Recommended method, device and storage medium for fault repair operation

Abstract

The present application discloses a fault repair operation recommendation method, device and storage medium, which belongs to the field of communication technology. In an embodiment of the present application, after obtaining fault information, a repair plan corresponding to the fault information can be found from a plan library. When a repair plan corresponding to the fault information cannot be found in the plan library, the fault information can be processed by a recommendation model to obtain a recommended plan, and then an operation is selected from one or more candidate operations included in the recommended plan as a fault repair operation, which solves the problem that there is no repair plan corresponding to the fault information in the plan library and a repair operation cannot be provided for the faulty node.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a fault repair operation recommendation method, device and storage medium. Background Art

[0002] In a software defined network (SDN) architecture, an SDN controller can centrally control multiple nodes. When any of the multiple nodes fails, the SDN controller can recommend a repair operation for the node to repair the failure.

[0003] In the related art, a plan library can be configured in the SDN controller based on manual experience, and the plan library stores the mapping relationship between fault information and repair plans. After the SDN controller receives fault information from the faulty node, it can search for a repair plan corresponding to the fault information from the plan library based on the fault information. If a repair plan corresponding to the fault information is found, the SDN controller can display the repair operations included in the repair plan found, and manually select a repair operation as a fault repair operation and send it to the faulty node, so that the faulty node can perform fault repair according to the fault repair operation. However, if the repair plan corresponding to the fault information does not exist in the plan library, the SDN controller will not be able to provide a repair operation for the faulty node, resulting in the inability to repair the fault. Summary of the invention

[0004] The present application provides a fault repair operation recommendation method, device and storage medium, which can be used to solve the problem in the related art that when there is no repair plan corresponding to the fault information in the plan library, it is impossible to provide a repair operation for the faulty node.

[0005] The technical solution is as follows:

[0006] In a first aspect, a method for recommending fault repair operations is provided, the method comprising: obtaining fault information of a faulty node; if a repair plan corresponding to the fault information does not exist in a plan library, determining a recommended plan through a recommendation model based on the fault information, the recommended plan including one or more candidate operations; determining a fault repair operation from the one or more candidate operations included in the recommended plan, and recommending the fault repair operation to the faulty node so that the faulty node performs fault repair according to the fault repair operation.

[0007] In an embodiment of the present application, when a repair plan corresponding to the fault information cannot be found in the plan library, the fault information can be processed through a recommendation model to obtain a recommended plan, and then an operation is selected from one or more candidate operations included in the recommended plan as a fault repair operation, thereby solving the problem that there is no repair plan corresponding to the fault information in the plan library and a repair operation cannot be provided for the faulty node.

[0008] Optionally, the fault information includes multiple fault parameters, the parameters of the recommended model include multiple fault characteristic factors, and each of the multiple fault characteristic factors corresponds to one of the multiple fault parameters. On this basis, according to the fault information, the implementation process of determining the recommended plan through the recommended model can be: using the multiple fault parameters as inputs of the recommended model, and determining the recommended plan through the recommended model according to the multiple fault characteristic factors.

[0009] Optionally, the parameters of the recommendation model also include a configuration influence factor, and the configuration influence factor includes the original configuration information of the faulty node and the physical topology information of the network where the faulty node is located. On this basis, the implementation process of determining the fault repair operation from one or more candidate operations included in the recommended plan can be: taking the first candidate operation as the input of the recommendation model, and determining the configuration influence degree corresponding to the first candidate operation through the recommendation model according to the configuration influence factor, the configuration influence degree is the predicted influence degree of the first candidate operation on the configuration of the faulty node and the physical topology of the network where it is located, and the first candidate operation is any candidate operation among the one or more candidate operations; and taking the candidate operation with the smallest configuration influence degree among the one or more candidate operations as the fault repair operation.

[0010] That is, the embodiments of the present application can determine the configuration impact degree of the candidate operation, and recommend a fault repair operation based on the configuration impact degree, so that the recommended fault repair operation has the least impact on the configuration of the faulty node and the physical topology of the network.

[0011] Optionally, after sending the fault repair operation to the faulty node, the method further includes: receiving a repair result fed back by the faulty node after the fault is repaired according to the fault repair operation, the repair result including routing information of the faulty node after the fault is repaired and physical topology information of the network in which the faulty node is located; using the routing information and the physical topology information after the fault is repaired as inputs of a data plane verification (DPV) model, and determining a data plane impact degree corresponding to the fault repair operation through the DPV model, wherein the data plane impact degree is a real impact degree caused by the fault repair operation on the configuration of the faulty node and the physical topology of the network in which the faulty node is located; generating a fault sample according to the data plane impact degree corresponding to the fault repair operation, the fault repair operation and the fault information; and adjusting parameters of the recommendation model according to the fault sample.

[0012] After recommending a fault repair operation, the DPV model can also be used to verify the impact of the fault repair operation on the data plane, and then generate fault samples based on the impact on the data plane, the fault repair operation and the fault information to further adjust the parameters of the recommendation model, thereby improving the accuracy of the recommended plan of the recommendation model.

[0013] In a second aspect, a method for recommending fault repair operations is provided, the method further comprising: obtaining fault information of a faulty node, and if a repair plan corresponding to the fault information exists in a plan library, taking the repair plan corresponding to the fault information as a recommended plan, the recommended plan including one or more candidate operations; predicting a comprehensive impact degree of each candidate operation in the one or more candidate operations, the comprehensive impact degree being used to indicate the size of the comprehensive impact of the corresponding candidate operation on the network where the faulty node is located; determining the candidate operation with the smallest corresponding comprehensive impact degree in the one or more candidate operations as a fault repair operation, and recommending the fault repair operation to the faulty node so that the faulty node performs fault repair according to the fault repair operation.

[0014] If there is a repair plan corresponding to the fault information in the plan library, the comprehensive impact of each candidate operation included in the repair plan can be determined, and then the fault repair operation can be determined based on the comprehensive impact, so as to minimize the comprehensive impact of the fault repair operation on the network where the faulty node is located.

[0015] Optionally, the predicting of the comprehensive impact of each of the one or more candidate operations included in the recommended plan includes: estimating the business impact of each of the one or more candidate operations on the business within the network where the faulty node is located; estimating the configuration impact of each of the one or more candidate operations on the network where the faulty node is located; and determining the comprehensive impact corresponding to each candidate operation based on the business impact and configuration impact corresponding to each candidate operation.

[0016] Optionally, the process of estimating the degree of configuration impact of each of the one or more candidate operations on the network where the faulty node is located may be: obtaining original configuration information of the faulty node and physical topology information of the network where the faulty node is located; generating predicted configuration information corresponding to the first candidate operation according to the first candidate operation and the original configuration information of the faulty node, where the first candidate operation is any candidate operation among the one or more candidate operations; using the predicted configuration information corresponding to the first candidate operation, the original configuration information of the faulty node, and the physical topology information as inputs of a configuration plane verification (CPV) model, and determining the degree of configuration impact corresponding to the first candidate operation through the CPV model.

[0017] Optionally, the implementation process of determining the comprehensive impact degree corresponding to each candidate operation based on the business impact degree and configuration impact degree corresponding to each candidate operation may be: obtaining a business impact weight and a configuration impact weight; and determining the comprehensive impact degree corresponding to the first candidate operation based on the business impact degree corresponding to the first candidate operation, the business impact weight, the configuration impact degree corresponding to the first candidate operation, and the configuration impact weight.

[0018] In a third aspect, a device for recommending a fault repair operation is provided, wherein the device has the function of implementing the fault repair operation recommendation method described in the first aspect or the second aspect. The device for recommending a fault repair operation includes at least one module, and the at least one module is used to implement the fault repair operation recommendation method described in the first aspect.

[0019] In a fourth aspect, a device for recommending a fault repair operation is provided, wherein the structure of the device for recommending a fault repair operation includes a processor and a memory, wherein the memory is used to store a program that supports the device for recommending a fault repair operation to execute the method for recommending a fault repair operation provided in the first aspect or the second aspect, and to store data involved in implementing the method for recommending a fault repair operation provided in the first aspect or the second aspect. The processor is configured to execute the program stored in the memory. The operating device of the storage device may also include a communication bus, which is used to establish a connection between the processor and the memory.

[0020] In a fifth aspect, a computer-readable storage medium is provided, wherein instructions are stored in the computer-readable storage medium, and when the computer-readable storage medium is executed on a computer, the computer executes the fault repair operation recommendation method described in the first aspect or the second aspect.

[0021] In a sixth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the fault repair operation recommendation method described in the first aspect or the second aspect.

[0022] The technical effects obtained by the above-mentioned third, fourth, fifth and sixth aspects are similar to the technical effects obtained by the corresponding technical means in the first and second aspects, and will not be repeated here.

[0023] The beneficial effects of the technical solution provided by this application include at least:

[0024] In an embodiment of the present application, when a repair plan corresponding to the fault information cannot be found in the plan library, the fault information can be processed through a recommendation model to obtain a recommended plan, and then an operation is selected from one or more candidate operations included in the recommended plan as a fault repair operation, thereby solving the problem that there is no repair plan corresponding to the fault information in the plan library and a repair operation cannot be provided for the faulty node. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a system architecture diagram involved in the fault repair operation recommendation method provided in the embodiment of the present application;

[0026] Figure 2 It is a structural diagram of a network device provided in an embodiment of the present application;

[0027] Figure 3 This is a flowchart of a recommended method for fault repair operation provided by an embodiment of the present application;

[0028] Figure 4 is a flowchart of another recommended method for fault repair operation provided by an embodiment of the present application;

[0029] Figure 5 It is a structural schematic diagram of a fault repair operation recommendation device provided in an embodiment of the present application;

[0030] Figure 6 is a structural schematic diagram of another fault repair operation recommendation device provided in an embodiment of the present application;

[0031] Figure 7 It is a structural diagram of another fault repair operation recommendation device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0033] Before explaining the embodiments of the present application in detail, the system architecture involved in the embodiments of the present application is first introduced.

[0034] Figure 1 1 is a system architecture diagram of the fault repair operation recommendation method provided in the embodiment of the present application. Figure 1 As shown in , the system includes a controller 101, an analyzer 102 and multiple physical nodes 103. The controller 101 can communicate with the analyzer 102 and the multiple physical nodes 103 respectively. In addition, the analyzer 102 can also communicate with the multiple physical nodes 103.

[0035] It should be noted that the controller 101 is used to centrally control the allocation of network resources. Exemplarily, the controller 101 can control data forwarding on multiple physical nodes 103. In addition, in the embodiment of the present application, the controller 101 can also receive fault information reported by the analyzer 102, and use the method for recommending fault repair operations provided in the embodiment of the present application to recommend fault repair operations to the faulty devices in the multiple physical nodes 103.

[0036] The analyzer 102 is used to monitor in real time whether each physical node 103 among multiple physical nodes 103 has a fault, and after monitoring a physical node 103 to have a fault, collect the fault information of the faulty node, and report the fault information to the controller 101, so that the controller 101 can recommend a repair operation to the faulty node according to the method provided in the embodiment of the present application.

[0037] The multiple physical nodes 103 may be multiple devices in a physical network. The multiple physical nodes 103 are used to receive and / or send service data.

[0038] It should be noted that the controller 101 and the analyzer 102 may be distributed in two different independent devices. Alternatively, the controller 101 and the analyzer 102 may be integrated into one device, which is not limited in the embodiment of the present application. Alternatively, the controller 101 may have the function of the analyzer 102, in which case the analyzer 102 may not be included in the above system. In addition, the multiple physical nodes 103 may be network devices such as switches and routers, which is not limited in the embodiment of the present application.

[0039] Figure 2 It is a structural diagram of a network device provided in an embodiment of the present application. Figure 1 The controller 101 in the Figure 2 See the network devices shown in Figure 2 The network device includes at least one processor 201, a communication bus 202, a memory 203 and at least one communication interface 204.

[0040] The processor 201 may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present application.

[0041] The communication bus 202 may include a pathway for transmitting information between the above-mentioned components.

[0042] The memory 203 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, an optical disc storage (including a compressed optical disc, a laser disc, an optical disc, a digital versatile disc, a Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 203 may exist independently and be connected to the processor 201 via the communication bus 202. The memory 203 may also be integrated with the processor 201.

[0043] The communication interface 204 uses any transceiver or other device for communicating with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Networks (WLAN), etc.

[0044] In a specific implementation, as an embodiment, the processor 201 may include one or more CPUs, such as Figure 2 CPU0 and CPU1 are shown in the figure.

[0045] In a specific implementation, as an embodiment, a computer device may include multiple processors, such as Figure 2 201 and processor 205 shown in FIG. Each of these processors may be a single-CPU processor or a multi-CPU processor. A processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0046] In a specific implementation, as an embodiment, the computer device may further include an output device 206 and an input device 207. The output device 206 communicates with the processor 201 and may display information in a variety of ways. For example, the output device 206 may be a liquid crystal display (LCD), a light emitting diode (LED) display device, a cathode ray tube (CRT) display device, or a projector. The input device 207 communicates with the processor 201 and may receive user input in a variety of ways. For example, the input device 207 may be a mouse, a keyboard, a touch screen device, or a sensor device.

[0047] The memory 203 is used to store program codes for executing the solution of the present application, and the execution is controlled by the processor 201. The processor 201 is used to execute the program code 208 stored in the memory 203. The program code 208 may include one or more software modules. Figure 1 The controller shown in FIG. 2 can recommend a fault repair operation for a faulty device through the processor 201 and one or more software modules in the program code 208 in the memory 203 .

[0048] Figure 2 The processor in the memory reads the code, which can be used as Figure 2 The network device shown performs part or all of the operations performed by the controller in each embodiment of the present application.

[0049] Next, the recommended method for troubleshooting operations provided in the embodiments of the present application is explained in detail.

[0050] Figure 3 is a flowchart of a fault repair operation recommendation method provided in an embodiment of the present application. The fault repair operation recommendation method can be applied to Figure 1 The controller 101 shown in FIG. Figure 3 , the method comprising:

[0051] Step 301: Obtain fault information of a faulty node.

[0052] In an embodiment of the present application, the controller may receive the fault information of the faulty node reported by the analyzer. In this case, the analyzer may monitor multiple physical nodes in real time. When a physical node is detected to have a fault, relevant information about the fault of the physical node may be collected and reported to the controller as fault information.

[0053] Optionally, in the absence of an analyzer, the controller may monitor multiple physical nodes, and when a physical node is detected to have a fault, collect relevant information about the fault of the physical node, and use the collected information as fault information. The physical node where the fault occurs is the fault node, and the fault of the fault node may be called the first fault.

[0054] In addition, the fault information may include multiple fault parameters. The multiple fault parameters may be two or more of the fault identifier, fault location, fault occurrence time, fault log, fault type and other information. Exemplarily, the fault identifier included in the fault information is an address resolution protocol (ARP) attack, which is used to indicate that the fault occurred is an ARP attack. The fault location is node 1 port 2, which is used to indicate that the fault node is node 1, and the fault occurred at port 2 of node 1. The fault occurred at 2019-07-14-08:30, and the fault type is 01. It should be noted that different fault identifiers may correspond to the same fault type, or may correspond to different fault types. That is, some faults with different names may be the same type of fault.

[0055] Optionally, in some scenarios, after a physical node fails, the physical node may continue to report fault information to the controller, so that within a certain period of time, the fault information received by the controller may contain multiple identical fault information repeatedly sent by certain faulty nodes. Alternatively, multiple physical nodes may have the same fault in the same time period. In this case, the controller will receive multiple similar fault information in the time period. Based on this, in an embodiment of the present application, the controller can also recommend repair operations according to preset time intervals. In this case, each time the controller recommends a repair operation, the controller can obtain multiple fault information received within a preset time interval closest to the current moment, and perform cluster analysis on the multiple fault information, thereby dividing the multiple fault information into multiple categories. Afterwards, the controller can recommend a repair operation for each category using the method described below.

[0056] Step 302: Search the plan library to see whether there is a repair plan corresponding to the fault information.

[0057] After acquiring the fault information, the controller can generate a first fault event according to the fault information, and the first fault event includes an event identifier, an event feature, and an event status. Among them, the event identifier is used to uniquely identify the first fault event, the event feature includes multiple fault parameters in the fault information, and the event status refers to the current state of the first fault event. With the subsequent processing progress of the fault event, the event status will be updated in real time. For example, when a recommended plan is obtained through the subsequent step 303, the event status of the first fault event can be updated to indicate that a recommended plan corresponding to the first fault event has been obtained.

[0058] After generating the first fault event, the controller can search the plan library for a second fault event similar to the first fault event based on the event features contained in the first fault event, and use the repair plan corresponding to the found second fault event as the repair plan corresponding to the fault information.

[0059] The plan library may store a variety of fault events and repair plans corresponding to each fault event. In addition, there may be one or more repair plans corresponding to each fault event. Each repair plan may include one or more repair operations, and the repair operation includes instructions for instructing the fault node to perform fault repair.

[0060] In the embodiment of the present application, the event feature includes multiple fault parameters. Based on this, the controller can calculate the similarity between the first fault event and various fault events in the plan library according to the multiple fault parameters included in the event feature of the first fault event and the fault parameters included in the event feature of each fault event in the plan library.

[0061] Exemplarily, the controller may determine the similarity between the first fault event and any fault event in the plan library by using the following algorithm (1).

[0062]

[0063] Wherein, Q is used to represent the first fault event, C is used to represent a certain fault event in the plan library, F is used to represent the event feature of the first fault event, f is used to represent the identifier of a certain fault parameter in the event feature of the first fault event, and w f Used to represent the weight corresponding to a certain fault parameter, q f is a fault parameter in the first fault event, c f is a certain fault event in the plan library and q f Fault parameters of the same type, σ f () is the similarity calculation function.

[0064] Through the above method, the controller can calculate the similarity between the first fault event and each fault event in the plan library. Afterwards, the controller can search whether there is a similarity greater than the first threshold value from the multiple similarities. If there is no similarity greater than the first threshold value among the multiple similarities calculated, the controller can determine that there is no repair plan corresponding to the fault information in the plan library, and then the controller can execute steps 303 and 304.

[0065] It should be noted that the main purpose of the fault event is to manage the received fault information. In the embodiment of the present application, the step of generating a fault event according to the fault information is an optional step. In this case, the plan library can store the mapping relationship between different fault parameter sets and repair plans. The controller can directly match the fault parameters contained in the fault information with the fault parameters in the different fault parameter sets stored in the plan library through the above algorithm (1), thereby obtaining the similarity between the fault information and each fault parameter set.

[0066] Step 303: If there is no repair plan corresponding to the fault information in the plan library, a recommended plan is determined according to the fault information through a recommendation model.

[0067] As can be seen from the above introduction, if it is found through similarity calculation that there is no fault event in the plan library whose similarity with the first fault event is greater than the first threshold, it means that there is no second fault event similar to the first fault event in the current plan library. The controller cannot obtain the repair plan corresponding to the fault information from the plan library. In this case, the controller can determine the recommended plan based on the fault information through the recommendation model.

[0068] The parameters of the recommended model may include multiple fault characteristic factors, and each fault characteristic factor may correspond to one of the multiple fault parameters included in the fault information. For example, the multiple fault parameters include the fault location, the fault occurrence time, and the fault source, and the multiple fault characteristic factors may include a topological characteristic factor corresponding to the fault location, a time characteristic factor corresponding to the fault occurrence time, and a source characteristic factor corresponding to the fault source.

[0069] The topological characteristic factor may be a networking association matrix, in which each element represents a node in the current physical network, and the value of the element corresponding to each node is determined according to the distance between each node and the faulty node. The more hops a node has from the faulty node, the farther the distance from the faulty node is, and accordingly, the smaller the value of the element corresponding to the node is.

[0070] The value of the time characteristic factor is determined by the time when the fault occurs. Since the same fault basically occurs at the same time, the time characteristic factor of the recommended model when processing fault information can be determined based on the time when the fault occurs.

[0071] The value of the source characteristic factor can be determined according to the source of the fault. The source of the fault refers to the source address of the fault. For example, when the first fault of the faulty node is an ARP attack, the source of the fault can be the attack source IP address of the ARP attack.

[0072] In addition, the parameters of the recommendation model may also include text-related feature factors, related application feature factors, etc. The value of the text-related feature factor may be determined according to the fault type, fault location, etc., and the value of the related application feature factor may be determined according to the application to which the fault location is related.

[0073] In this step, the controller can use multiple fault parameters as inputs of the recommendation model, and determine a recommended plan through the recommendation model according to multiple fault characteristics, wherein each characteristic factor included in the recommendation model can be used as the weight of the corresponding fault parameter. The recommended plan can include one or more candidate operations.

[0074] Step 304: determine a fault repair operation from one or more candidate operations included in the recommended plan, and send the fault repair operation to the faulty node.

[0075] After the recommendation model processes the fault information to obtain a recommended plan, it can further perform a configuration impact analysis on each of the one or more candidate operations included in the recommended plan, and then derive a fault repair operation from the one or more candidate operations.

[0076] The parameters of the recommendation model may also include a configuration influencing factor, which includes the original configuration information of the faulty node and the physical topology information of the network where the faulty node is located. The original configuration information of the faulty node may be the original configuration baseline of the faulty node. The physical topology information of the network where the faulty node is located may be used to indicate the topological relationship of each node in the physical network.

[0077] On this basis, the recommendation model can process each candidate operation and configuration impact factor to obtain the configuration impact degree corresponding to the corresponding candidate operation, and then use the candidate operation with the smallest configuration impact degree among one or more candidate operations as the fault repair operation. Taking any candidate operation among one or more candidate operations as an example, it is called the first candidate operation. The first candidate operation can be used as the input of the recommendation model. According to the configuration impact factor, the configuration impact degree corresponding to the first candidate operation is determined by the recommendation model. For each other candidate operation, the configuration impact degree corresponding to the corresponding candidate operation can be obtained by referring to the processing of the first candidate operation. After that, the recommendation model can output the candidate operation with the smallest configuration impact degree as the fault repair operation. Among them, the configuration impact degree is the predicted impact degree of the first candidate operation on the configuration of the faulty node and the physical topology of the network. In addition, the configuration impact degree can characterize the degree of connectivity between the faulty node and other nodes in the physical network after the first fault occurring on the faulty node is repaired by the corresponding candidate operation.

[0078] Optionally, in some possible implementations, the recommendation model can also perform a business impact analysis on each candidate operation to obtain the business impact level corresponding to each candidate operation. Afterwards, the recommendation model can determine the comprehensive impact level of each candidate operation based on the configuration impact level and business impact level corresponding to each candidate operation, and then select the candidate operation with the smallest comprehensive impact level as the fault repair operation.

[0079] Optionally, in some possible implementations, after obtaining the configuration impact degree corresponding to each candidate operation, the recommendation model may also output all candidate operations and the corresponding configuration impact degrees. In this case, the controller may search for candidate operations whose corresponding configuration impact degree is less than the second threshold value, and if found, one of the candidate operations found is used as the fault repair operation. For example, the candidate operation with the smallest corresponding configuration impact degree among the candidate operations found is used as the fault repair operation. If not found, the controller may reprocess the fault information through the recommendation model, and the parameters of the recommendation model may be adjusted during the reprocessing process.

[0080] Optionally, in some possible implementations, if the recommended plan includes only one candidate operation, the recommendation model may not perform configuration impact analysis, but directly output the candidate operation as the fault repair operation.

[0081] After obtaining the fault repair operation, the controller may send the fault repair operation to the fault node so that the fault node performs fault repair according to the fault repair operation.

[0082] Optionally, after the fault node repairs the fault according to the fault repair operation, the fault node may also feedback the repair result to the controller. The repair result includes the repaired routing information of the fault node and the physical topology information of the current network. The repaired routing information of the fault node may include the repaired forwarding table or routing table of the fault node.

[0083] After receiving the repair result fed back by the faulty node, the controller can use the repaired routing information and the physical topology information of the network included in the repair result as the input of the DPV model, and determine the real impact of the fault repair operation on the configuration of the faulty node and the physical topology of the network through the DPV model. Afterwards, the controller can generate a fault sample based on the data plane impact, the fault repair operation and the fault information, and then adjust the parameters of the recommendation model based on the fault sample. That is, the fault sample can be used as sample data to train the recommendation model.

[0084] It can be seen that the DPV model can be used to obtain the actual impact of the fault repair operation on the network where the faulty node is located, while the CPV model can be used to obtain the predicted impact of the candidate operation on the network where the faulty node is located. In this way, further adjustments to the recommendation model based on the data plane impact determined by the DPV model can improve the recommendation accuracy of the recommendation model.

[0085] Optionally, in some possible situations, the controller can not only determine the degree of impact on the data plane based on the repair results, but also analyze the impact of the fault repair operation on the business traffic in the physical network based on the real-time business flow in the network where the faulty node is located. After that, the controller can determine the comprehensive impact after the repair based on the business traffic impact and the data plane impact, and then generate a fault sample based on the comprehensive impact after the repair to adjust the parameters of the recommended model.

[0086] In an embodiment of the present application, when a repair plan corresponding to the fault information cannot be found in the plan library, the fault information can be processed through a recommendation model to obtain a recommended plan, and then an operation is selected from one or more candidate operations included in the recommended plan as a fault repair operation, thereby solving the problem that there is no repair plan corresponding to the fault information in the plan library and a repair operation cannot be provided for the faulty node.

[0087] In addition, in the embodiment of the present application, after obtaining the recommended plan, the configuration impact of each candidate operation included in the recommended plan can be estimated in advance, and then the candidate operation with the smallest configuration impact can be used as the fault repair operation. In this way, the impact of the fault node on each node in the network can be minimized when the fault is repaired through the fault repair operation.

[0088] The above embodiment mainly introduces the implementation process of how to recommend a fault repair operation to a faulty node based on the fault information of the faulty node when there is no repair plan corresponding to the fault information in the plan library. Figure 4 This paper introduces the implementation process of the controller recommending fault repair operations to the faulty node according to the fault information when there is fault information in the plan library.

[0089] Step 401: Obtain fault information of a faulty node.

[0090] The implementation method of this step can refer to step 301 in the aforementioned embodiment, and the embodiment of the present application will not be repeated here.

[0091] Step 402: Search the plan library to see whether there is a repair plan corresponding to the fault information.

[0092] In this step, the controller can determine the similarity between the first fault event and each fault event in the plan library by referring to the method described in step 302 in the above embodiment to obtain multiple similarities. If there is a similarity greater than the first threshold value among the multiple similarities, it can be determined that there is a repair plan corresponding to the fault information in the plan library, and then the controller can execute steps 403-405.

[0093] Step 403: If there is a repair plan corresponding to the fault information in the plan library, the repair plan corresponding to the fault information is used as a recommended plan.

[0094] As can be seen from the introduction in step 402, if it is found through similarity calculation that there is a fault event in the plan library whose similarity with the first fault event is greater than the first threshold, it means that there is a fault event similar to the first fault event in the current plan library. In this case, the controller can obtain the first similarity from the similarities greater than the first threshold, and use the fault event with the first similarity as the second fault event. After that, the controller can obtain the repair plan corresponding to the second fault event and use the repair plan as the recommended plan. Among them, the first similarity can be the maximum similarity among the similarities greater than the first threshold.

[0095] Optionally, in another possible implementation, the controller may take all fault events whose similarity to the first fault event is greater than a first threshold as second fault events. That is, there may be multiple second fault events. In this case, the controller may obtain multiple recommended plans from the plan library according to the second fault event.

[0096] Step 404: predicting the comprehensive impact of each of the one or more candidate operations included in the recommended plan.

[0097] After obtaining the recommended plan, the controller may estimate the comprehensive impact of each candidate operation on the network where the faulty node is located for one or more candidate operations included in the recommended plan.

[0098] Exemplarily, the controller may estimate the impact of each of the one or more candidate operations included in the recommended plan on the services in the network where the faulty node is located. Estimate the impact of each of the one or more candidate operations on the configuration of the network where the faulty node is located; and determine the comprehensive impact corresponding to each candidate operation based on the service impact and configuration impact corresponding to each candidate operation.

[0099] Take any candidate operation among one or more candidate operations as an example, and refer to it as the first candidate operation. The controller can estimate the impact on the business currently carried by the faulty node after the fault on the faulty node is repaired by the first candidate operation. The smaller the impact, the smaller the business impact. For example, assuming that the first candidate operation is to isolate a port on the faulty node, the controller can estimate whether isolating the port will affect the existing virtual machine business. If it does, the business impact can be determined as the first value, otherwise, the business impact is determined as the second value.

[0100] For each candidate operation in the one or more candidate operations, the controller may use the above method to estimate the business impact degree corresponding to each candidate operation.

[0101] In an embodiment of the present application, the controller may also estimate the degree of influence of each candidate operation on the configuration of the network where the faulty node is located. The degree of configuration influence may characterize the degree of connectivity between the faulty node and other nodes in the physical network after the first fault occurring on the faulty node is repaired using the corresponding candidate operation. Still taking any candidate operation among one or more candidate operations as an example, for the convenience of description, it is referred to as the first candidate operation.

[0102] Exemplarily, the controller may obtain the original configuration information of the faulty node and the physical topology information of the network where the faulty node is located; based on the first candidate operation and the original configuration information of the faulty node, generate the predicted configuration information corresponding to the first candidate operation, where the first candidate operation is any candidate operation among one or more candidate operations; use the predicted configuration information corresponding to the first candidate operation, the original configuration information of the faulty node, and the physical topology information as inputs of the configuration plane verification CPV model, and determine the configuration impact degree corresponding to the first candidate operation through the CPV model. The configuration impact degree may be a third value or a fourth value. When the configuration impact degree is the third value, it is used to indicate that the first candidate operation has no impact on the configuration of the faulty node, that is, the first candidate operation has no impact on the connectivity between the faulty node and other nodes. When the configuration impact degree is the fourth value, it is used to indicate that the first candidate operation has an impact on the configuration of the faulty node, that is, the first candidate operation has an impact on the connectivity between the faulty node and other nodes.

[0103] The controller may store the physical topology information of the entire physical network and the configuration baseline of each physical node in the physical network. The controller may obtain the physical topology information and the configuration baseline of the faulty node and use the configuration baseline as the original configuration information of the faulty node.

[0104] After obtaining the original configuration information of the faulty node, the controller can generate predicted configuration information corresponding to the first candidate operation based on the first candidate operation and the original configuration information. As can be seen from the above introduction, the first candidate operation may include instructions for indicating to repair the fault. Based on this, the instructions included in the first candidate operation can be added to the original configuration information to obtain the predicted configuration information, or the instructions included in the original configuration information can be modified according to the instructions included in the first candidate operation to obtain the predicted configuration information.

[0105] After obtaining the original configuration information, physical topology information, and predicted configuration information corresponding to the first candidate operation of the faulty node, the controller may process the above information through a CPV model to obtain the configuration impact degree corresponding to the first candidate operation.

[0106] For each candidate operation in the one or more candidate operations, the controller may use the above method to estimate the configuration impact degree corresponding to each candidate operation.

[0107] After determining the business impact degree and configuration impact degree corresponding to each candidate operation, the controller can obtain the business impact weight and the configuration impact weight. Then, according to the business impact degree and business impact weight corresponding to the first candidate operation, the configuration impact degree and configuration impact weight corresponding to the first candidate operation, the comprehensive impact degree corresponding to the first candidate operation is determined.

[0108] The business impact weight and the configuration impact weight may be pre-configured. Based on this, the controller may determine the product of the business impact weight and the business impact degree of the first candidate operation, determine the product of the configuration impact weight and the configuration impact degree corresponding to the first candidate operation, and use the sum of the two products as the comprehensive impact degree corresponding to the first candidate operation.

[0109] For each candidate operation, the controller may refer to the above method to determine the comprehensive impact degree corresponding to the corresponding candidate operation.

[0110] It should be noted that, in a possible situation, the controller may not perform a service impact analysis, that is, may not estimate the service impact degree. In this case, the determined configuration impact degree may be directly used as the comprehensive impact degree.

[0111] Step 405: Determine the candidate operation with the smallest comprehensive impact among the one or more candidate operations as the fault repair operation, and send the fault repair operation to the faulty node so that the faulty node performs fault repair according to the fault repair operation.

[0112] After obtaining the comprehensive impact degree corresponding to each candidate operation, the controller can determine the candidate operation with the smallest comprehensive impact degree as the fault repair operation, and send the fault repair operation to the faulty node so that the faulty node can repair the first fault that occurs according to the fault repair operation.

[0113] In an embodiment of the present application, when a repair plan corresponding to the fault information is found in the plan library, the repair plan corresponding to the fault information can be used as a recommended plan. After obtaining the recommended plan, the comprehensive impact of each candidate operation included in the recommended plan can be estimated in advance. Among them, the comprehensive impact can be obtained by comprehensively calculating the estimated business impact and configuration impact, and the candidate operation with the smallest comprehensive impact is used as the fault repair operation. In this way, the fault node can minimize the impact on the business carried by the fault node and the connected nodes when the fault is repaired through the fault repair operation.

[0114] See also Figure 5, an embodiment of the present application provides a fault repair operation recommendation device 500, the device 500 comprising:

[0115] Acquisition module 501, used to execute step 301 in the above embodiment;

[0116] A first determining module 502, configured to execute step 303 in the above embodiment;

[0117] The recommendation module 503 is used to execute step 304 in the above embodiment.

[0118] Optionally, the fault information includes a plurality of fault parameters, the parameters of the recommended model include a plurality of fault characteristic factors, and each of the plurality of fault characteristic factors corresponds to a fault parameter of the plurality of fault parameters;

[0119] The first determining module 502 is specifically used for:

[0120] A plurality of fault parameters are used as inputs of the recommendation model, and a recommended plan is determined through the recommendation model according to a plurality of fault characteristic factors.

[0121] Optionally, the parameters of the recommendation model further include a configuration influencing factor, and the configuration influencing factor includes original configuration information of the faulty node and physical topology information of the network where the faulty node is located;

[0122] The recommendation module 503 is specifically used for:

[0123] The first candidate operation is used as an input of the recommendation model, and the configuration impact degree corresponding to the first candidate operation is determined by the recommendation model according to the configuration impact factor, where the configuration impact degree is the predicted impact degree of the first candidate operation on the configuration of the faulty node and the physical topology of the network where the first candidate operation is located, and the first candidate operation is any candidate operation among the one or more candidate operations;

[0124] A candidate operation with the least configuration impact among the one or more candidate operations is used as a fault repair operation.

[0125] Alternatively, see Figure 6 , the device 500 further includes:

[0126] The receiving module 504 is used to receive the repair result fed back by the fault node after the fault is repaired according to the fault repair operation, and the repair result includes the routing information of the fault node after the fault is repaired and the physical topology information of the network where the fault node is located;

[0127] The second determination module 505 is used to use the routing information and physical topology information after the fault is repaired as inputs of the data plane verification DPV model, and determine the data plane impact degree corresponding to the fault repair operation through the DPV model. The data plane impact degree is the actual impact degree of the fault repair operation on the configuration of the faulty node and the physical topology of the network in which it is located;

[0128] A generating module 506, configured to generate a fault sample according to the data plane impact degree corresponding to the fault repair operation, the fault repair operation and the fault information;

[0129] The adjustment module 507 is used to adjust the parameters of the recommendation model according to the fault samples.

[0130] To summarize, in an embodiment of the present application, when a repair plan corresponding to the fault information cannot be found in the plan library, the fault information can be processed through a recommendation model to obtain a recommended plan, and then an operation is selected from one or more candidate operations included in the recommended plan as a fault repair operation, thereby solving the problem that there is no repair plan corresponding to the fault information in the plan library, and a repair operation cannot be provided for the faulty node.

[0131] See also Figure 7 , the embodiment of the present application provides a fault repair operation recommendation device 700, the device 700 includes:

[0132] Acquisition module 701, used to execute step 401 in the above embodiment;

[0133] Determining module 702, used to execute step 403 in the above embodiment;

[0134] Prediction module 703, used to execute step 404 in the above embodiment;

[0135] The recommendation module 704 is used to execute step 405 in the above embodiment.

[0136] Optionally, the prediction module 703 is specifically used for:

[0137] estimating a degree of service impact of each of the one or more candidate operations on services in a network where the failed node is located;

[0138] estimating the degree of impact of each of the one or more candidate operations on the configuration of the network where the failed node is located;

[0139] According to the business impact degree and configuration impact degree corresponding to each candidate operation, the comprehensive impact degree corresponding to each candidate operation is determined.

[0140] Optionally, the prediction module 703 is specifically used for:

[0141] Obtain the original configuration information of the faulty node and the physical topology information of the network where the faulty node is located;

[0142] Generate predicted configuration information corresponding to the first candidate operation according to the first candidate operation and the original configuration information of the faulty node, where the first candidate operation is any candidate operation among the one or more candidate operations;

[0143] The predicted configuration information corresponding to the first candidate operation, the original configuration information of the faulty node, and the physical topology information are used as inputs of the configuration plane verification CPV model, and the configuration impact degree corresponding to the first candidate operation is determined through the CPV model.

[0144] Optionally, the recommendation module 704 is specifically used for:

[0145] Obtain business impact weights and configuration impact weights;

[0146] The comprehensive impact degree corresponding to the first candidate operation is determined according to the business impact degree and the business impact weight corresponding to the first candidate operation, and the configuration impact degree and the configuration impact weight corresponding to the first candidate operation.

[0147] In summary, in an embodiment of the present application, when a repair plan corresponding to fault information is found in the plan library, the repair plan corresponding to the fault information can be used as a recommended plan. After obtaining the recommended plan, the comprehensive impact of each candidate operation included in the recommended plan can be estimated in advance. Among them, the comprehensive impact can be obtained by comprehensively calculating the estimated business impact and configuration impact, and the candidate operation with the smallest comprehensive impact is used as the fault repair operation. In this way, the impact on the business carried by the faulty node and the connected nodes can be minimized when the faulty node is repaired through the fault repair operation.

[0148] It should be noted that: the fault repair operation recommendation device provided in the above embodiment only uses the division of the above functional modules as an example when recommending fault repair operations. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the fault repair operation recommendation device provided in the above embodiment and the fault repair operation recommendation method embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0149] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (for example: coaxial cable, optical fiber, data subscriber line (Digital Subscriber Line, DSL)) or wireless (for example: infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integration. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0150] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0151] The above-mentioned embodiments are provided for the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A fault repair operation recommendation method, characterized in that: The method comprises: Get the fault information of the faulty node; If there is no repair plan corresponding to the fault information in the plan library, determine a recommended plan based on the fault information through a recommendation model, wherein the plan library is used to store multiple fault events and repair plans corresponding to each of the multiple fault events, and the recommended plan includes one or more candidate operations; Predicting a configuration impact degree of each candidate operation among the one or more candidate operations, wherein the configuration impact degree is a predicted impact degree of the corresponding candidate operation on the configuration of the faulty node and the physical topology of the network in which the node is located; A fault recovery operation is determined from the one or more candidate operations based on the configuration impact degree, and the fault recovery operation is recommended to the fault node so that the fault node performs fault recovery according to the fault recovery operation.

2. The method according to claim 1, characterized in that The fault information includes a plurality of fault parameters, the parameters of the recommended model include a plurality of fault characteristic factors, each of the plurality of fault characteristic factors corresponds to one of the plurality of fault parameters; Determining a recommended plan through a recommendation model according to the fault information includes: The multiple fault parameters are used as inputs of the recommendation model, and the recommended plan is determined through the recommendation model according to the multiple fault characteristic factors.

3. The method according to claim 2, characterized in that The parameters of the recommendation model also include a configuration influencing factor, and the configuration influencing factor includes the original configuration information of the faulty node and the physical topology information of the network where the faulty node is located; The determining of the fault recovery operation from the one or more candidate operations based on the configuration impact degree includes: Taking a first candidate operation as an input of the recommendation model, and determining a configuration influence degree corresponding to the first candidate operation through the recommendation model according to the configuration influence factor, wherein the first candidate operation is any candidate operation among the one or more candidate operations; The candidate operation with the least corresponding configuration impact among the one or more candidate operations is used as the fault repair operation.

4. The method according to any one of claims 1 to 3, characterized in that: After recommending the fault repair operation to the fault node, the method further includes: Receiving a repair result fed back by the faulty node after the fault is repaired according to the fault repair operation, the repair result including routing information of the faulty node after the fault is repaired and physical topology information of the network in which the faulty node is located; The routing information and the physical topology information after the fault repair are used as inputs of a data plane verification DPV model, and the data plane impact degree corresponding to the fault repair operation is determined through the DPV model, where the data plane impact degree is the actual impact degree of the fault repair operation on the configuration of the faulty node and the physical topology of the network in which it is located; generating a fault sample according to the data plane impact degree corresponding to the fault repair operation, the fault repair operation and the fault information; The parameters of the recommendation model are adjusted according to the fault sample.

5. A fault repair operation recommendation method, characterized in that: The method comprises: Get the fault information of the faulty node; If a repair plan corresponding to the fault information exists in the plan library, the repair plan corresponding to the fault information is used as a recommended plan, where the recommended plan includes one or more candidate operations; Predicting a comprehensive impact degree of each candidate operation among the one or more candidate operations, wherein the comprehensive impact degree is used to indicate the magnitude of the comprehensive impact of the corresponding candidate operation on the network where the faulty node is located; A candidate operation with the smallest comprehensive impact among the one or more candidate operations is determined as a fault repair operation, and the fault repair operation is recommended to the fault node so that the fault node performs fault repair according to the fault repair operation.

6. The method according to claim 5, characterized in that The predicting of the comprehensive impact of each of the one or more candidate operations included in the recommended plan includes: estimating the degree of service impact of each of the one or more candidate operations on services in the network where the faulty node is located; estimating the degree of influence of each of the one or more candidate operations on the configuration of the network where the faulty node is located; According to the business impact degree and configuration impact degree corresponding to each candidate operation, the comprehensive impact degree corresponding to each candidate operation is determined.

7. The method according to claim 6, characterized in that The estimating the degree of influence of each of the one or more candidate operations on the configuration of the network where the faulty node is located includes: Obtaining original configuration information of the faulty node and physical topology information of the network where the faulty node is located; Generate predicted configuration information corresponding to the first candidate operation according to the first candidate operation and the original configuration information of the faulty node, where the first candidate operation is any candidate operation among the one or more candidate operations; The predicted configuration information corresponding to the first candidate operation, the original configuration information of the faulty node, and the physical topology information are used as inputs of a configuration plane verification CPV model, and the configuration impact degree corresponding to the first candidate operation is determined through the CPV model.

8. The method according to claim 6 or 7, characterized in that: The determining, according to the business impact degree and configuration impact degree corresponding to each candidate operation, the comprehensive impact degree corresponding to each candidate operation includes: Obtain business impact weights and configuration impact weights; The comprehensive impact degree corresponding to the first candidate operation is determined according to the business impact degree corresponding to the first candidate operation, the business impact weight, the configuration impact degree corresponding to the first candidate operation, and the configuration impact weight.

9. A fault repair operation recommendation device, characterized in that: The device comprises: An acquisition module, used to acquire fault information of a faulty node; A first determination module is configured to determine a recommended plan through a recommendation model according to the fault information if a repair plan corresponding to the fault information does not exist in a plan library, wherein the plan library is used to store multiple fault events and repair plans corresponding to each of the multiple fault events, and the recommended plan includes one or more candidate operations; A recommendation module is used to predict the configuration impact degree of each candidate operation in the one or more candidate operations, where the configuration impact degree is the predicted impact degree of the corresponding candidate operation on the configuration of the faulty node and the physical topology of the network in which it is located; determine a fault repair operation from the one or more candidate operations based on the configuration impact degree, and recommend the fault repair operation to the faulty node so that the faulty node performs fault repair according to the fault repair operation.

10. The device according to claim 9, characterized in that The fault information includes a plurality of fault parameters, the parameters of the recommended model include a plurality of fault characteristic factors, each of the plurality of fault characteristic factors corresponds to one of the plurality of fault parameters; The first determining module is specifically used for: The multiple fault parameters are used as inputs of the recommendation model, and the recommended plan is determined through the recommendation model according to the multiple fault characteristic factors.

11. The device according to claim 10, characterized in that The parameters of the recommendation model also include a configuration influencing factor, and the configuration influencing factor includes the original configuration information of the faulty node and the physical topology information of the network where the faulty node is located; The recommendation module is specifically used for: Taking a first candidate operation as an input of the recommendation model, and determining a configuration influence degree corresponding to the first candidate operation through the recommendation model according to the configuration influence factor, wherein the first candidate operation is any candidate operation among the one or more candidate operations; The candidate operation with the least corresponding configuration impact among the one or more candidate operations is used as the fault repair operation.

12. The device according to any one of claims 9 to 11, characterized in that: The device also includes: A receiving module, used to receive a repair result fed back by the fault node after the fault is repaired according to the fault repair operation, wherein the repair result includes routing information of the fault node after the fault is repaired and physical topology information of the network in which the fault node is located; A second determination module is configured to use the routing information and the physical topology information after the fault is repaired as inputs of a data plane verification DPV model, and determine the data plane impact degree corresponding to the fault repair operation through the DPV model, wherein the data plane impact degree is the actual impact degree caused by the fault repair operation on the configuration of the faulty node and the physical topology of the network in which it is located; A generating module, configured to generate a fault sample according to the data plane impact degree corresponding to the fault repair operation, the fault repair operation and the fault information; An adjustment module is used to adjust the parameters of the recommendation model according to the fault sample.

13. A fault repair operation recommendation device, characterized in that: The device comprises: An acquisition module, used to acquire fault information of a faulty node; a determination module, configured to use the repair plan corresponding to the fault information as a recommended plan if a repair plan corresponding to the fault information exists in a plan library, wherein the recommended plan includes one or more candidate operations; A prediction module, used to predict the comprehensive impact degree of each candidate operation in the one or more candidate operations, wherein the comprehensive impact degree is used to indicate the magnitude of the comprehensive impact of the corresponding candidate operation on the network where the faulty node is located; The recommendation module is used to determine the candidate operation with the smallest comprehensive impact among the one or more candidate operations as the fault repair operation, and recommend the fault repair operation to the fault node so that the fault node performs fault repair according to the fault repair operation.

14. The device according to claim 13, characterized in that The prediction module is specifically used for: estimating the degree of service impact of each of the one or more candidate operations on services in the network where the faulty node is located; estimating the degree of influence of each of the one or more candidate operations on the configuration of the network where the faulty node is located; According to the business impact degree and configuration impact degree corresponding to each candidate operation, the comprehensive impact degree corresponding to each candidate operation is determined.

15. The device according to claim 14, characterized in that The prediction module is specifically used for: Obtaining original configuration information of the faulty node and physical topology information of the network where the faulty node is located; Generate predicted configuration information corresponding to the first candidate operation according to the first candidate operation and the original configuration information of the faulty node, where the first candidate operation is any candidate operation among the one or more candidate operations; The predicted configuration information corresponding to the first candidate operation, the original configuration information of the faulty node, and the physical topology information are used as inputs of a configuration plane verification CPV model, and the configuration impact degree corresponding to the first candidate operation is determined through the CPV model.

16. The device according to claim 14 or 15, characterized in that The recommendation module is specifically used for: Obtain business impact weights and configuration impact weights; The comprehensive impact degree corresponding to the first candidate operation is determined according to the business impact degree corresponding to the first candidate operation, the business impact weight, the configuration impact degree corresponding to the first candidate operation, and the configuration impact weight.

17. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a computer, the computer executes the fault repair operation recommendation method according to any one of claims 1 to 8.

Images