A server fault diagnosis method, device, equipment and storage medium

Abstract

A server fault diagnosis method, device, equipment and storage medium are disclosed. The method comprises: in response to a server fault diagnosis request, determining at least two target images of a server display screen through a baseboard management controller (BMC) of the server; selecting a fault image from the at least two target images and determining a fault category corresponding to the fault image; processing the fault image to obtain fault description information of the fault image; and generating a fault diagnosis result of the server according to the fault category and the fault description information. The technical solution of the present application can comprehensively analyze the automatic fault of the display screen image of the server and generate an accurate fault diagnosis result, which is helpful for subsequent targeted maintenance of the server.

Description

Technical Field

[0001] This invention relates to the field of computers, and more particularly to a server fault diagnosis method, apparatus, device, and storage medium. Background Technology

[0002] Servers need to provide stable services to customers. If a server malfunctions during operation and is not troubleshooted and repaired in a timely manner, it will seriously affect the use of the client.

[0003] Currently, after a server failure, relevant personnel need to review the failure recordings and conduct troubleshooting and repairs, resulting in a waste of human resources.

[0004] Therefore, how to perform more comprehensive and accurate automatic fault analysis on server display images to facilitate subsequent targeted fault repair is an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a server fault diagnosis method, apparatus, device, and storage medium, which can perform more comprehensive automatic fault analysis on the server's display screen image, generate accurate fault diagnosis results, and help with subsequent targeted maintenance of the server.

[0006] According to one aspect of the present invention, a server fault diagnosis method is provided, comprising:

[0007] In response to a server fault diagnosis request, at least two target images of the server display screen are determined via the server's Baseboard Management Controller (BMC).

[0008] Select a fault image from the at least two target images and determine the fault category corresponding to the fault image;

[0009] The fault image is processed to obtain fault description information of the fault image;

[0010] Based on the fault category and the fault description information, generate fault diagnosis results for the server.

[0011] Optionally, a fault image is selected from the at least two target images, and the fault category corresponding to the fault image is determined, including:

[0012] The target image is coarsely classified using a preset first coarse classification model. If the coarse classification result of the target image indicates that the target image is abnormal, then the target image is regarded as a fault image.

[0013] The fault image is coarsely classified using a preset second coarse classification model to obtain the fault category corresponding to the fault image.

[0014] The above technical solution uses a first coarse classification model to coarsely classify the target image, which can accurately determine the fault image from the target image based on the classification result. By using a second coarse classification model to coarsely classify the fault image, the fault category of the fault image can be further accurately identified, thereby improving the comprehensiveness of server fault diagnosis.

[0015] Optionally, the fault image is processed to obtain fault description information of the fault image, including:

[0016] The fault image is subjected to text detection based on the object detection model, and the fault keywords contained in the fault image are obtained based on the text detection results.

[0017] Based on the pre-defined correspondence between fault keywords and fault description information, the fault description information contained in the fault image is determined.

[0018] The above technical solution utilizes a target detection model to perform text detection on fault images. Based on the text detection results, it can accurately extract fault keywords contained in the fault images. By utilizing the correspondence between fault keywords and fault description information, it can determine the fault description information corresponding to the fault keywords, thereby achieving more granular fault analysis of fault images.

[0019] Optionally, the fault description information includes at least one of the following: reaching peak server access, requests exceeding server capacity, server disk full, server overload, server encountering malicious attack, and server memory exhaustion.

[0020] The above technical solution, by further refining the fault description information to include at least one of the following: reaching peak server access, requests exceeding server capacity, server disk full, server overload, server encountering malicious attacks, and server memory exhaustion, helps to conduct more comprehensive fault diagnosis of the server and improves the richness of the fault description information.

[0021] Optionally, the target image of the server display screen is determined via the server's baseboard management controller (BMC), including:

[0022] The server's Baseboard Management Controller (BMC) determines the video data displayed on the server's screen within a preset time period.

[0023] The candidate images contained in the video data are filtered to determine the target image for the server display screen.

[0024] The above technical solution utilizes the server's Baseboard Management Controller (BMC) to acquire video data from the server's display screen more effectively and accurately without relying on external devices. By filtering candidate images corresponding to the video data, the target image can be determined, avoiding fault analysis of all candidate images and improving the efficiency of subsequent server fault diagnosis.

[0025] Optionally, the fault category is black screen, blue screen, fault prompt box, distorted screen or white screen.

[0026] The above technical solution, by further refining the fault categories to include black screen, blue screen, fault prompt box, distorted screen, and white screen, helps to perform more comprehensive fault diagnosis of servers, determine accurate fault categories, and improve the richness of fault categories.

[0027] Optionally, the first coarse classification model and the second coarse classification model are convolutional neural network (CNN) models, and the object detection model is a model based on the object detection YOLO algorithm.

[0028] The above technical solution further refines the first and second coarse classification models into convolutional neural network (CNN) models and the object detection model into a model based on the YOLO object detection algorithm. It provides an implementation method for constructing coarse classification and object detection models, which helps to perform more comprehensive, accurate and effective fault diagnosis of servers.

[0029] According to another aspect of the present invention, a server fault diagnosis device is provided, comprising:

[0030] The image determination module is used to determine at least two target images of the server display screen in response to a server fault diagnosis request, via the server's Baseboard Management Controller (BMC).

[0031] A category determination module is used to select a fault image from the at least two target images and determine the fault category corresponding to the fault image;

[0032] An information determination module is used to process the fault image to obtain fault description information of the fault image;

[0033] The generation module is used to generate fault diagnosis results for the server based on the fault category and the fault description information.

[0034] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0035] At least one processor; and

[0036] A memory communicatively connected to the at least one processor; wherein,

[0037] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the server fault diagnosis method according to any embodiment of the present invention.

[0038] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the server fault diagnosis method according to any embodiment of the present invention.

[0039] The technical solution of this invention, in response to a server fault diagnosis request, determines at least two target images of the server display screen through the server's Baseboard Management Controller (BMC), selects a fault image from the at least two target images, determines the fault category corresponding to the fault image, processes the fault image to obtain fault description information, and generates a fault diagnosis result for the server based on the fault category and fault description information. By utilizing the server's Baseboard Management Controller (BMC), video data from the server display screen can be acquired more effectively and accurately without relying on external devices. By first determining the fault category of the fault image and further processing the fault image to determine more granular fault description information, a more comprehensive analysis of the server's fault condition can be performed, leading to an accurate fault diagnosis result. This helps relevant personnel to perform targeted repairs on the server. Simultaneously, the technical solution of this invention can achieve fully automated fault diagnosis, avoiding manual acquisition and analysis of fault images, thereby improving the efficiency and accuracy of fault diagnosis.

[0040] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0042] Figure 1 This is a flowchart of a server fault diagnosis method provided in Embodiment 1 of the present invention;

[0043] Figure 2 This is a flowchart of a server fault diagnosis method provided in Embodiment 2 of the present invention;

[0044] Figure 3 This is a flowchart of a server fault diagnosis method provided in Embodiment 3 of the present invention;

[0045] Figure 4 This is a schematic diagram of a server fault diagnosis device according to Embodiment 4 of the present invention;

[0046] Figure 5 This is a schematic diagram of the structure of an electronic device that implements the server fault diagnosis method of this invention. Detailed Implementation

[0047] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0048] It should be noted that the terms "first," "second," "target," "candidate," "alternative," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0049] Example 1

[0050] Figure 1 This is a flowchart of a server fault diagnosis method according to Embodiment 1 of the present invention. This embodiment is applicable to server fault diagnosis from the perspective of fault category and fault description information. The method can be executed by a server fault diagnosis device, which can be implemented in hardware and / or software. The server fault diagnosis device can be configured in an electronic device and executed by the fault analysis module in the server. Figure 1 As shown, the server fault diagnosis method includes:

[0051] S101. In response to a server fault diagnosis request, at least two target images of the server display screen are determined through the server's Baseboard Management Controller (BMC).

[0052] The server fault diagnosis request refers to a request to diagnose faults in the server's display screen. The Baseboard Management Controller (BMC) is a module configured within the server for monitoring server status and recording video of server faults. The target image refers to an image that characterizes the display status of the server's screen.

[0053] It should be noted that the fault video recording function is an important auxiliary function for locating OS (Operating System) faults in the server BMC.

[0054] Optionally, the server's fault analysis module can monitor relevant signals of the server's OS in real time. If a server fault signal is detected, it will automatically generate a server fault diagnosis request. Alternatively, it can automatically generate a server fault diagnosis request at preset intervals to perform a fault diagnosis on the server display screen. The server's fault analysis module can also generate a server fault diagnosis request when it receives a fault command from relevant personnel.

[0055] Optionally, when the server's fault analysis module detects a server fault diagnosis request, it can respond to the server fault diagnosis request by sending a fault video acquisition request to the server's BMC, and determine at least two target images of the server display screen based on the video data of the server display screen fed back by the server's BMC.

[0056] Optionally, the server's baseboard management controller (BMC) determines at least two target images for the server display screen, including: determining video data of the server display screen within a preset time period using the server's baseboard management controller (BMC); and filtering candidate images contained in the video data to determine at least two target images for the server display screen.

[0057] The preset time period refers to the time period prior to detecting a server fault diagnosis request, or it can be the time period prior to detecting a server fault signal. The server display screen refers to the computer display screen of the server. Video data refers to the video stream data about the server's OS captured by the BMC on the server display screen. Candidate images refer to images contained within the video data.

[0058] Optionally, the server's fault analysis module can send a fault video acquisition request to the server BMC. When the server BMC receives the fault video acquisition request, it can use a time period preset before receiving the fault video acquisition request as the target time period, and further filter out the corresponding video data from the historical recorded videos based on the target time period and feed it back to the server fault analysis module.

[0059] Optionally, the server's fault analysis module can send a fault video acquisition request to the server BMC based on the fault time when the server fault signal is detected. This allows the server BMC to use a preset time period before the fault time as the target time period. Furthermore, based on this target time period, the BMC can filter out the corresponding video data from historical recorded videos and feed it back to the server fault analysis module, thus determining the video data of the server display screen within the preset time period.

[0060] Optionally, after determining the video data of the server display screen, each frame of the video data can be used as a candidate image, that is, the candidate images contained in the video data can be determined. Further, based on preset filtering conditions, such as the image clarity being higher than a certain threshold, the image containing the server display screen, or the image being undamaged, the candidate images are filtered, and the candidate images that meet the filtering conditions are determined as at least two target images of the server display screen.

[0061] This optional solution utilizes the server's Baseboard Management Controller (BMC) to acquire video data from the server's display screen more effectively and accurately without relying on external devices. By filtering candidate images corresponding to the video data, the target image can be determined, avoiding fault analysis of all candidate images and improving the efficiency of subsequent server fault diagnosis.

[0062] It should be noted that in related technologies, relevant personnel need to view server recordings and locate problems. In addition, the analysis of fault causes is based on fault signals outside the OS system. When the OS system fails, there may be situations where fault signals cannot be obtained, thus making fault analysis impossible. However, the solution provided by this invention can utilize the server BMC to directly acquire the video stream data of the server OS without relying on other image acquisition devices, and use a pre-trained model for fault analysis, thereby improving the efficiency and accuracy of server fault diagnosis.

[0063] S102. Select a fault image from at least two target images and determine the fault category corresponding to the fault image.

[0064] Here, a fault image refers to an image in the target image that can characterize the server's fault condition. The fault category refers to the type of fault the fault image belongs to; the fault category is coarser-grained fault information than the specific fault description.

[0065] Optionally, the fault category can be black screen, blue screen, fault message box, distorted screen, or white screen. A black screen means the server display screen has a black background. A blue screen means the server display screen has a blue background. A fault message box is a text box on the server display screen used to display information about server faults. A distorted screen refers to the appearance of varying numbers of spots or lines on the server display screen. A white screen means the server display screen has a white background.

[0066] This optional solution further refines the fault categories, including black screen, blue screen, fault prompt box, distorted screen, and white screen, which helps to perform more comprehensive fault diagnosis of the server, determine the accurate fault category, and improve the richness of the fault categories.

[0067] Optionally, the fault analysis module can input at least two target images into a pre-trained coarse classification model to obtain the classification results of whether each target image is a fault image and the fault category to which it belongs. Based on the classification results, the target image with a classification result of no fault is determined as a fault image. That is, the fault image is selected from at least two target images, and the fault category corresponding to the classification result is determined as the fault category of the fault image.

[0068] For example, a pre-trained coarse classification model can obtain six classification results: no fault, black screen, blue screen, fault prompt box, distorted screen, or white screen. After inputting the target image into the pre-trained classification model, if the classification result is that the target image is no fault, then the target image is determined not to be a fault image. If the classification result is that the target image belongs to black screen, blue screen, fault prompt box, distorted screen, or white screen, then the target image is determined to be a fault image. At the same time, the corresponding fault category can be determined based on the classification result of the target image.

[0069] Optionally, the fault analysis module can extract and analyze image features such as color features or contour features of the target image based on a preset classification algorithm, determine whether each target image is a fault image, and determine the fault category of the fault image.

[0070] Optionally, if the target image does not contain a faulty image, the subsequent S103-S104 operations are not performed. Instead, the system monitors in real time whether a server fault diagnosis request is received. When a server fault diagnosis request is detected again, the S101 operation described above in this embodiment is performed.

[0071] S103. Process the fault image to obtain fault description information of the fault image.

[0072] Among them, fault description information refers to the descriptive information in the fault image that can characterize the server fault situation. Fault description information is more granular than fault category.

[0073] Optionally, the fault description information includes at least one of the following: server access peak is reached, requests exceed server capacity, server disk is full, server is overloaded, server is under malicious attack, and server memory is exhausted.

[0074] Optionally, after the fault analysis module determines the fault category of the fault image, it can determine the target detection model corresponding to the fault category, and process the fault image based on the determined target detection model to obtain the fault description information of the fault image.

[0075] This optional solution further refines the fault description information, including at least one of the following: reaching peak server access, requests exceeding server capacity, server disk full, server overload, server under malicious attack, and server memory exhaustion. This helps to conduct more comprehensive fault diagnosis of the server and improves the richness of the fault description information.

[0076] Optionally, the object detection model is a model based on the YOLO object detection algorithm. Convolutional Neural Networks (CNNs) are models used for binary or multi-class classification of images. YOLO (You Only Look Once) is an object detection algorithm.

[0077] It should be noted that different target detection models can be trained in advance for fault images of different fault categories, or the same target detection model can be used for fault images of different fault categories. This invention does not limit this.

[0078] This alternative approach further refines the target detection model into a model based on the YOLO target detection algorithm, providing a feasible method for constructing a target detection model, which helps to perform more comprehensive, accurate and effective fault diagnosis of servers.

[0079] S104. Generate fault diagnosis results for the server based on the fault category and fault description information.

[0080] The fault diagnosis result refers to the result generated after fault analysis of the server. The fault diagnosis result can include fault category and fault description information.

[0081] Optionally, a fault diagnosis result for the server can be generated based on the fault category and fault description information, combined with the fault image and the fault time corresponding to the fault image; alternatively, the fault category number, fault description number, fault image, and fault time can be combined based on the preset correspondence between fault category and fault category number, as well as the correspondence between fault description information and fault description number, to generate a fault diagnosis result for the server.

[0082] Optionally, after determining the fault category of the fault image, steps S103-104 as described in this embodiment of the invention can be executed uniformly; alternatively, steps S103-104 as described in this embodiment of the invention can be executed only when the fault category is a black screen, a blue screen, or a fault prompt box. When the fault category is a distorted screen or a white screen, considering that meaningful fault description information cannot be extracted from the target image, steps S103-104 can be refused to be executed, and the server fault diagnosis result can be generated directly based on the determined fault category.

[0083] Optionally, after generating the server fault diagnosis results, the fault diagnosis results can be sent to relevant personnel based on a preset notification method, such as sending the fault diagnosis results to maintenance personnel via email.

[0084] The technical solution of this invention, in response to a server fault diagnosis request, determines at least two target images of the server display screen through the server's Baseboard Management Controller (BMC), selects a fault image from the at least two target images, determines the fault category corresponding to the fault image, processes the fault image to obtain fault description information, and generates a fault diagnosis result for the server based on the fault category and fault description information. By utilizing the server's Baseboard Management Controller (BMC), video data from the server display screen can be acquired more effectively and accurately without relying on external devices. By first determining the fault category of the fault image and further processing the fault image to determine more granular fault description information, a more comprehensive analysis of the server's fault condition can be performed, resulting in an accurate fault diagnosis result, which helps relevant personnel to perform targeted repairs on the server. Simultaneously, the technical solution of this invention can achieve fully automated fault diagnosis, avoiding manual acquisition and analysis of fault images, thereby improving the efficiency and accuracy of fault diagnosis.

[0085] Example 2

[0086] Figure 2 This is a flowchart of a server fault diagnosis method according to Embodiment 2 of the present invention. This embodiment is optimized and improved based on the above technical solutions.

[0087] Furthermore, the process of "selecting a fault image from at least two target images and determining the fault category corresponding to the fault image" is refined to "using a preset first coarse classification model to perform coarse classification on the target image. If the coarse classification result of the target image indicates that the target image is abnormal, then the target image is taken as a fault image; using a preset second coarse classification model to perform coarse classification on the fault image to obtain the fault category corresponding to the fault image" in order to improve the method of determining the fault image and the fault category corresponding to the fault image.

[0088] Furthermore, the process of "processing fault images to obtain fault description information of fault images" is refined into "performing text detection on fault images based on target detection models, and obtaining fault keywords contained in fault images based on text detection results; determining the fault description information contained in fault images based on the pre-defined correspondence between fault keywords and fault description information" to improve the method of determining fault description information of fault images.

[0089] like Figure 2 As shown, the method includes the following specific steps:

[0090] S201. In response to a server fault diagnosis request, at least two target images of the server display screen are determined through the server's Baseboard Management Controller (BMC).

[0091] S202. Use the preset first coarse classification model to perform coarse classification on the target image. If the coarse classification result of the target image is that the target image is abnormal, then the target image is regarded as a fault image.

[0092] The first coarse classification model refers to a pre-trained binary classification model that can classify a target image as either having an anomaly or not having an anomaly.

[0093] Optionally, at least two target images can be input into a pre-trained first coarse classification model to obtain a classification result indicating whether each target image is abnormal. If the classification result indicates that the target image is abnormal, then the target image is regarded as a faulty image; otherwise, the target image is excluded without further processing. In this way, a faulty image can be selected from at least two target images.

[0094] Optionally, the training method for the first coarse classification model can be as follows: collect a preset number of server display screen images as sample images, and use a preset image annotation tool, such as LABELIMG, to annotate each sample image based on the known information of whether each sample image is a fault image. Input the annotated sample images and the annotation information of each sample image into the preset coarse classification model. Based on the deviation between the prediction result of the coarse classification model and the pre-annotated information of the sample images, iteratively train the preset coarse classification model to obtain the first coarse classification model that can classify images into fault images and non-fault images.

[0095] Optionally, if the coarse classification result of the first coarse classification model is that the target image does not have any anomalies, then the target image can be considered to be fault-free, and thus the target image can be discarded without performing subsequent operations on it. In this way, the waste of resources caused by performing subsequent fault analysis on all target images can be avoided, thereby improving the efficiency of server fault diagnosis.

[0096] S203. Use the preset second coarse classification model to perform coarse classification on the fault image to obtain the fault category corresponding to the fault image.

[0097] The second coarse classification model refers to a multi-classification model that can classify the target image into different fault categories. For example, the classification result of the second coarse classification model can be a black screen, a blue screen, a fault prompt box, a distorted screen, or a white screen.

[0098] Optionally, the first and second coarse classification models can be convolutional neural network (CNN) models. This optional approach, by further refining the first and second coarse classification models into CNN models, provides a feasible method for constructing coarse classification models, which helps to perform more comprehensive, accurate, and effective fault diagnosis of servers.

[0099] Optionally, the determined fault images can be input into a pre-trained second coarse classification model to obtain the classification results for each fault image. Based on the fault identifier in the classification results, and the correspondence between the fault identifier and the fault category, the fault category corresponding to the fault image can be determined. For example, the correspondence between the fault identifier and the fault category could be that fault identifiers 1-5 correspond to fault categories such as black screen, blue screen, fault prompt box, distorted screen, or white screen, respectively.

[0100] S204. Process the fault image to obtain fault description information of the fault image.

[0101] Optionally, after determining the fault category of the fault image, a target detection model corresponding to the fault category can be determined for each fault image, and the fault image can be processed using the target detection model to obtain the fault description information of the fault image.

[0102] Optionally, the fault image is processed to obtain fault description information of the fault image, including: performing text detection on the fault image based on the target detection model, and obtaining fault keywords contained in the fault image according to the text detection results; determining the fault description information contained in the fault image according to the preset correspondence between fault keywords and fault description information.

[0103] Among them, fault keywords refer to keywords that can characterize server fault conditions. Fault keywords can be fault numbers corresponding to fault description information, the same as fault description information, or content that expresses fault description information. For example, if the fault description information is "server memory exhausted," the fault keyword can be the preset number Error101, or the same as the fault description information "server memory exhausted," or content such as "Memory exhausted, please clean up memory!" or "OUT of memory."

[0104] Optionally, the fault image can be input into the corresponding target detection model to perform text detection on the fault image, determine the text content contained in the fault image, i.e., the text detection result, and further match the text content in the text detection result with preset fault keywords to determine the fault keywords contained in the text content, i.e., obtain the fault keywords contained in the fault image.

[0105] Optionally, after determining the fault keywords contained in the fault image, the fault description information corresponding to the fault keywords can be determined based on the preset correspondence between the fault keywords and the fault description information.

[0106] This alternative solution utilizes a target detection model to perform text detection on fault images. Based on the text detection results, it can accurately extract fault keywords contained in the fault images. By utilizing the correspondence between fault keywords and fault description information, it can determine the fault description information corresponding to the fault keywords, thereby achieving more granular fault analysis of fault images.

[0107] S205. Based on the fault category and fault description information, generate fault diagnosis results for the server.

[0108] Optionally, after generating the server fault diagnosis results, it is also possible to obtain the unidentified fault description information fed back by relevant personnel, and correct the weight parameters of the target detection model, the first coarse classification model, and the second coarse classification model based on the unidentified fault description information and the corresponding fault category. This will facilitate subsequent operations in response to server fault diagnosis requests, allowing fault diagnosis to be performed based on the corrected target detection model, the first coarse classification model, and the second coarse classification model, thereby improving the efficiency and accuracy of fault diagnosis.

[0109] The technical solution of this invention, in response to a server fault diagnosis request, determines at least two target images of the server display screen through the server's Baseboard Management Controller (BMC). It then performs a coarse classification of the target images using a preset first coarse classification model. If the coarse classification result indicates an anomaly in the target image, it is designated as a fault image. A second coarse classification model is then used to further classify the fault image, obtaining its corresponding fault category. The fault image is then processed to obtain fault description information. Finally, based on the fault category and description information, a fault diagnosis result for the server is generated. By utilizing the first coarse classification model to coarsely classify the target images, the fault image can be accurately identified from the target images based on the classification results. Utilizing the second coarse classification model further refines the identification of the fault category, thereby improving the comprehensiveness of server fault diagnosis.

[0110] Example 3

[0111] Figure 3 This is a flowchart of a server fault diagnosis method according to Embodiment 3 of the present invention. Based on the above embodiments, this embodiment provides a preferred example.

[0112] like Figure 3 As shown, the method includes the following specific steps:

[0113] S301. In response to a server fault diagnosis request, determine the video data of the server display screen within a preset time period through the server's Baseboard Management Controller (BMC).

[0114] S302. Filter the candidate images contained in the video data to determine at least two target images for the server display screen.

[0115] S303. Use the preset first coarse classification model to perform coarse classification on the target image. If the coarse classification result of the target image is that the target image is abnormal, then the target image is regarded as a fault image.

[0116] It should be noted that if the coarse classification result of the target image is that there are no anomalies, the process ends and the subsequent operations S304-S308 are not executed.

[0117] S304. Use the preset second coarse classification model to perform coarse classification on the fault image to obtain the fault category corresponding to the fault image. If the fault category is black screen, blue screen or fault prompt box, then execute S305-S307. If the fault category is distorted screen or white screen, then execute S308.

[0118] S305. If the fault category is black screen, blue screen or fault prompt box, then perform text detection on the fault image based on the target detection model, and obtain the fault keywords contained in the fault image based on the text detection results.

[0119] S306. Determine the fault description information contained in the fault image based on the preset correspondence between fault keywords and fault description information.

[0120] S307. Generate fault diagnosis results for the server based on the fault category and fault description information.

[0121] S308. If the fault category is screen flickering or white screen, then generate a fault diagnosis result for the server based on the fault category.

[0122] Example 4

[0123] Figure 4 This is a schematic diagram of a server fault diagnosis device according to Embodiment 4 of the present invention. The server fault diagnosis device provided in this embodiment of the present invention is applicable to situations where server fault diagnosis is performed from the perspective of fault category and fault description information. This server fault diagnosis device can be implemented in hardware and / or software, configured in a server, and executed by a fault analysis module within the server, such as... Figure 4 As shown, the device specifically includes: an image determination module 401, a category determination module 402, an information determination module 403, and a generation module 404. Among them,

[0124] Image determination module 401 is used to determine at least two target images of the server display screen in response to a server fault diagnosis request via the server's baseboard management controller (BMC).

[0125] The category determination module 402 is used to select a fault image from the at least two target images and determine the fault category corresponding to the fault image;

[0126] The information determination module 403 is used to process the fault image to obtain fault description information of the fault image;

[0127] The generation module 404 is used to generate a fault diagnosis result for the server based on the fault category and the fault description information.

[0128] The technical solution of this invention, in response to a server fault diagnosis request, determines at least two target images of the server display screen through the server's Baseboard Management Controller (BMC), selects a fault image from the at least two target images, determines the fault category corresponding to the fault image, processes the fault image to obtain fault description information, and generates a fault diagnosis result for the server based on the fault category and fault description information. By utilizing the server's Baseboard Management Controller (BMC), video data from the server display screen can be acquired more effectively and accurately without relying on external devices. By first determining the fault category of the fault image and further processing the fault image to determine more granular fault description information, a more comprehensive analysis of the server's fault condition can be performed, leading to an accurate fault diagnosis result. This helps relevant personnel to perform targeted repairs on the server. Simultaneously, the technical solution of this invention can achieve fully automated fault diagnosis, avoiding manual acquisition and analysis of fault images, thereby improving the efficiency and accuracy of fault diagnosis.

[0129] Furthermore, the category determination module 402 is specifically used for:

[0130] The target image is coarsely classified using a preset first coarse classification model. If the coarse classification result of the target image indicates that the target image is abnormal, then the target image is regarded as a fault image.

[0131] The fault image is coarsely classified using a preset second coarse classification model to obtain the fault category corresponding to the fault image.

[0132] Furthermore, the information determination module 403 is specifically used for:

[0133] The fault image is subjected to text detection based on the object detection model, and the fault keywords contained in the fault image are obtained based on the text detection results.

[0134] Based on the pre-defined correspondence between fault keywords and fault description information, the fault description information contained in the fault image is determined.

[0135] Furthermore, the fault description information includes at least one of the following: reaching peak server access, requests exceeding server capacity, server disk full, server overload, server encountering malicious attack, and server memory exhaustion.

[0136] Furthermore, the image determination module 401 is specifically used for:

[0137] The server's Baseboard Management Controller (BMC) determines the video data displayed on the server's screen within a preset time period.

[0138] The candidate images contained in the video data are filtered to determine at least two target images for the server display screen.

[0139] Furthermore, the fault categories are black screen, blue screen, fault prompt box, distorted screen, or white screen.

[0140] Furthermore, the first coarse classification model and the second coarse classification model are convolutional neural network (CNN) models, and the object detection model is a model based on the object detection YOLO algorithm.

[0141] Example 5

[0142] Figure 5 This is a schematic diagram of the structure of an electronic device that implements the server fault diagnosis method of this invention. Figure 5 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0143] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0144] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0145] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as server fault diagnosis methods.

[0146] In some embodiments, the server fault diagnosis method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the server fault diagnosis method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the server fault diagnosis method by any other suitable means (e.g., by means of firmware).

[0147] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0148] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0149] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0150] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0151] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0152] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0153] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0154] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A server fault diagnosis method, characterized in that, include: In response to a server fault diagnosis request, at least two target images of the server display screen are determined via the server's Baseboard Management Controller (BMC). Select a fault image from the at least two target images and determine the fault category corresponding to the fault image; The fault image is processed to obtain fault description information of the fault image, including: for each fault image, determining the target detection model corresponding to the fault category, and using the target detection model to process the fault image to obtain fault description information of the fault image; Based on the fault category and the fault description information, generate fault diagnosis results for the server; The process of selecting a fault image from the at least two target images and determining the fault category corresponding to the fault image includes: The target image is coarsely classified using a preset first coarse classification model. If the coarse classification result of the target image indicates that the target image has an anomaly, then the target image is regarded as a fault image. Here, the first coarse classification model refers to a pre-trained binary classification model used to classify the target image as having an anomaly or not having an anomaly. The fault image is coarsely classified using a preset second coarse classification model to obtain the fault category corresponding to the fault image; wherein, the second coarse classification model is a multi-classification model used to classify the target image into different fault categories; the fault category is black screen, blue screen, fault prompt box, distorted screen or white screen.

2. The method according to claim 1, characterized in that, The fault image is processed to obtain fault description information, including: The fault image is subjected to text detection based on the object detection model, and the fault keywords contained in the fault image are obtained based on the text detection results. Based on the pre-defined correspondence between fault keywords and fault description information, the fault description information contained in the fault image is determined.

3. The method according to claim 1, characterized in that, The fault description information includes at least one of the following: reaching peak server access, requests exceeding server capacity, server disk full, server overload, server under malicious attack, and server memory exhaustion.

4. The method according to claim 1, characterized in that, The server's Baseboard Management Controller (BMC) determines at least two target images for the server display, including: The server's Baseboard Management Controller (BMC) determines the video data displayed on the server's screen within a preset time period. The candidate images contained in the video data are filtered to determine at least two target images for the server display screen.

5. The method according to claim 1, characterized in that, The first coarse classification model and the second coarse classification model are convolutional neural network (CNN) models, and the object detection model is a model based on the object detection YOLO algorithm.

6. A server fault diagnosis device, characterized in that, include: The image determination module is used to determine at least two target images of the server display screen in response to a server fault diagnosis request, via the server's Baseboard Management Controller (BMC). A category determination module is used to select a fault image from the at least two target images and determine the fault category corresponding to the fault image; An information determination module is used to process the fault image to obtain fault description information of the fault image; The generation module is used to generate fault diagnosis results for the server based on the fault category and the fault description information; Specifically, the information determination module is used to: determine the target detection model corresponding to the fault category for each fault image, and use the target detection model to process the fault image to obtain the fault description information of the fault image; Specifically, the category determination module is used to: perform coarse classification on the target image using a preset first coarse classification model; if the coarse classification result of the target image indicates that the target image has an anomaly, then the target image is regarded as a fault image; wherein, the first coarse classification model refers to a pre-trained binary classification model used to classify the target image as having an anomaly or not having an anomaly. The fault image is coarsely classified using a preset second coarse classification model to obtain the fault category corresponding to the fault image; wherein, the second coarse classification model is a multi-classification model used to classify the target image into different fault categories; the fault category is black screen, blue screen, fault prompt box, distorted screen or white screen.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the server fault diagnosis method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute the server fault diagnosis method according to any one of claims 1-5.

Images