Machine room fault manual operation tracing method, device and system and medium

Abstract

The invention provides a machine room fault manual operation tracing method, device and system and a medium, and relates to the technical field of fault and video analysis, and the method comprises the steps: responding to a system fault monitored by system fault monitoring software, and carrying out the analysis and obtaining the generation equipment and generation time period of manual operation causing the system fault; and obtaining a monitoring video of a machine room where the generation equipment is located in the generation time period, and capturing manual operation causing the system fault in the monitoring video. According to the invention, machine room environment monitoring is associated with software monitoring, so that the information that corresponding physical equipment is manually operated recently can be automatically obtained at the first time when the software monitoring system gives an alarm.

Description

Technical Field

[0001] The present disclosure at least relates to the field of fault and video analysis technology, and in particular to a method for tracing the source of a human operation caused by a computer room fault, a device for tracing the source of a human operation caused by a computer room fault, a system for tracing the source of a human operation caused by a computer room fault, and a computer-readable storage medium. Background Art

[0002] At present, large-scale enterprise data rooms, core rooms and other equipment rooms are the foundation of digital enterprise operations. Monitoring of computer rooms and tracing of human operations are particularly important. How enterprises establish and improve a reasonable computer room monitoring process system is becoming increasingly important.

[0003] The prior art processes the fault analysis of the equipment in the computer room separately from the video surveillance viewing of the computer room, resulting in low efficiency when attempting to obtain the human operation that caused the fault of the equipment in the computer room. Summary of the invention

[0004] The technical problem to be solved by the present disclosure is to provide a method for tracing the source of human operation of a computer room fault, a device for tracing the source of human operation of a computer room fault, a system for tracing the source of human operation of a computer room fault, and a computer-readable storage medium in order to solve the problem of how to automatically obtain the information on the recent human operation of the corresponding physical equipment as soon as the software monitoring system alarm is generated.

[0005] In a first aspect, the present disclosure provides a method for tracing the source of a machine room failure caused by a human operation.

[0006] The methods include:

[0007] In response to the system fault monitoring software monitoring the system fault, analyzing and obtaining the occurrence device and occurrence time period of the human operation that caused the system fault;

[0008] Acquire surveillance video of the computer room where the device where the fault occurs is located during the occurrence period, and capture the human operation that causes the system fault in the surveillance video.

[0009] Further, in response to the system fault monitoring software monitoring the system fault, analyzing and obtaining the occurrence device and occurrence time period of the human operation that caused the system fault, specifically including:

[0010] Receiving system fault alarm information sent by system fault monitoring software in response to monitoring a system fault;

[0011] The knowledge graph is used to perform root cause analysis on the system fault alarm information to locate the device and time period of the human operation that caused the system fault.

[0012] Further, wherein:

[0013] The knowledge graph is pre-constructed by collecting the association between historical system fault alarm information and historical human operation information, wherein the historical human operation information includes the root cause device causing the historical system fault and the time when the erroneous human operation was performed on the root cause device.

[0014] Furthermore, capturing the human operation causing the system failure in the monitoring video specifically includes:

[0015] Processing the surveillance video using an inter-frame difference method to group images in the surveillance video;

[0016] The Gaussian mixture model is called to process the grouped images to obtain the personnel who have performed manual operations on the generating device.

[0017] Further, the monitoring video is processed using an inter-frame difference method to group images in the monitoring video, specifically including:

[0018] Calculating the inter-frame intensity difference between each two adjacent frames of the surveillance video in sequence;

[0019] Selecting a number of frames whose inter-frame difference strength is higher than a threshold as key frames of the monitoring video image;

[0020] All images in the surveillance video are grouped with the key frame as the starting frame of each group of images.

[0021] Further, calling the Gaussian mixture model to process the grouped images to obtain the personnel who performed manual operations on the generating device, specifically including:

[0022] Calling a Gaussian mixture model to process the key frame, and obtaining a first key frame having an active person in the foreground of the image;

[0023] Calling the Gaussian mixture model to process all images of the group where the first key frame is located to obtain the activity trajectory of the active person;

[0024] It is analyzed according to the activity track whether the activity personnel have performed manual operations on the generating device.

[0025] Furthermore, the method further comprises:

[0026] In response to capturing the human operation that caused the system failure from the monitoring video, obtain a second time period of the human operation that caused the system failure in the monitoring video, obtain an operation log of the causing device in the second time period, and obtain the content of the erroneous operation of the causing device by the active person based on the operation log.

[0027] In a second aspect, the present disclosure provides a device for tracing the source of a human operation in a computer room fault, the device comprising:

[0028] A fault location module is used to analyze and obtain the device and time period of the human operation that caused the system fault in response to the system fault monitoring software monitoring the system fault;

[0029] The video analysis module is connected to the fault location module and is used to obtain the monitoring video of the computer room where the fault device is located during the occurrence period and capture the human operation that causes the system fault in the monitoring video.

[0030] In a third aspect, the present disclosure provides a computer room fault human operation tracing system, characterized in that the system comprises:

[0031] A device for tracing the source of a machine room fault caused by a human operation, used to implement the method for tracing the source of a machine room fault caused by a human operation as described above;

[0032] System fault monitoring software, connected to the computer room fault manual operation tracing device and the computer room equipment, is used to monitor whether the computer room equipment has a system fault, and send system fault alarm information to the computer room fault manual operation tracing device after a system fault occurs;

[0033] The computer room monitoring equipment and the computer room fault artificial operation tracing device are used to provide the computer room's monitoring video to the computer room fault artificial operation tracing device.

[0034] In a fourth aspect, the present disclosure provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, the method for tracing the source of a human operation caused by a computer room fault as described above is implemented.

[0035] The present invention provides a method for tracing the source of human operation of computer room faults, a device for tracing the source of human operation of computer room faults, a system for tracing the source of human operation of computer room faults, and a computer-readable storage medium, which associate computer room environment monitoring with software monitoring, and aim to automatically obtain the information of human operation of the corresponding physical equipment in the near future as soon as the alarm of the software monitoring system is generated. It includes: monitoring the alarm of the software system deployed on the equipment in the computer room, associating the fault alarm information with the information of the computer room and the camera; extracting information from the video stream captured by the camera in the computer room, and tracing the source of the human operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a flow chart of a method for tracing the source of a human operation in a computer room fault according to an embodiment of the present disclosure;

[0037] Figure 2 It is a structural schematic diagram of a device for tracing the source of a human operation of a machine room fault according to an embodiment of the present disclosure;

[0038] Figure 3It is a schematic diagram of a computer room fault manual operation tracing system according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solution of the present disclosure, the embodiments of the present disclosure will be further described in detail below with reference to the accompanying drawings.

[0040] It should be understood that the specific embodiments and drawings described herein are only used to explain the present disclosure rather than to limit the present disclosure.

[0041] It can be understood that, in the absence of conflict, the various embodiments of the present disclosure and the various features in the embodiments can be combined with each other.

[0042] It can be understood that, for the convenience of description, the drawings of the present disclosure only show parts related to the present disclosure, while parts irrelevant to the present disclosure are not shown in the drawings.

[0043] It can be understood that each unit or unit involved in the embodiments of the present disclosure may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units or units may be integrated into one physical structure.

[0044] It will be understood that, in the absence of conflict, the functions and steps marked in the flowcharts and block diagrams of the present disclosure may occur in an order different from that marked in the drawings.

[0045] It is understood that the flowcharts and block diagrams of the present disclosure illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the embodiments of the present disclosure. Each box in the flowchart or block diagram may represent a unit, unit, program segment, or code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or may be implemented by a combination of hardware and computer instructions.

[0046] It can be understood that the units and units involved in the embodiments of the present disclosure can be implemented by software or hardware, for example, the units and units can be located in a processor.

[0047] Embodiment 1:

[0048] like Figure 1 As shown, the present disclosure provides a method for tracing the source of a human operation in a computer room fault, the method comprising:

[0049] S11, in response to the system fault monitoring software monitoring a system fault, analyzing and obtaining the device and time period of the human operation that caused the system fault;

[0050] S12: Obtain a surveillance video of a computer room where the device where the fault occurs is located during the occurrence period, and capture the human operation that causes the system fault in the surveillance video.

[0051] In this embodiment, the method associates the computer room environment monitoring with the software monitoring, aiming to automatically obtain the information about the recent human operation of the corresponding physical equipment as soon as the software monitoring system alarm is generated. It includes: monitoring the alarm of the software system deployed on the computer room equipment, associating the fault alarm information with the computer room information and camera information; extracting information from the video stream captured by the computer room camera, and tracing the source of the human operation. The device corresponding to the method is as follows Figure 2 As shown, it includes a fault location module 11 and a video analysis module 12, which are used to execute steps S11 and S12 respectively.

[0052] Specifically, this embodiment can provide a method for tracing the source of human operations in the computer room based on Gaussian mixture models and knowledge graphs to locate human operations that cause equipment failures in the computer room. Taking the operator network computer room monitoring as an example, a city has a large number of managed computer rooms, involving a large number of cameras. By analyzing the surveillance videos of these cameras, it is possible to obtain information about people entering the computer room to operate the equipment. However, the amount of data is huge, and how to efficiently analyze these surveillance videos has become an urgent problem to be solved. First, this embodiment performs alarm monitoring on the software system deployed on the computer room equipment, and associates the fault alarm information with the computer room information and camera information through the knowledge graph; then, the real-time video stream captured by the computer room camera is extracted through the inter-frame difference method and Gaussian mixture model to extract key frames, and a set of behavioral motion trajectory diagrams and corresponding camera information are extracted to capture the human operations that cause failures. Among them, the Gaussian mixture model, namely GMM (Gaussian Mixture Module), is a mixture model formed by superimposing multiple Gaussian models, and is often used to separate the foreground and background in image processing; the knowledge graph is a technology that describes concepts and their relationships in a structured manner. The graph database mainly used in this embodiment is a graph-oriented database. The data is mainly reflected in nodes and relationships, and key-value pairs can also be processed. It is more efficient for complex relationship problems. This embodiment reduces the labor cost and increases the timeliness of fault location analysis by designing a human operation tracing process in the computer room based on the Gaussian mixture model and the knowledge graph.

[0053] In one embodiment, S11, in response to the system fault monitoring software monitoring a system fault, analyzing and obtaining the device and time period of the human operation that caused the system fault, specifically includes:

[0054] Receiving system fault alarm information sent by system fault monitoring software in response to monitoring a system fault;

[0055] The knowledge graph is used to perform root cause analysis on the system fault alarm information to locate the device and time period of the human operation that caused the system fault.

[0056] In this embodiment, if Figure 2 The deployment of the equipment room fault human operation tracing device is shown as follows Figure 3 In the system shown in the figure, the software system is deployed on hardware devices such as servers, the computer room camera is responsible for monitoring these hardware devices, and the system fault monitoring software is responsible for monitoring the system status; the monitoring software collects system fault alarm information from the computer room software and hardware equipment, including the fault content and alarm time, and provides it to the computer room fault human operation tracing device; the computer room monitoring equipment takes real-time video streams and transmits them to the computer room fault human operation tracing device; the computer room fault human operation tracing device comprehensively analyzes the information of the two, traces back the system fault alarm information through the knowledge graph, searches for the equipment information related to the fault, conducts in-depth analysis to accurately locate the cause of the fault, and then analyzes the corresponding monitoring video to locate the human operation that causes the fault. Among them, the system fault alarm information is analyzed by using the knowledge graph to locate the occurrence equipment and occurrence period of the human operation that causes the system fault. For example, first, a large number of system fault alarm information is root-caused to locate the root cause equipment of the fault. The root cause location can be realized by referring to the existing technology, and the direct or indirect relationship between the root cause equipment of the fault and the camera is queried; then, the time point when the root cause equipment starts to fail is determined, and the approximate time before and after the time point is determined as the basis for obtaining the video. For example, if the root cause of the fault is located in area 1 of room A, the system will automatically retrieve the camera list related to area 1 in the knowledge graph and filter out the video streams shot before and after the fault time. This can directly associate the fault time with the specific camera information, thereby obtaining the on-site video stream when the fault occurred, providing intuitive evidence for fault analysis. Root cause location narrows the scope of video viewing and helps to quickly locate the location where the erroneous human operation occurred.

[0057] In one embodiment, wherein:

[0058] The knowledge graph is pre-constructed by collecting the association between historical system fault alarm information and historical human operation information, wherein the historical human operation information includes the root cause device causing the historical system fault and the time when the erroneous human operation was performed on the root cause device.

[0059] In this embodiment, the knowledge graph is pre-established based on historical fault information and actual deployment information of the computer room. A knowledge graph consisting of entity information such as computer room cameras, computer room equipment, equipment deployment software, and association relationships obtained from historical fault analysis is designed as the basis for fault alarm information analysis.

[0060] like Figure 3As shown in the figure, the system fault monitoring software is responsible for monitoring the system status. When the system fails, the software transmits the system information, fault information, alarm information, etc. to the graph database of the knowledge graph, and associates the pre-entered computer room information, equipment information, camera information, system information, etc. in the knowledge graph. Finally, the camera information and fault time corresponding to the faulty system equipment are associated and transmitted to the camera monitoring system.

[0061] The camera monitoring system has received the real-time video stream captured by the camera in the computer room in advance, obtained the camera information and the computer room monitoring video corresponding to the fault time, called the deployed video image analysis algorithm (inter-frame difference method and Gaussian mixture model), and performed key frame extraction and human trajectory highlighting on the video stream within a certain period before the fault time point. By analyzing the human activities in the key frames, the human operation trajectory is automatically identified and highlighted. Human operation is often the direct or indirect cause of the fault. By highlighting, the operation point that may affect the stability of the system can be quickly located, providing important clues for troubleshooting. Finally, several recent human operation trajectories are locked. "Several" specific operation trajectories refer to all human operations within a set time window before and after the fault time point (for example, 2 hours before and 1 hour after). The length of the time window can be dynamically adjusted according to the fault type and environmental changes to ensure that all possible related operations are covered. In this way, the system failure that may be caused by human operation in the computer room can be traced more quickly, which is more efficient than manually reviewing the monitoring video for the entire period of time.

[0062] The graph structure of the knowledge graph is composed of many entities and relationships. The entity and relationship information constitutes the core basis for understanding the cause of the fault, locating the source of the fault, and evaluating the impact of the fault. A comprehensive fault analysis framework is constructed through this information. Entities include cameras, computer rooms, equipment, and systems. The relationships are mainly fault association relationships, equipment association systems, camera association equipment, and computer room association cameras. For example, the basic information of the camera includes: camera number, computer room location, shooting equipment, etc.; the basic information of the computer room includes: computer room name, camera distribution information in the computer room, etc.; the basic information of the equipment includes: equipment model, equipment operation time, etc.; the basic information of the system includes: system name, system abbreviation, etc.

[0063] Neo4j (a graph database software) can be used as a graph structure storage database for knowledge graphs. Compared with traditional relational databases and other NoSQL (Not Only SQL) solutions, Neo4j has shown significant superiority, especially in the management of associated data. It can not only intuitively map and display the complex connections between data, but also has the characteristic of maintaining stable query performance as the data scale increases. Neo4j's graph data model allows it to process highly interconnected data sets in a natural and efficient way, providing better performance and flexibility than traditional methods.

[0064] In one implementation, capturing the human operation causing the system failure in the monitoring video in S12 specifically includes:

[0065] Processing the surveillance video using an inter-frame difference method to group images in the surveillance video;

[0066] The Gaussian mixture model is called to process the grouped images to obtain the personnel who have performed manual operations on the generating device.

[0067] In this embodiment, the real-time video stream captured by the camera in the computer room is used to extract key frame groups through the inter-frame difference method, and a group of behavior motion trajectory diagrams and corresponding camera information are extracted through the Gaussian mixture model.

[0068] In one embodiment, the surveillance video is processed using an inter-frame difference method to group images in the surveillance video, specifically including:

[0069] Calculating the inter-frame intensity difference between each two adjacent frames of the surveillance video in sequence;

[0070] Selecting a number of frames whose inter-frame difference strength is higher than a threshold as key frames of the monitoring video image;

[0071] All images in the surveillance video are grouped with the key frames as the first and last frames of each group of images.

[0072] In this embodiment, the real-time video stream is truncated according to the time period when the fault occurs, and then the inter-frame difference method is used for a single video to obtain key frame groups. Specifically, the direct intensity difference between two adjacent frames is calculated in turn, and then the average differential intensity of the entire video is obtained. All frames are sorted according to the average inter-frame differential intensity, and several frames with average inter-frame differential intensity higher than a threshold (such as the latter frame of two adjacent frames) are selected as key frames of the video. The key frame reflects that the content of the two images before and after has a large change. For example, if an active person enters the computer room, the adjacent video segments of the key frame are divided into one group.

[0073] In one implementation, calling a Gaussian mixture model to process the grouped images to obtain a person who has performed a manual operation on the generating device specifically includes:

[0074] Calling a Gaussian mixture model to process the key frame, and obtaining a first key frame having an active person in the foreground of the image;

[0075] Calling the Gaussian mixture model to process all images of the group where the first key frame is located to obtain the activity trajectory of the active person;

[0076] It is analyzed according to the activity track whether the activity personnel have performed manual operations on the generating device.

[0077] In this embodiment, after the morphological opening operation is performed on the key frame image, the Gaussian mixture model is called to perform background removal, separate the foreground (active personnel) and background (computer room) of the key frame image, obtain the key frame in which the active personnel appear, and select the active personnel on the group of original images; in this way, what is finally presented after association is a group image of personnel activities within the required time, and the personnel activities are captured and highlighted by the algorithm throughout the process, so that abnormal situations can be discovered and located more quickly, and it is convenient to trace the source of human operations, such as locating the equipment where the fault occurred and confirming whether the active personnel operated the equipment.

[0078] In one embodiment, the method further comprises:

[0079] In response to capturing the human operation that caused the system failure from the monitoring video, obtain a second time period of the human operation that caused the system failure in the monitoring video, obtain an operation log of the causing device in the second time period, and obtain the content of the erroneous operation of the causing device by the active person based on the operation log.

[0080] In this embodiment, when performing computer room monitoring and manual operation tracing, one or more of the following superposition methods may be used:

[0081] 1) Operation log review: refers to checking the operation log of the corresponding equipment and analyzing it to determine whether the personnel's operation behavior is correct. When personnel enter the machine room or operate the equipment, the personnel's identity will be authenticated. Only authorized personnel can operate, so the operator can also be located. However, a large amount of operation log data needs to be analyzed, and there are high inspection delays and certain restrictions on equipment types and operation types.

[0082] 2) Camera monitoring: Multiple cameras are deployed in the computer room to shoot the computer room in real time from multiple angles, thereby monitoring the behavior of personnel in the computer room. Through 24-hour on-duty monitoring or by viewing and replaying videos within a certain period of time in the later stage, the operation behavior of personnel can be traced. The disadvantage of this method is that there are too many videos to monitor at the same time, which makes it difficult to monitor manually in real time. By viewing and replaying, it is often only possible to locate a time period with a large span, and manual playback is required to check frame by frame, which is labor-intensive and time-consuming.

[0083] In this embodiment, in response to analyzing the activity trajectory that the active person has performed manual operations on the device, the second time period in which the active person has performed manual operations on the device is obtained, the operation log of the device in the second time period is obtained, and the content of the erroneous operation of the active person on the device is obtained according to the operation log. For example, pictures of the active person performing manual operations on the device where the fault occurs can be obtained, and the second time period of these pictures when the video is taken can be obtained. In combination with the device operation log during this period, it can be further confirmed whether the person has performed an erroneous operation, and the content of the erroneous operation can be determined for correction.

[0084] Embodiment 2:

[0085] like Figure 2 As shown, the present disclosure provides a device for tracing the source of a machine room fault caused by a human operation, the device comprising:

[0086] The fault location module 11 is used for analyzing and obtaining the device and time period of the human operation that caused the system fault in response to the system fault monitoring software monitoring the system fault;

[0087] The video analysis module 12 is connected to the fault location module 11, and is used to obtain the monitoring video of the computer room where the fault device is located during the occurrence period, and capture the human operation that causes the system fault in the monitoring video.

[0088] In one implementation, the fault location module 11 specifically includes:

[0089] A receiving unit, used for receiving system fault alarm information sent by the system fault monitoring software in response to monitoring a system fault;

[0090] The alarm analysis unit is connected to the receiving unit and is used to use the knowledge graph to perform root cause analysis on the system fault alarm information to locate the device and time period of the human operation that caused the system fault.

[0091] In one embodiment, wherein:

[0092] The knowledge graph is pre-constructed by collecting the association between historical system fault alarm information and historical human operation information, wherein the historical human operation information includes the root cause device causing the historical system fault and the time when the erroneous human operation was performed on the root cause device.

[0093] In one embodiment, the video analysis module 12 specifically includes:

[0094] A grouping unit, configured to process the surveillance video using an inter-frame difference method to group images in the surveillance video;

[0095] The tracing unit is connected to the grouping unit and is used to call the Gaussian mixture model to process the grouped images to obtain the person who has performed manual operations on the generating device.

[0096] In one embodiment, the grouping unit is specifically used for:

[0097] Calculating the inter-frame intensity difference between each two adjacent frames of the surveillance video in sequence;

[0098] Selecting a number of frames whose inter-frame difference strength is higher than a threshold as key frames of the monitoring video image;

[0099] All images in the surveillance video are grouped with the key frame as the starting frame of each group of images.

[0100] In one embodiment, the traceability unit is specifically used to:

[0101] Calling a Gaussian mixture model to process the key frame, and obtaining a first key frame having an active person in the foreground of the image;

[0102] Calling the Gaussian mixture model to process all images of the group where the first key frame is located to obtain the activity trajectory of the active person;

[0103] It is analyzed according to the activity track whether the activity personnel have performed manual operations on the generating device.

[0104] In one embodiment, the device further comprises:

[0105] A log analysis module is used to, in response to capturing the human operation that causes the system failure from the monitoring video, obtain a second time period of the human operation that causes the system failure in the monitoring video, obtain the operation log of the device that caused the failure in the second time period, and obtain the content of the erroneous operation of the person involved in the device that caused the failure based on the operation log.

[0106] Embodiment 3:

[0107] like Figure 3As shown, the present disclosure provides a system for tracing the source of human operation of a computer room fault, the system comprising:

[0108] A device for tracing the source of a machine room fault caused by a human operation, used to implement the method for tracing the source of a machine room fault caused by a human operation as described in Example 1;

[0109] System fault monitoring software, connected to the computer room fault manual operation tracing device and the computer room equipment, is used to monitor whether the computer room equipment has a system fault, and send system fault alarm information to the computer room fault manual operation tracing device after a system fault occurs;

[0110] The computer room monitoring equipment and the computer room fault artificial operation tracing device are used to provide the computer room's monitoring video to the computer room fault artificial operation tracing device.

[0111] In this embodiment, the equipment room includes: Figure 3 The remaining contents of the hardware devices and software devices shown have been described in detail in the embodiments and will not be repeated here.

[0112] Embodiment 4:

[0113] Embodiment 4 of the present disclosure provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the method for tracing the source of artificial operation of a computer room failure as described in Embodiment 1 is implemented, or the device or system for tracing the source of artificial operation of a computer room failure as described in Embodiments 2 and / or 3 is implemented.

[0114] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program elements or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0115] In addition, the present disclosure may also provide a computer device, including a memory and a processor, wherein the memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes the method for tracing the source of artificial operations in a computer room failure as described in Example 1. The computer device may be a device or system for tracing the source of artificial operations in a computer room failure as described in Examples 2 and / or 3.

[0116] The memory is connected to the processor, the memory may be a flash memory or a read-only memory or other memory, and the processor may be a central processing unit or a single-chip microcomputer.

[0117] Embodiments 1-4 of the present disclosure provide a method for tracing the source of human operation of a computer room fault, a device for tracing the source of human operation of a computer room fault, a system for tracing the source of human operation of a computer room fault, and a computer-readable storage medium, which associate computer room environment monitoring with software monitoring, and aim to automatically obtain information about human operation of the corresponding physical equipment in the near future as soon as the alarm of the software monitoring system is generated. It includes: monitoring the alarm of the software system deployed on the equipment in the computer room, associating the fault alarm information with the information of the computer room and the camera; extracting information from the video stream captured by the camera in the computer room, and tracing the source of the human operation.

[0118] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present disclosure, but the present disclosure is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and substance of the present disclosure, and these modifications and improvements are also considered to be within the scope of protection of the present disclosure.

Claims

1. A method for tracing the source of human operation in a computer room failure. It is characterized in that The method comprises: In response to the system fault monitoring software monitoring the system fault, analyzing and obtaining the occurrence device and occurrence time period of the human operation that caused the system fault; Acquire surveillance video of the computer room where the device where the fault occurs is located during the occurrence period, and capture the human operation that causes the system fault in the surveillance video.

2. The method according to claim 1, It is characterized in that In response to the system fault monitoring software monitoring a system fault, analyzing and obtaining the occurrence device and occurrence time period of the human operation that caused the system fault, specifically including: Receiving system fault alarm information sent by system fault monitoring software in response to monitoring a system fault; The knowledge graph is used to perform root cause analysis on the system fault alarm information to locate the device and time period of the human operation that caused the system fault.

3. The method according to claim 2, It is characterized in that in: The knowledge graph is pre-constructed by collecting the association between historical system fault alarm information and historical human operation information, wherein the historical human operation information includes the root cause device causing the historical system fault and the time when the erroneous human operation was performed on the root cause device.

4. The method according to any one of claims 1 to 3, It is characterized in that Capturing the human operation that causes the system failure in the monitoring video specifically includes: Processing the surveillance video using an inter-frame difference method to group images in the surveillance video; The Gaussian mixture model is called to process the grouped images to obtain the personnel who have performed manual operations on the generating device.

5. The method according to claim 4, It is characterized in that Processing the surveillance video using an inter-frame difference method to group images in the surveillance video specifically includes: Calculating the inter-frame intensity difference between each two adjacent frames of the surveillance video in sequence; Selecting a number of frames whose inter-frame difference strength is higher than a threshold as key frames of the monitoring video image; All images in the surveillance video are grouped with the key frame as the starting frame of each group of images.

6. The method according to claim 5, It is characterized in that The Gaussian mixture model is called to process the grouped images to obtain the personnel who have performed manual operations on the generating device, specifically including: Calling a Gaussian mixture model to process the key frame, and obtaining a first key frame having an active person in the foreground of the image; Calling the Gaussian mixture model to process all images of the group where the first key frame is located to obtain the activity trajectory of the active person; It is analyzed according to the activity track whether the activity personnel have performed manual operations on the generating device.

7. The method according to any one of claims 1 to 6, It is characterized in that The method further comprises: In response to capturing the human operation that caused the system failure from the monitoring video, obtain a second time period of the human operation that caused the system failure in the monitoring video, obtain an operation log of the causing device in the second time period, and obtain the content of the erroneous operation of the causing device by the active person based on the operation log.

8. A device for tracing the source of human operation in a machine room failure. It is characterized in that The device comprises: A fault location module is used to analyze and obtain the device and time period of the human operation that caused the system fault in response to the system fault monitoring software monitoring the system fault; The video analysis module is connected to the fault location module and is used to obtain the monitoring video of the computer room where the fault device is located during the occurrence period and capture the human operation that causes the system fault in the monitoring video.

9. A system for tracing the source of human operation in a computer room. It is characterized in that The system comprises: A device for tracing the source of a machine room fault caused by a human operation, used to implement the method for tracing the source of a machine room fault caused by a human operation as described in any one of claims 1 to 7; System fault monitoring software, connected to the computer room fault manual operation tracing device and the computer room equipment, is used to monitor whether the computer room equipment has a system fault, and send system fault alarm information to the computer room fault manual operation tracing device after a system fault occurs; The computer room monitoring equipment and the computer room fault artificial operation tracing device are used to provide the computer room's monitoring video to the computer room fault artificial operation tracing device.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method for tracing the source of a human operation caused by a computer room fault as described in any one of claims 1 to 7 is implemented.

Images