An online system troubleshooting method, device and electronic equipment

Abstract

The application provides an online system troubleshooting method, device and electronic equipment, and the method comprises the following steps: when the online system fails, a troubleshooting system acquires a system log of the online system, and extracts failure information of the online system recorded in the system log; the failure information is input into a trained convolutional neural network model to determine whether the failure corresponding to the failure information is a failure that needs to be investigated; when it is determined that the failure information is a failure that needs to be investigated, the failure information is input into a trained fasttext model to determine the failure cause of the failure corresponding to the failure information. Through the online system troubleshooting method, device and electronic equipment provided in the embodiments of the application, the troubleshooting of the online system can be automatically completed, and the execution efficiency of the troubleshooting of the online system is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically, to an online system troubleshooting method, apparatus, and electronic device. Background Technology

[0002] Currently, when problems occur in online systems, the common solution is to restore online services first, and then conduct further investigations. However, the investigation process requires manual operation, which makes the investigation progress very slow. Summary of the Invention

[0003] To address the aforementioned problems, the present invention aims to provide an online system troubleshooting method, apparatus, and electronic device.

[0004] In a first aspect, embodiments of the present invention provide an online system troubleshooting method, including:

[0005] When an online system malfunctions, the troubleshooting system retrieves the system logs of the online system and extracts the fault information of the online system recorded in the system logs.

[0006] The fault information is input into the trained convolutional neural network model to determine whether the fault information corresponds to a fault that needs to be investigated.

[0007] When the fault information is determined to be a fault that needs to be investigated, the fault information is input into the trained fasttext model to determine the cause of the fault corresponding to the fault information.

[0008] Secondly, embodiments of the present invention also provide an online system troubleshooting device, comprising:

[0009] The acquisition module is used to acquire the system logs of the online system when a failure occurs, and extract the failure information of the online system recorded in the system logs.

[0010] The first processing module is used to input the fault information into the trained convolutional neural network model and determine whether the fault information corresponds to a fault that needs to be investigated.

[0011] The second processing module is used to input the fault information into the trained fasttext model when it is determined that the fault information is a fault that needs to be investigated, and to determine the cause of the fault corresponding to the fault information.

[0012] Thirdly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, performs the steps of the method described in the first aspect above.

[0013] Fourthly, embodiments of the present invention also provide an electronic device, the electronic device including a memory, a processor and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor using the steps of the method described in the first aspect above.

[0014] In the solutions provided by the first to fourth aspects of the present invention, when an online system malfunctions, the troubleshooting system obtains the system logs of the online system, extracts the fault information recorded in the system logs, and processes the fault information using a trained convolutional neural network model. If the fault information is determined to be a fault requiring investigation, it is input into a trained fasttext model to determine the fault type corresponding to the fault information. This completes the troubleshooting work for the online system. Compared to the manual troubleshooting methods in related technologies, the troubleshooting system can automatically complete the troubleshooting work without manual intervention, greatly improving the efficiency of online system fault troubleshooting.

[0015] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart of an online system troubleshooting method provided in Embodiment 1 of the present invention is shown;

[0018] Figure 2 This diagram illustrates the structure of an online system troubleshooting device provided in Embodiment 2 of the present invention.

[0019] Figure 3 A schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention is shown. Detailed Implementation

[0020] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0021] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0022] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0023] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] Example 1

[0025] This embodiment proposes an online system troubleshooting method, in which the troubleshooting system is the executing entity.

[0026] The troubleshooting system can run on a server. Furthermore, it can run on the same server as the online system or on a different server; and it can interact with the online system.

[0027] See Figure 1 The flowchart shown illustrates a troubleshooting method for an online system. This embodiment proposes a troubleshooting method for an online system, including the following specific steps:

[0028] Step 100: When a failure occurs in the online system, the troubleshooting system obtains the system log of the online system and extracts the failure information of the online system recorded in the system log.

[0029] In step 100 above, the troubleshooting system monitors the operation of the online system in real time. When a failure occurs in the online system, the troubleshooting system will obtain the system log of the online system and extract the failure information of the online system recorded in the system log.

[0030] Furthermore, the troubleshooting system is networked with other troubleshooting systems to form a blockchain system that stores fault information that identifies faults requiring investigation. The troubleshooting system registers on this blockchain system, becoming a regular node within it.

[0031] When an online system malfunctions, some faults can be resolved automatically without troubleshooting, such as software update failures or high data load. However, other faults require troubleshooting, such as system crashes or inability to communicate with external devices over the network.

[0032] All of these fault information will be recorded in the system log of the online system. When the troubleshooting system detects a fault in the online system, it needs to identify the fault that needs to be investigated.

[0033] Step 102: Input the fault information into the trained convolutional neural network model to determine whether the fault information corresponds to a fault that needs to be investigated.

[0034] In step 102 above, the trained convolutional neural network model is pre-trained using fault information from faults that do not require troubleshooting and fault information from faults that require troubleshooting. The specific training process is existing technology and will not be described in detail here.

[0035] To determine whether the fault information corresponds to a fault that needs to be investigated, step 102 above can be performed by executing the following steps (1) to (3):

[0036] (1) Perform word segmentation on the fault information to obtain the word segmentation of the fault information;

[0037] (2) The word2vec model is used to process the word segmentation of the fault information to obtain the word segmentation vector of the fault information;

[0038] (3) Input the word segmentation vector of the fault information into the trained convolutional neural network model to detect whether the fault information is a fault that needs to be investigated.

[0039] In step (3) above, the specific process by which the trained convolutional neural network model detects whether the fault information is a fault that needs to be investigated is existing technology and will not be described in detail here.

[0040] The convolutional neural network model includes: D convolutional layers, each followed by an activation function and a pooling layer. The result of convolution, activation, and pooling of the word segmentation vector of the fault information is called a feature map, denoted by Z; M fully connected layers, where the kernel size of the convolutional layers is Kn, the stride is Sn, and the number of kernels is Xn, where 1 ≤ n ≤ D; the last fully connected layer has 2 neurons, and the other fully connected layers have Cm neurons, where 1 ≤ m ≤ M; and the activation function after each convolutional layer is a Max-Feature-Map, as shown below:

[0041]

[0042] Where 1≤k≤n; Z∈R h×w×2n The number of Z is 2n;

[0043] Then the derivative of B is:

[0044]

[0045] Where w is the width of the feature map; h is the height of the feature map; i is the i-th column of the feature map, 0 ≤ i < w; j is the j-th row of the feature map, 0 ≤ j < h; 1 ≤ k ≤ n, and R is the real number space. This represents the feature value in the i-th column and j-th row of the k-th feature map. Z represents the feature value in the i-th column and j-th row of the (k+n)-th feature map; k This represents the k-th feature map.

[0046] When the fault information is determined to be a fault that needs to be investigated, continue to the next step 104.

[0047] Step 104: When it is determined that the fault information is a fault that needs to be investigated, the fault information is input into the trained fasttext model to determine the cause of the fault corresponding to the fault information.

[0048] In step 104 above, the trained FastText model is obtained by inputting the fault name, fault phenomenon, and known fault causes into the FastText model. The specific training process is existing technology and will not be described in detail here.

[0049] The known causes of the fault are those summarized by humans. The fault phenomena are the system operation information of the online system at the time of the fault with the given fault name.

[0050] In order to determine the cause of the fault corresponding to the fault information, step 104 above can perform the following steps (1) to (3):

[0051] (1) Process the word segmentation of the fault information to generate n-gram features, character repetition features and character category features of the word segmentation of the fault information, so as to serve as the vector representation of the word segmentation of the fault information;

[0052] (2) Obtain the system operation information of the online system, process the system operation information, and obtain the vector representation of the system operation information;

[0053] (3) Input the vector representation of the fault information and the vector representation of the system operation information into the fasttext model to determine the fault cause corresponding to the fault information.

[0054] In step (2) above, the vector representation of the system operation information also includes: n-gram features, character repetition feature information features, and character category features.

[0055] The specific process of obtaining the vector representation of the system operation information is similar to the process of obtaining the vector representation of the word segmentation of the fault information described above, and will not be repeated here.

[0056] In step (3) above, the specific process of inputting the vector representation of the fault information and the vector representation of the system operation information into the trained fasttext model to determine the fault cause corresponding to the fault information is existing technology and will not be described in detail here.

[0057] Since the fault to be investigated is a serious fault in the online system, it is necessary to avoid the fault information being tampered with or deleted as much as possible. Therefore, when the cause of the fault corresponding to the fault information is determined, the online system troubleshooting method proposed in this embodiment may also include the following steps (1) to (3):

[0058] (1) When the cause of the fault corresponding to the fault information is determined, a fault information identifier is assigned to the fault information;

[0059] (2) Store the fault information in the blockchain system where the troubleshooting system is located, and obtain the block address of the fault information stored by the blockchain system.

[0060] (3) Perform a hash calculation on the block address storing the fault information to obtain the block address hash value, generate a correspondence between the fault information identifier and the block address hash value, and send the block address hash value to the blockchain system, so that when the blockchain system receives the block address hash value, it generates a correspondence between the block address storing the fault information and the received block address hash value, and stores the generated correspondence between the block address storing the fault information and the received block address hash value.

[0061] In step (1) above, the fault information identifier assigned to the fault information is generated by the troubleshooting system and is set in the fault information.

[0062] In step (2) above, after storing the fault information, the blockchain system can obtain the block address where the fault information is stored, and then send the obtained block address to the troubleshooting system.

[0063] As can be seen from the process described in steps (1) to (3) above, the fault information of the online system that has been investigated can be stored in the blockchain system, and the block address storing the fault information of the investigated fault can be hashed to obtain the block address hash value. The calculated block address hash value is then stored in the blockchain system. Thus, the fault information of the investigated fault can be stored through the blockchain system, and the fault information of the investigated fault and the block address storing the fault information of the investigated fault can be managed through the blockchain system. By utilizing the traceability and immutability of the blockchain system, it is ensured that the fault information will not be tampered with or deleted, thereby ensuring the accuracy of the fault information of the online system that has been investigated.

[0064] When it is necessary to query the fault information of the faults that have been investigated in the online system, the online system troubleshooting method proposed in this embodiment further includes the following steps (10) to (14):

[0065] (10) Obtain fault query information, wherein the fault query information includes: the user identifier of the user who issued the fault query information, the query time, and the fault information identifier of the fault to be queried;

[0066] (11) When the fault information identifier in the fault query information can be used to query the block address hash value corresponding to the fault information identifier, a fault information query instruction is generated using the queried block address hash value corresponding to the fault information identifier.

[0067] (12) Send the generated fault information query instruction to the blockchain system, so that the blockchain system can query the block address corresponding to the block address hash value carried in the received fault information query instruction according to the block address hash value carried in the received fault information query instruction, read the fault information from the queried block address, and feed back the read fault information to the troubleshooting system.

[0068] (13) Receive the fault information fed back by the blockchain system and feed back the received fault information to the user who issued the fault query information;

[0069] (14) The fault query information is sent to the blockchain system, so that the blockchain system stores the fault query information in the query log, wherein the query log is set in the blockchain system.

[0070] As can be seen from the descriptions of steps (10) to (14) above, after the query is completed, the troubleshooting system sends the fault query information to the blockchain system, so that the blockchain system stores the fault query information in the query log, thereby recording the fault query information and facilitating data access security management; when fault information is leaked, tampered with, or lost, an investigation can be conducted based on the fault query information.

[0071] In summary, this embodiment proposes an online system troubleshooting method. When an online system malfunctions, the troubleshooting system obtains the system logs and extracts the fault information recorded in the logs. It then processes the fault information using a trained convolutional neural network model. If the fault information is determined to be a fault requiring investigation, it is input into a trained fasttext model to determine the fault type. This completes the online system fault investigation. Compared to the manual troubleshooting methods in related technologies, this method allows the troubleshooting system to automatically complete the online system fault investigation without manual intervention, significantly improving the efficiency of online system fault investigation.

[0072] Example 2

[0073] This embodiment proposes an online system troubleshooting device for executing the online system troubleshooting method proposed in Embodiment 1 above.

[0074] See Figure 2 The online system troubleshooting device shown includes:

[0075] The acquisition module 200 is used to acquire the system log of the online system and extract the fault information of the online system recorded in the system log when the online system fails.

[0076] The first processing module 202 is used to input the fault information into the trained convolutional neural network model and determine whether the fault information corresponds to a fault that needs to be investigated.

[0077] The second processing module 204 is used to input the fault information into the trained fasttext model when it is determined that the fault information is a fault that needs to be investigated, and to determine the fault cause corresponding to the fault information.

[0078] The first processing module 202 is specifically used for:

[0079] The fault information is segmented into words to obtain the segmented fault information.

[0080] The fault information is processed using the word2vec model to obtain the word segmentation vector of the fault information;

[0081] The word segmentation vector of the fault information is input into a trained convolutional neural network model to detect whether the fault information is a fault that needs to be investigated. The convolutional neural network model includes: D convolutional layers, each followed by an activation function and a pooling layer. The result of convolution, activation, and pooling of the fault information word segmentation vector is called a feature map, denoted by Z; M fully connected layers, where the kernel size of each convolutional layer is Kn, the stride is Sn, and the number of kernels is Xn, where 1 ≤ n ≤ D; the last fully connected layer has 2 neurons, and the other fully connected layers have Cm neurons, where 1 ≤ m ≤ M; and the activation function after each convolutional layer is a Max-Feature-Map, as shown below:

[0082]

[0083] Where 1≤k≤n; Z∈R h×w×2n The number of Z is 2n;

[0084] Then the derivative of B is:

[0085]

[0086] Where w is the width of the feature map; h is the height of the feature map; i is the i-th column of the feature map, 0 ≤ i < w; j is the j-th row of the feature map, 0 ≤ j < h; 1 ≤ k ≤ n, and R is the real number space. This represents the feature value in the i-th column and j-th row of the k-th feature map. Z represents the feature value in the i-th column and j-th row of the (k+n)-th feature map; k This represents the k-th feature map.

[0087] Specifically, the second processing module 204 is used to input the fault information into the trained fasttext model to determine the fault cause corresponding to the fault information, including:

[0088] The fault information is segmented into words to generate n-gram features, character repetition features, and character category features of the fault information segments, which are used as vector representations of the fault information segments.

[0089] Obtain the system operation information of the online system, process the system operation information, and obtain a vector representation of the system operation information;

[0090] The vector representations of the fault information and the vector representations of the system operation information are input into the fasttext model to determine the cause of the fault corresponding to the fault information.

[0091] In summary, this embodiment proposes an online system troubleshooting device. When an online system malfunctions, the troubleshooting system acquires the system logs and extracts the fault information recorded in the logs. It then processes the fault information using a trained convolutional neural network model. If the fault information is determined to be a fault requiring investigation, it is input into a trained fasttext model to determine the fault type. This completes the online system fault investigation. Compared to the manual troubleshooting methods in related technologies, this device can automatically complete the online system fault investigation without manual intervention, significantly improving the efficiency of online system fault investigation.

[0092] Example 3

[0093] This embodiment proposes a computer-readable storage medium storing a computer program. When the computer program is run by a processor, it executes the steps of the online system troubleshooting method described in Embodiment 1 above. For specific implementation details, please refer to Method Embodiment 1, which will not be repeated here.

[0094] In addition, see Figure 3 The diagram shows the structure of an electronic device. This embodiment also proposes an electronic device, which includes a bus 51, a processor 52, a transceiver 53, a bus interface 54, a memory 55, and a user interface 56. The electronic device includes a memory 55.

[0095] In this embodiment, the electronic device further includes: one or more programs stored in the memory 55 and executable on the processor 52, configured to be executed by the processor to perform the one or more programs for the following steps (1) to (3):

[0096] (1) When a failure occurs in the online system, the troubleshooting system obtains the system log of the online system and extracts the failure information of the online system recorded in the system log.

[0097] (2) Input the fault information into the trained convolutional neural network model to determine whether the fault information corresponds to a fault that needs to be investigated;

[0098] (3) When the fault information is determined to be a fault that needs to be investigated, the fault information is input into the trained fasttext model to determine the fault cause corresponding to the fault information.

[0099] Transceiver 53 is used to receive and send data under the control of processor 52.

[0100] The bus architecture (represented by bus 51) can include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 52 and memory represented by memory 55. Bus 51 can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be further described in this embodiment. Bus interface 54 provides an interface between bus 51 and transceiver 53. Transceiver 53 can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. For example, transceiver 53 receives external data from other devices. Transceiver 53 is used to transmit data processed by processor 52 to other devices. Depending on the nature of the computing system, a user interface 56 may also be provided, such as a keypad, display, speaker, microphone, or joystick.

[0101] Processor 52 is responsible for managing bus 51 and general processing, such as running a general-purpose operating system as described above. Memory 55 can be used to store data used by processor 52 during operation.

[0102] Optionally, the processor 52 may be, but is not limited to, a central processing unit, a microcontroller, a microprocessor, or a programmable logic device.

[0103] It is understood that the memory 55 in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM). The memory 55 of the systems and methods described in this embodiment is intended to include, but is not limited to, these and any other suitable types of memory.

[0104] In some implementations, memory 55 stores elements such as executable modules or data structures, or subsets thereof, or extended sets thereof: operating system 551 and application programs 552.

[0105] The operating system 551 includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic business functions and handle hardware-based tasks. The application program 552 includes various applications, such as a media player and a browser, used to implement various application functions. The program implementing the method of this embodiment can be included in the application program 552.

[0106] In summary, this embodiment proposes an online system troubleshooting method, apparatus, and electronic device. When an online system malfunctions, the troubleshooting system acquires the system logs and extracts the fault information recorded in the logs. It then processes the fault information using a trained convolutional neural network model. If the fault information is determined to be a fault requiring investigation, it is input into a trained fasttext model to determine the fault type. This completes the online system fault investigation. Compared to the manual troubleshooting methods in related technologies, the troubleshooting system can automatically complete the online system fault investigation without manual intervention, significantly improving the efficiency of online system fault investigation.

[0107] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for troubleshooting an online system, characterized in that, The method, applied to a troubleshooting system where the troubleshooting system and the online system run on the same server or different servers and interact with the online system, includes: The troubleshooting system monitors the operation of the online system in real time. When a failure occurs in the online system, the troubleshooting system obtains the system log of the online system and extracts the failure information recorded in the system log. The troubleshooting system is connected to other troubleshooting systems to form a blockchain system that stores failure information that determines the failures that need to be investigated. The failure information includes failure information of failures that do not require troubleshooting or failure information of failures that require troubleshooting. The fault information is input into the trained convolutional neural network model to determine whether the fault information corresponds to a fault that needs to be investigated. When the fault information is determined to be a fault requiring investigation, the fault information is input into a trained fasttext model to determine the cause of the fault corresponding to the fault information; the step of inputting the fault information into the trained fasttext model to determine the cause of the fault corresponding to the fault information includes: The fault information is segmented into words to generate n-gram features, character repetition features, and character category features of the fault information segments, which are used as vector representations of the fault information segments. Obtain the system operation information of the online system, process the system operation information, and obtain a vector representation of the system operation information; The vector representation of the fault information and the vector representation of the system operation information are input into the fasttext model to determine the fault cause corresponding to the fault information; wherein, the trained fasttext model is trained by inputting the fault name, fault phenomenon and known fault cause into the fasttext model.

2. The method according to claim 1, characterized in that, The fault information is input into a trained convolutional neural network model to determine whether the fault corresponding to the fault information is a fault that needs to be investigated, including: The fault information is segmented into words to obtain the segmented fault information. The fault information is processed using the word2vec model to obtain the word segmentation vector of the fault information; The word segmentation vector of the fault information is input into a trained convolutional neural network model to detect whether the fault information is a fault that needs to be investigated. The convolutional neural network model includes: D convolutional layers, each followed by an activation function and a pooling layer. The result of convolution, activation, and pooling of the fault information word segmentation vector is called a feature map, denoted by Z; M fully connected layers, where the kernel size of each convolutional layer is Kn, the stride is Sn, and the number of kernels is Xn, where 1 ≤ n ≤ D; the last fully connected layer has 2 neurons, and the other fully connected layers have Cm neurons, where 1 ≤ m ≤ M; and the activation function after each convolutional layer is a Max-Feature-Map, as shown below: Where 1≤k≤n; The number of Z is 2n; Then the derivative of B is: Where w is the width of the feature map; h is the height of the feature map; i is the i-th column of the feature map, 0 ≤ i < w; j is the j-th row of the feature map, 0 ≤ j < h; 1 ≤ k ≤ n, and R is the real number space. This represents the feature value in the i-th column and j-th row of the k-th feature map. This represents the feature value in the i-th column and j-th row of the k+n-th feature map; This represents the k-th feature map.

3. The method according to claim 1, characterized in that, Also includes: When the cause of the fault corresponding to the fault information is determined, a fault information identifier is assigned to the fault information; The fault information is stored in the blockchain system where the troubleshooting system is located, and the block address where the fault information is stored is obtained from the blockchain system. A hash calculation is performed on the block address storing the fault information to obtain a block address hash value. A correspondence between the fault information identifier and the block address hash value is generated, and the block address hash value is sent to the blockchain system. When the blockchain system receives the block address hash value, it generates a correspondence between the block address storing the fault information and the received block address hash value, and stores the generated correspondence between the block address storing the fault information and the received block address hash value.

4. The method according to claim 3, characterized in that, Also includes: Obtain fault query information, which includes: the user identifier of the user who issued the fault query information, the query time, and the fault information identifier of the fault to be queried; When the fault information identifier in the fault query information can be used to query the block address hash value corresponding to the fault information identifier, a fault information query instruction is generated using the queried block address hash value corresponding to the fault information identifier. The generated fault information query command is sent to the blockchain system, which then queries the block address corresponding to the block address hash value carried in the received fault information query command, reads the fault information from the queried block address, and feeds the read fault information back to the troubleshooting system. Receive fault information from the blockchain system and send the received fault information back to the user who issued the fault query. The fault query information is sent to the blockchain system, so that the blockchain system stores the fault query information in the query log, wherein the query log is set in the blockchain system.

5. An online system troubleshooting device, characterized in that, An apparatus for use in troubleshooting systems, where the troubleshooting system and the online system run on the same server or different servers, and the apparatus interacts with the online system, comprising: The monitoring module is used to monitor the operation of the online system in real time; The acquisition module is used to acquire the system log of the online system when the online system fails, and extract the fault information of the online system recorded in the system log; wherein, the troubleshooting system is connected to other troubleshooting systems to form a blockchain system for storing fault information that determines faults that need to be investigated; the fault information includes fault information of faults that do not need to be investigated or fault information of faults that need to be investigated. The first processing module is used to input the fault information into the trained convolutional neural network model and determine whether the fault information corresponds to a fault that needs to be investigated. The second processing module is used to input the fault information into the trained fasttext model when it is determined that the fault information is a fault that needs to be investigated, and to determine the fault cause corresponding to the fault information. The second processing module is specifically used to process the word segmentation of the fault information, and generate n-gram features, character repetition feature information features, and character category features of the word segmentation of the fault information, so as to serve as the vector representation of the word segmentation of the fault information; Obtain the system operation information of the online system, process the system operation information, and obtain a vector representation of the system operation information; The vector representation of the fault information and the vector representation of the system operation information are input into the fasttext model to determine the fault cause corresponding to the fault information; wherein, the trained fasttext model is trained by inputting the fault name, fault phenomenon and known fault cause into the fasttext model.

6. The apparatus according to claim 5, characterized in that, The first processing module is specifically used for: The fault information is segmented into words to obtain the segmented fault information. The fault information is processed using the word2vec model to obtain the word segmentation vector of the fault information; The word segmentation vector of the fault information is input into a trained convolutional neural network model to detect whether the fault information is a fault that needs to be investigated. The convolutional neural network model includes: D convolutional layers, each followed by an activation function and a pooling layer. The result of convolution, activation, and pooling of the fault information word segmentation vector is called a feature map, denoted by Z; M fully connected layers, where the kernel size of each convolutional layer is Kn, the stride is Sn, and the number of kernels is Xn, where 1 ≤ n ≤ D; the last fully connected layer has 2 neurons, and the other fully connected layers have Cm neurons, where 1 ≤ m ≤ M; and the activation function after each convolutional layer is a Max-Feature-Map, as shown below: Where 1≤k≤n; The number of Z is 2n; Then the derivative of B is: Where w is the width of the feature map; h is the height of the feature map; i is the i-th column of the feature map, 0 ≤ i < w; j is the j-th row of the feature map, 0 ≤ j < h; 1 ≤ k ≤ n, and R is the real number space. This represents the feature value in the i-th column and j-th row of the k-th feature map. This represents the feature value in the i-th column and j-th row of the k+n-th feature map; This represents the k-th feature map.

7. A computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program is run by the processor, it performs the steps of the method described in any one of claims 1-4.

8. An electronic device, characterized in that, The electronic device includes a memory, a processor, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor of the steps of the method according to any one of claims 1-4.

Images