, that is, the real-time service request may cause damage to the service system carried by the service server, in response to determining that the real-time service request includes at least one illegal service parameter, a service request rejection instruction is sent to instruct the service server to refuse data interaction with the client, so as to avoid damage to the service system carried by the service server as much as possible. At the same time, a pre-trained target service request model is obtained, and at least one illegal service parameter is used as input and input into the target service request model to obtain a simulated illegal service request output by the target service request model, wherein the simulated illegal service request can be understood as, when it is determined that the client will send a real-time service request including at least one illegal service parameter, before the client sends the real-time service request, the client may have sent other service requests to the service server that may cause damage to the service system carried by the service server. Simulated business response information is obtained based on the simulated illegal business request, and illegal business request countermeasure information is obtained based on the simulated business response information, and the illegal business request countermeasure information is sent to instruct the business server to perform corresponding processing based on the illegal business request countermeasure information, so as to eliminate the possibility of the business system crashing due to damage to the business system carried by the business server caused by other business requests sent by the client, thereby improving the stability of the business system.

[0163] The present disclosure also discloses an electronic device, Figure 7 A schematic structural block diagram of an electronic device according to an embodiment of the present disclosure is shown. Figure 7 As shown, the electronic device includes a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement any method in the embodiments of the present disclosure.

[0164] Figure 8 A schematic diagram showing the structure of a computer system suitable for implementing the method according to an embodiment of the present disclosure is shown.

[0165] like Figure 8As shown, the computer system includes a processing unit, which can perform the various methods in the above-mentioned embodiments according to the program stored in the read-only memory (ROM) or the program loaded from the storage part into the random access memory (RAM). In the RAM, various programs and data required for the operation of the computer system are also stored. The processing unit, ROM and RAM are connected to each other through a bus. The input / output (I / O) interface is also connected to the bus.

[0166] The following components are connected to the I / O interface: an input part including a keyboard, a mouse, etc.; an output part including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage part including a hard disk, etc.; and a communication part including a network interface card such as a LAN card, a modem, etc. The communication part performs a communication process via a network such as the Internet. The drive is also connected to the I / O interface as needed. Removable media, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on the drive as needed so that the computer program read therefrom is installed into the storage part as needed. Among them, the processing unit can be implemented as a processing unit such as a CPU, a GPU, a TPU, an FPGA, an NPU, etc.

[0167] In particular, according to an embodiment of the present disclosure, the method described above can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program tangibly contained on a machine-readable medium, and the computer program includes a program code for executing the above method. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium.

[0168] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0169] The units or modules involved in the embodiments described in the present disclosure may be implemented by software or programmable hardware. The units or modules described may also be set in a processor, and the names of these units or modules do not constitute limitations on the units or modules themselves in some cases.

[0170] As another aspect, the present disclosure further provides a computer-readable storage medium, which may be a computer-readable storage medium included in the electronic device or computer system in the above embodiment; or a computer-readable storage medium that exists independently and is not assembled into a device. The computer-readable storage medium stores one or more programs, and the programs are used by one or more processors to execute the method described in the present disclosure.

[0171] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present disclosure is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other.

Claims

1. A method for processing a business request, characterized in that: The method comprises: Intercepting real-time interaction data between a client and a service server, and acquiring at least one real-time service request sent by the client to the service server according to the real-time interaction data; In response to determining that the real-time service request includes at least one illegal service parameter, sending a service request rejection instruction, wherein the service request rejection instruction is used to instruct the service server to reject data interaction with the client; Obtain a pre-trained target business request model; Obtaining a target client identifier corresponding to the client sending the real-time service request; Taking the at least one illegal service parameter and the target client identifier as input, and inputting them into the target service request model, so as to obtain a simulated illegal service request output by the target service request model; Acquire simulated business response information according to the simulated illegal business request, and acquire illegal business request countermeasure information according to the simulated business response information, wherein the illegal business request countermeasure information is used to instruct the business system mounted on the business server how to process the corresponding business application, the relevant data generated in response to the corresponding business application, and the software module activated in response to the corresponding business application, so as to prevent the corresponding business application from causing damage to the business system; The illegal service request countermeasure information is sent.

2. The service request processing method according to claim 1, characterized in that: Before obtaining the pre-trained target service request model, the method further includes: Acquire historical interaction data between the client and the service server and historical logs of the service server; In response to the historical log including at least one fault event, a fault time and a fault service parameter corresponding to the fault event are obtained, and at least one associated service request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to a client sending the associated service request are obtained according to the historical interaction data, wherein parameter values ​​of the service parameters in the associated service request are all outside a legal value range of the corresponding service parameters; A service request model is obtained, the fault service parameter and the associated client identifier are used as input, the associated service request is used as output, and the service request model is trained to obtain the target service request model.

3. The service request processing method according to claim 2, characterized in that: The taking the fault service parameter and the associated client identifier as input, taking the associated service request as output, and training the service request model to obtain the target service request model includes: Acquire a private service request model according to the service request model; receiving an updated weight parameter sent by the edge server, and updating the private service request model according to the updated weight parameter; Taking the fault service parameter and the associated client identifier as input and the associated service request as output, training the updated service request model; When the trained service request model has not converged, a gradient update vector is obtained according to the trained service request model, and the gradient update vector is sent, and the edge server is used to aggregate the gradient update vector, and update the weight parameter of the common service request model of the edge server according to the aggregated gradient update vector to obtain the updated weight parameter; When the trained service request model converges, the target service request model is obtained according to the trained service request model.

4. The service request processing method according to claim 3, characterized in that: Before obtaining the private service request model according to the service request model, the method further includes: Receive private data upload instructions; In response to the private data upload instruction, sending the fault service parameter, the associated client identifier and the associated service request; Receiving initial weight parameters sent by the edge server; The acquiring of the private service request model according to the service request model comprises: The initial service request model is updated according to the initial weight parameter to obtain the private service request model.

5. A service request processing device, characterized in that: The device comprises: A real-time request acquisition module is configured to intercept real-time interaction data between the client and the service server, and acquire at least one real-time service request sent by the client to the service server according to the real-time interaction data; an illegal request rejection module, configured to send a service request rejection instruction in response to determining that the real-time service request includes at least one illegal service parameter, wherein the service request rejection instruction is used to instruct the service server to reject data interaction with the client; The simulation request acquisition module is configured to acquire a pre-trained target service request model, acquire a target client identifier corresponding to the client sending the real-time service request, and input the at least one illegal service parameter and the target client identifier into the target service request model to acquire a simulated illegal service request output by the target service request model; a request countermeasure acquisition module, configured to acquire simulated business response information according to the simulated illegal business request, and acquire illegal business request countermeasure information according to the simulated business response information, wherein the illegal business request countermeasure information is used to instruct the business system mounted on the business server how to process the corresponding business application, the relevant data generated in response to the corresponding business application, and the software module activated in response to the corresponding business application, so as to prevent the corresponding business application from causing damage to the business system; The request countermeasure sending module is configured to send the illegal service request countermeasure information.

6. The service request processing device according to claim 5, characterized in that: The simulation request acquisition module is further configured to: Acquire historical interaction data between the client and the service server and historical logs of the service server; In response to the historical log including at least one fault event, a fault time and a fault service parameter corresponding to the fault event are obtained, and at least one associated service request associated with the fault event within a target time length before the fault time and an associated client identifier corresponding to a client sending the associated service request are obtained according to the historical interaction data, wherein parameter values ​​of the service parameters in the associated service request are all outside a legal value range of the corresponding service parameters; A service request model is obtained, the fault service parameter and the associated client identifier are used as input, the associated service request is used as output, and the service request model is trained to obtain the target service request model.

7. An electronic device, characterized in that: The electronic device comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 4.

8. A readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed by a processor, the method steps described in any one of claims 1 to 4 are implemented.

Images