Equipment cluster fault analysis method and device, equipment and storage medium
By constructing a topology matrix and a dependency matrix, a fault link graph is generated, which solves the problems of accuracy and efficiency in equipment cluster fault analysis and enables efficient location and impact analysis of equipment cluster faults.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-22
- Publication Date
- 2026-03-31
AI Technical Summary
In existing technologies, equipment cluster fault analysis relies on human experience, resulting in poor accuracy and efficiency, especially in complex fault situations.
By constructing the original topology matrix of the device set, calculating propagation parameters, establishing a dependency matrix, and generating a fault link diagram based on the current operating data, the degree of dependency between devices and the impact of faults are displayed.
It improves the accuracy and efficiency of equipment cluster fault analysis, clearly showing the first faulty device and its affected related devices, supporting maintenance personnel to quickly locate and resolve problems.
Smart Images
Figure CN121770973A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of fault analysis technology, and in particular to a method, apparatus, device and storage medium for fault analysis of equipment clusters. Background Technology
[0002] The normal operation of a business often requires real-time support from its internal equipment clusters, such as multiple computers or servers located in a company's data center. Occasionally, one or more individual devices in the equipment cluster may fail. To ensure the stability of the business operation, it is necessary to perform timely fault analysis on the equipment cluster to facilitate subsequent maintenance.
[0003] Currently, the common method for fault analysis of equipment clusters is for maintenance personnel to rely on their human experience to diagnose and analyze the equipment clusters. However, due to the strong subjectivity of human experience, the accuracy and efficiency of fault analysis for equipment clusters are both poor when facing complex fault situations. Summary of the Invention
[0004] To improve the accuracy and efficiency of fault analysis for equipment clusters, this application provides a method, apparatus, device, and storage medium for fault analysis of equipment clusters.
[0005] In a first aspect, this application provides a method for analyzing equipment cluster faults, including:
[0006] An initial topology matrix is constructed based on the device set and the connection relationships between devices, and the initial topology matrix is obtained by initializing the initial topology matrix.
[0007] Calculate the propagation parameters corresponding to each topology element in the initial topology matrix to obtain the propagation parameter set;
[0008] Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, a dependency matrix is constructed;
[0009] A fault link diagram is generated based on the current operating data of each device in the device set and the dependency matrix.
[0010] Secondly, this application provides a device for analyzing equipment cluster faults, comprising:
[0011] The matrix generation module is used to construct an original topology matrix based on the device set and the connection relationship between devices, and to initialize the original topology matrix to obtain an initialized topology matrix.
[0012] The parameter calculation module is used to calculate the propagation parameters corresponding to each topology element in the initialization topology matrix, and obtain the propagation parameter set.
[0013] The matrix construction module is used to construct a dependency matrix based on the initial topology matrix, the propagation parameter set, and the preset inter-device dependency relationships;
[0014] The link determination module is used to generate a fault link diagram based on the current operating data of each device in the device set and the dependency matrix.
[0015] Thirdly, this application provides a computer device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method.
[0016] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method.
[0017] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.
[0018] The aforementioned device cluster fault analysis method, apparatus, equipment, and storage medium construct an initial topology matrix based on the device set and the connection relationships between devices, and initialize the initial topology matrix to obtain an initial topology matrix. Propagation parameters corresponding to each topology element in the initial topology matrix are calculated to obtain a propagation parameter set. A dependency matrix is constructed based on the initial topology matrix, the propagation parameter set, and preset inter-device dependencies. A fault link diagram is generated based on the current operating data of each device in the device set and the dependency matrix. Through the above implementation, by calculating the propagation parameters between devices in the device set, the degree of dependency between devices is first determined. This degree of dependency is displayed through the dependency matrix. Then, the current operating data of the devices is used to determine whether a device has failed. When a failure occurs, the current operating data and the dependency matrix determine the fault link diagram formed by the failed devices. This fault link diagram clearly shows the device that first failed and other related devices affected by it, thus effectively improving the accuracy and efficiency of fault analysis for device clusters.
[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description
[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0021] Figure 1 This is a flowchart of a device cluster fault analysis method provided in the embodiments of this application;
[0022] Figure 2 This is a schematic diagram of the structure of a device cluster provided in an embodiment of this application;
[0023] Figure 3 This is a fault link diagram provided in the embodiments of this application;
[0024] Figure 4 This is a schematic diagram of the structure of a device cluster fault analysis device provided in the embodiments of this application;
[0025] Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application;
[0026] Figure 6 This is an internal structural diagram of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation
[0027] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this disclosure.
[0028] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings herein 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 the embodiments 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, apparatus, product, or device 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 devices.
[0029] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.
[0030] Example 1
[0031] Figure 1 This is a flowchart of a device cluster fault analysis method provided in Embodiment 1 of this application, for reference. Figure 1 The method can be executed by a device that performs the method, which can be implemented in software and / or hardware, and the method includes:
[0032] S110. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0033] It should be noted that enterprises often set up device clusters to maintain business operations. A device cluster consists of multiple individual devices, and there may be direct connections between these individual devices. For example, the computers or servers in an enterprise's data center can form a device cluster, and there may be direct connections between these computers or servers. These direct connections can be physical connections between devices, communication connections between microservices, or direct connections between software devices, etc., and are not specifically limited.
[0034] To perform fault analysis on a device cluster, the first step is to determine the direct connections between the devices in the cluster. These direct connections are pre-defined knowledge, but to visualize them and incorporate them into subsequent fault analysis calculations, they need to be quantified.
[0035] Specifically, assume the device set V is { , , , , }, and the equipment With equipment The direct connection relationship between the devices is denoted as If the equipment With equipment If a direct connection exists, then let =1; if the device With equipment If there is no direct connection between them, then let =0; Furthermore, in this embodiment, the direct connection relationship between each device is quantified in the form of a matrix, and this matrix is denoted as the original topology matrix A.
[0036] in, ;
[0037] It should be noted that, All are 0, where the positive integer i ranges from [1, n].
[0038] For example, suppose there are 4 devices in the device cluster, and the device set V = { , , , See also Figure 2 The diagram shows a device connection configuration. Based on this configuration, the original topology matrix A is calculated as follows:
[0039] .
[0040] It should be noted that although the original topology matrix A can quantify the direct connection relationship between devices in the device cluster, subsequent fault analysis focuses more on whether there is an actual data propagation relationship between devices. Therefore, it is necessary to further establish a matrix based on the original topology matrix A to experience the data propagation relationship between devices; for this purpose, the original topology matrix A needs to be initialized.
[0041] Specifically, all elements in the original topological matrix A are set to 0, thereby initializing the original topological matrix A, and the initialized topological matrix A is a matrix of all zeros, denoted as the initialized topological matrix.
[0042] S120. Calculate the propagation parameters corresponding to each topology element in the initialization topology matrix to obtain the propagation parameter set.
[0043] In this process, the elements in the initial topology matrix are denoted as topology elements, which can be used to reflect the corresponding two devices. Taking one topology element as an example, the two devices corresponding to the topology element are denoted as the first device and the second device, respectively. There are one or more data association indicators between the first device and the second device. For example, the data association indicators include, but are not limited to, business association degree and communication quality. By performing corresponding calculations on the data association indicators of the first device and the second device, multiple different data association indicators are unified into one indicator to reflect the data dependency relationship between the two devices, and this indicator is marked as a propagation parameter.
[0044] Specifically, the propagation parameters corresponding to each topology element in the initial topology matrix are calculated respectively. And each propagation parameter The set of propagation parameters is denoted as the propagation parameter set.
[0045] S130. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0046] In this set of propagation parameters, each propagation parameter corresponds one-to-one with each topological element in the initialization topology matrix.
[0047] Specifically, each propagation parameter in the propagation parameter set will replace the corresponding topological element in the initial topological matrix to obtain the initial dependency matrix.
[0048] It should be noted that this embodiment also pre-defines inter-device dependencies, which are used to characterize whether there is a data dependency between any two devices in the device set. If there is a data dependency between two devices, the propagation parameters corresponding to the two devices in the initial dependency matrix remain unchanged. If there is no data dependency between two devices, the propagation parameters corresponding to the two devices in the initial dependency matrix are set to 0, thereby obtaining the target dependency matrix. This dependency matrix can not only quantify whether there is a data dependency between any two devices in the device set, but also, based on the determination that there is a device dependency, further determine the degree of data dependency through the value of the propagation parameters.
[0049] S140. Based on the current operating data of each device in the device set and the dependency matrix, generate a fault link diagram.
[0050] It should be noted that, taking one device in the device set as an example, the device will generate corresponding operating data during operation, and the operating data of the device detected at the current stage will be recorded as the current operating data; for example, the current operating data includes network traffic, CPU utilization and memory utilization, etc., without being specifically limited.
[0051] Among them, by detecting the current operating data, it can be determined whether the corresponding device is operating abnormally, that is, whether a fault has occurred; in addition, the dependency matrix can also identify other devices that have data dependencies on the faulty device. If other devices determine that a fault has occurred through their corresponding current operating data, then multiple faulty devices with data dependencies can form a link, and this link is recorded as a fault link diagram.
[0052] It should be noted that this embodiment constructs an initial topology matrix based on the device set and the connection relationships between devices, and initializes the initial topology matrix to obtain an initial topology matrix; calculates the propagation parameters corresponding to each topology element in the initial topology matrix to obtain a propagation parameter set; constructs a dependency matrix based on the initial topology matrix, the propagation parameter set, and preset inter-device dependencies; and generates a fault link diagram based on the current operating data of each device in the device set and the dependency matrix. Through the above implementation, by calculating the propagation parameters between each device in the device set, the degree of dependency between each device is first determined, and the degree of dependency is displayed through the dependency matrix. Then, the current operating data of the devices is used to determine whether a device has failed. When a failure occurs, the current operating data and the dependency matrix are used to determine the fault link diagram formed by each failed device. This fault link diagram can clearly show the device that first failed, as well as other related devices affected by that device, thus effectively improving the accuracy and efficiency of fault analysis of the device cluster.
[0053] Example 2
[0054] This application provides a device cluster fault analysis method in Embodiment 2, which optimizes the "calculation of propagation parameters corresponding to each topology element in the initialization topology matrix to obtain a propagation parameter set" in Embodiment 1. It should be noted that for parts not detailed in this embodiment, please refer to the descriptions in other embodiments. The method includes:
[0055] S210. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0056] S221. Determine the target device and the device corresponding to each topology element in the initialization topology matrix.
[0057] Taking a topology element in the initialization topology matrix as an example, this topology element corresponds to two devices in the device set, denoted as the target device and the device corresponding to the element, respectively.
[0058] S222. Calculate the service correlation and communication quality between the target device and the element corresponding to each topology element.
[0059] In this example, taking the target device corresponding to a topology element and the device corresponding to the element, there may be corresponding data association indicators between the target device and the device corresponding to the element. In this embodiment, the data association indicators specifically include: service association degree and communication quality.
[0060] Among them, business relevance Used to reflect the target device Corresponding device for elements The strength of interaction between them; the degree of business relevance Specifically, it is the ratio between the number of interactions between two devices detected in the current stage and the preset maximum number of interactions; for example, the target device detected in the current stage Corresponding device for elements The number of interactions between them is 180, and the maximum number of interactions is 200. What is the business relevance? =180 / 200=0.9.
[0061] Among them, communication quality is used to reflect the target device. Corresponding device for elements The degree of communication quality between them; this communication quality degree Based on target device Corresponding device for elements The communication delay τ and packet loss rate b are calculated. The calculation formula is: = (1-b); where the unit of communication delay is ms; for example, the target device Corresponding device for elements If the communication delay τ is 10ms and the packet loss rate is 0.02, then the communication quality score is... = (1-0.02)=0.36.
[0062] S223. Based on the service correlation degree and communication quality corresponding to each of the topology elements, determine the propagation parameters corresponding to each of the topology elements to obtain a propagation parameter set.
[0063] It should be noted that this embodiment aims to determine the degree of data dependency between any two devices in the device set using a single parameter. This single parameter can be obtained through the aforementioned business correlation. With communication quality The calculation yields this parameter, which is then denoted as the propagation parameter. .
[0064] Among them, propagation parameters The calculation formula is as follows:
[0065] =λ +(1-λ) ;
[0066] Wherein, λ is a weight adjustment factor, and the value range of λ is [0,1]. In this embodiment, the weight adjustment factor λ is preferably 0.6.
[0067] It should be noted that the propagation parameters for each topological element can be calculated. And the propagation parameters corresponding to each topological element The set of propagation parameters is denoted as the propagation parameter set.
[0068] S230. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0069] S240. Based on the current operating data of each device in the device set and the dependency matrix, generate a fault link diagram.
[0070] Example 3
[0071] This application provides a device cluster fault analysis method in Embodiment 3, which optimizes the "generating a fault link diagram based on the current operating data of each device in the device set and the dependency matrix" in Embodiment 1. It should be noted that for parts not detailed in this embodiment, please refer to the descriptions in other embodiments. The method includes:
[0072] S310. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0073] S320. Calculate the propagation parameters corresponding to each topology element in the initialization topology matrix to obtain the propagation parameter set.
[0074] S330. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0075] S341. Determine the devices in the device set whose current operating data is abnormal, and obtain the faulty devices; wherein, the current operating data includes at least the current network traffic, current CPU utilization, and current memory usage.
[0076] It should be noted that, taking one device in the device set as an example, the device will generate corresponding running data during the current stage of operation, and this running data will be recorded as the current running data. In this embodiment, the current running data specifically includes: current network traffic, current CPU utilization, and current memory utilization.
[0077] Among them, the current network traffic is the ratio of the sum of the number of bytes sent and received by the network card counter within a preset data collection period of the current stage to the collection period; the current CPU utilization rate is the proportion of time that the CPU uses to execute effective work tasks (instead of idle waiting) within a preset data collection period of the current stage; and the current memory usage rate is the ratio of the amount of memory used by the device to the total amount of memory within a preset data collection period of the current stage.
[0078] Specifically, to determine whether a device is currently abnormal, this embodiment sets corresponding thresholds for current network traffic, current CPU utilization, and current memory utilization. If there is a situation where the current network traffic is greater than the corresponding threshold, the current CPU utilization is greater than the corresponding threshold, or the current memory utilization is greater than the corresponding threshold, then the corresponding device is determined to be an abnormal device and recorded as a faulty device.
[0079] S342. Determine the dependency elements in the dependency matrix that correspond to the faulty device.
[0080] In this step, the faulty device identified in the previous step has a corresponding set of dependency elements in the dependency matrix, and non-zero dependency elements are selected from this set of dependency elements as the dependency elements corresponding to the faulty device.
[0081] It should be noted that the dependency elements corresponding to the faulty device mentioned above are used to reflect other devices that have data dependencies on the faulty device, and these dependencies can be propagated through the dependency elements, i.e., the parameters. The size of the dependency further reflects the degree of dependence; therefore, after identifying the faulty device, the impact of the faulty device on other devices that have dependencies on the data can be effectively determined by the corresponding dependency elements of the faulty device.
[0082] S343. Based on the current operating data of the faulty device and the dependency elements, generate a fault link diagram.
[0083] To analyze device failures in a device cluster, the first step is to identify the device that first failed and the subsequent devices that failed due to the failure of the first device. Based on these two types of devices, a fault link diagram is generated for maintenance personnel to refer to directly. This fault link diagram includes link nodes composed of each faulty device, and the direction of the arrows between adjacent link nodes depends on the data dependency between them. For example, if the first faulty device depends on the second faulty device in terms of data, then the arrow between the first and second faulty devices will point from the second faulty device to the first faulty device.
[0084] Specifically, in order to generate the above-mentioned fault link diagram, this embodiment pre-sets a trained fault analysis model. This fault analysis model is used to analyze and process the current operating data of the faulty device and the corresponding dependency elements, and then outputs the fault link diagram.
[0085] For example, the above analysis and processing procedure is as follows: the fault analysis model determines the degree of impact on several related faulty devices (such as device D1, device D2, device D3, device D4, and device D5) based on the current operating data of the faulty device and the corresponding dependency elements:
[0086] D1, The degree of impact is 100% (fault source);
[0087] D2, The degree of influence is 80%;
[0088] D3, The degree of influence is 70%;
[0089] D4, The degree of influence is 60%;
[0090] D5, The degree of influence is 50%.
[0091] Then, based on the degree of impact on each faulty device, a fault chain diagram is determined, such as... Figure 3 As shown.
[0092] Example 4
[0093] This application provides a device cluster fault analysis method in Embodiment 2, which supplements the steps after "determining the device whose current operating data is abnormal in the device set, and obtaining the faulty device" in Embodiment 3. It should be noted that for parts not detailed in this embodiment, please refer to the descriptions in other embodiments. The method includes:
[0094] S410. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0095] S420. Calculate the propagation parameters corresponding to each topology element in the initialization topology matrix to obtain the propagation parameter set.
[0096] S430. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0097] S441. Determine the devices in the device set whose current operating data is abnormal, and obtain the faulty devices; wherein, the current operating data includes at least the current network traffic, current CPU utilization, and current memory usage.
[0098] S442. Determine the dependency elements in the dependency matrix that correspond to the faulty device.
[0099] S443. Based on the current operating data of the faulty device and the dependency elements, generate a fault link diagram.
[0100] S444. Fit the real-time operating data of the faulty equipment with the prior operating data to obtain the performance change curve.
[0101] It should be noted that the fault chain diagram can provide maintenance personnel with a clear cause-and-effect relationship of the fault, which can help improve the efficiency and accuracy of subsequent fault analysis. This embodiment aims to further improve the efficiency and accuracy of fault analysis by generating deeper fault analysis results, such as fault causes and fault impacts, based on the determination of the real-time operating data corresponding to the faulty device.
[0102] Therefore, the first step is to determine how the performance of the faulty equipment changes in the current time period compared to past time periods.
[0103] Among them, the real-time operating data of the faulty equipment is the operating data corresponding to the current time period (defined from a certain time in the past to the current time). In this embodiment, the operating data of the faulty equipment in a time period between the current time period (referred to as the past time period) is also collected and referred to as the prior operating data.
[0104] To determine the performance change of the faulty equipment in the current period compared with past periods, this embodiment also presets a data fitting function. This data fitting function is used to fit the real-time operating data of the faulty equipment with the previous operating data, thereby obtaining the performance change curve of the faulty equipment.
[0105] S445. Based on the performance change curve and the preset warning value, determine the instantaneous operating data of the target.
[0106] It should be noted that in the past, the faulty equipment had not yet failed and its performance was relatively high. In the current period, due to the failure of the faulty equipment, its performance has begun to decline. If the performance is low, it indicates that the faulty equipment has suffered a significant failure. Therefore, the operating data when the performance is low has a high value for fault analysis.
[0107] To determine the specific moment when performance is low, this embodiment presets a warning value. The moment corresponding to the intersection of the performance change curve and the warning value is the specific moment when performance is low. The operating data collected by the faulty device in the second corresponding to the low performance is recorded as the target instantaneous operating data.
[0108] S446. Analyze the instantaneous operating data of the target to obtain the fault analysis results.
[0109] In this embodiment, the preset fault analysis model can analyze the instantaneous operating data of the target to determine the fault analysis result.
[0110] The fault analysis model can analyze the instantaneous operating data of the target as follows: combining the instantaneous operating data of the target with a Bayesian network to determine the equipment status of the faulty equipment; wherein, the equipment status is used to describe the degree of fault of the faulty equipment; calculating the posterior probability set based on the equipment status; taking the maximum value in the posterior probability set as the root cause; analyzing the root cause to obtain the fault analysis result; wherein, in this embodiment, the fault analysis result includes: fault cause, fault impact, and recommended measures.
[0111] Example 5
[0112] This application provides a device cluster fault analysis method in Embodiment 5, which optimizes the "generating a fault link diagram based on the current operating data of each device in the device set and the dependency matrix" in Embodiment 1. It should be noted that for parts not detailed in this embodiment, please refer to the descriptions in other embodiments. The method includes:
[0113] S510. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0114] S520. Calculate the propagation parameters corresponding to each topology element in the initialization topology matrix to obtain the propagation parameter set.
[0115] S530. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0116] S541. Based on the current operating data of each device in the device set and the abnormal threshold corresponding to each device, determine the health of the operating data of each device.
[0117] It should be noted that, taking one device in the device set as an example, this device can detect the corresponding current operating data. In this embodiment, the current operating data includes the current network traffic. Current CPU utilization and current memory usage In this implementation, the current network traffic is also considered. Current CPU utilization and current memory usage Each has a preset corresponding anomaly threshold, and the current network traffic is... The abnormal threshold is denoted as the network traffic threshold F, and the current CPU utilization is... The abnormal threshold is denoted as the CPU utilization threshold U, and the current memory utilization is also recorded. The abnormal threshold is denoted as the memory usage threshold M.
[0118] By analyzing various current operating data of the device and their corresponding anomaly thresholds, the device's health status, including traffic health status, can be calculated for different operating data points. CPU health and memory health .
[0119] Specifically, traffic health =1- / F; CPU health =1- / U; Memory health =1- / M.
[0120] And the device's traffic health CPU health and memory health This is collectively referred to as the health of operational data.
[0121] S542. Based on the health status of the operating data of each device and the preset coefficient group, determine the comprehensive fault impact value of each device.
[0122] It should be noted that the operational health data of the faulty equipment can be used to calculate the degree of impact of the fault on the equipment, and this degree of impact is recorded as the comprehensive fault impact value Z.
[0123] Among them, the comprehensive fault impact value Z= · + · · ;
[0124] in, , as well as For the preset coefficient group, + =1.
[0125] S543. Based on the comprehensive fault impact value of each device and the dependency matrix, determine the data flow direction of each device.
[0126] It should be noted that the comprehensive failure impact value can be calculated for each faulty device, and the dependency matrix can identify another device that has a data dependency relationship with each device. By measuring the comprehensive failure impact values of the device and the corresponding other device, the data flow direction between the two devices can be determined, where the data flow direction is from the device with the larger comprehensive failure impact value to the device with the smaller comprehensive failure impact value.
[0127] S544. Based on the comprehensive fault impact value of each device and the data flow direction, generate a fault link diagram.
[0128] It should be noted that the corresponding comprehensive fault impact value and data flow direction can be calculated for each faulty device. The data flow direction can identify two devices that are related in terms of data. The device with the larger comprehensive fault impact value is recorded as the initial faulty device, and the device with the smaller comprehensive fault impact value is recorded as the affected faulty device.
[0129] For example, taking a set of initial faulty devices and affected faulty devices corresponding to a faulty device, the initial faulty device may be the fault source device or the affected faulty device of another faulty device; the affected faulty device may also be the initial faulty device of another faulty device; by sorting and connecting the initial faulty devices and affected faulty devices corresponding to different faulty devices, a fault link diagram can be obtained.
[0130] Example 6
[0131] This application provides a device cluster fault analysis method in Embodiment Six, which supplements the steps preceding "determining the health of the operating data of each device based on the current operating data of each device in the device set and the corresponding abnormal threshold of each device" in Embodiment Five. It should be noted that for parts not detailed in this embodiment, please refer to the descriptions in other embodiments. The method includes:
[0132] S610. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0133] S620. Calculate the propagation parameters corresponding to each topology element in the initialization topology matrix to obtain the propagation parameter set.
[0134] S630. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0135] S641. Determine the device type of the devices in the device set, and obtain the historical operation data corresponding to the device type.
[0136] The device type refers to the category to which the device belongs, such as server, router, switch, database, etc. The device type can be determined by information such as the device's model number, function, and purpose. For example, a device with the model number "AR600" belongs to the router type.
[0137] In this embodiment, historical operational data includes historical network traffic, historical CPU utilization, and historical memory usage.
[0138] S642. Filter out abnormal operation data from the historical operation data.
[0139] Specifically, taking one type of historical operational data as an example, such as historical network traffic, this embodiment uses the Z-Score method to filter out abnormal network traffic from historical network traffic as abnormal operational data.
[0140] S643. Calculate the standard deviation of the abnormal operating data to obtain the abnormal threshold corresponding to the device.
[0141] Taking the aforementioned historical network traffic as an example, the standard deviation of each abnormal network traffic obtained by filtering the historical network traffic is used as the abnormal threshold for the device in terms of network traffic operation data; the same applies to CPU utilization and memory utilization, which will not be elaborated here.
[0142] S644. Based on the current operating data of each device in the device set and the corresponding abnormal threshold of each device, determine the health of the operating data of each device.
[0143] S645. Based on the health status of the operating data of each device and the preset coefficient group, determine the comprehensive fault impact value of each device.
[0144] S646. Based on the comprehensive fault impact value of each device and the dependency matrix, determine the data flow direction of each device.
[0145] S647. Based on the comprehensive fault impact value of each device and the data flow direction, generate a fault link diagram.
[0146] Example 7
[0147] This application provides a device cluster fault analysis method in Embodiment 7, which optimizes the "determining the data flow direction of each device based on the comprehensive fault impact value of each device and the dependency matrix" in Embodiment 5. It should be noted that for parts not detailed in this embodiment, please refer to the descriptions in other embodiments. The method includes:
[0148] S710. Construct an original topology matrix based on the device set and the connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix.
[0149] S720. Calculate the propagation parameters corresponding to each topology element in the initialization topology matrix to obtain the propagation parameter set.
[0150] S730. Based on the initialization topology matrix, the propagation parameter set, and the preset inter-device dependencies, construct a dependency matrix.
[0151] S741. Based on the current operating data of each device in the device set and the abnormal threshold corresponding to each device, determine the health of the operating data of each device.
[0152] S742. Based on the health status of the operating data of each device and the preset coefficient group, determine the comprehensive fault impact value of each device.
[0153] S743A. Based on the comprehensive fault impact value of each device and the dependency matrix, determine the source device and dependent device corresponding to each device.
[0154] Each device in the device set has a corresponding comprehensive fault impact value and a corresponding matrix element in the dependency matrix. The matrix element can be used to reflect another device that has a data dependency relationship with the device. The device with the higher comprehensive fault impact value is denoted as the source device, and the other device is denoted as the dependent device. The dependent device depends on the source device in terms of data. If the source device fails, the dependent device will also fail.
[0155] S743B, Determine the first task processing type of the source device and the second task processing type of the corresponding dependent device for each of the aforementioned devices.
[0156] It should be noted that for two devices with a data dependency relationship, the source device and the dependent device, their respective task processing types also have a sequential relationship in the data processing logic. For example, the task processing type of the source device is a data preprocessing task, and the task processing type of the dependent device is a data analysis task. In terms of data processing logic, the data preprocessing task precedes the data analysis task. Therefore, by using the corresponding task processing types of the source device and the dependent device, the actual data flow between the source device and the dependent device can be determined.
[0157] In this context, the task processing type of the source device is designated as the first task processing type, and the task processing type of the dependent device is designated as the second task processing type.
[0158] S743C: Based on the first task processing type and the second task processing type corresponding to each of the devices, determine the data flow direction of each device.
[0159] By analyzing the data processing logic of the first and second task processing types using a pre-defined semantic model, the actual data flow between the source device and the dependent device can be determined.
[0160] S744. Based on the comprehensive fault impact value of each device and the data flow direction, generate a fault link diagram.
[0161] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.
[0162] Example 8
[0163] Based on the same inventive concept, this embodiment also provides an equipment cluster fault analysis device for implementing the equipment cluster fault analysis method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more equipment cluster fault analysis device embodiments provided below can be found in the limitations of the equipment cluster fault analysis method described above, and will not be repeated here.
[0164] In this embodiment, as Figure 4 As shown, a device cluster fault analysis apparatus is provided, comprising:
[0165] The matrix generation module is used to construct an original topology matrix based on the device set and the connection relationship between devices, and to initialize the original topology matrix to obtain an initialized topology matrix.
[0166] The parameter calculation module is used to calculate the propagation parameters corresponding to each topology element in the initialization topology matrix, and obtain the propagation parameter set.
[0167] The matrix construction module is used to construct a dependency matrix based on the initial topology matrix, the propagation parameter set, and the preset inter-device dependency relationships;
[0168] The link determination module is used to generate a fault link diagram based on the current operating data of each device in the device set and the dependency matrix.
[0169] In an optional embodiment, calculating the propagation parameters corresponding to each topological element in the initial topology matrix to obtain a propagation parameter set includes:
[0170] Determine the target device and the device corresponding to each topology element in the initialization topology matrix;
[0171] Calculate the service correlation and communication quality between the target device and the device corresponding to each topology element;
[0172] Based on the service relevance and communication quality corresponding to each of the topology elements, the propagation parameters corresponding to each of the topology elements are determined to obtain a propagation parameter set.
[0173] In an optional embodiment, generating a fault link graph based on the current operating data of each device in the device set and the dependency matrix includes:
[0174] The faulty devices are identified by determining which devices in the set have abnormal current operating data; wherein, the current operating data includes at least current network traffic, current CPU utilization, and current memory usage.
[0175] Determine the dependency elements in the dependency matrix that correspond to the faulty device;
[0176] A fault link diagram is generated based on the current operating data of the faulty device and the dependency elements.
[0177] In an optional embodiment, the device cluster fault analysis apparatus further includes:
[0178] The curve calculation module is used to fit the real-time operating data of the faulty equipment with the prior operating data to obtain the performance change curve;
[0179] The data determination module is used to determine the instantaneous operating data of the target based on the performance change curve and the preset warning value;
[0180] The data analysis module is used to analyze the instantaneous operating data of the target to obtain fault analysis results.
[0181] In an optional embodiment, generating a fault link graph based on the current operating data of each device in the device set and the dependency matrix includes:
[0182] Based on the current operating data of each device in the device set and the corresponding abnormal threshold of each device, the health of the operating data of each device is determined.
[0183] Based on the operational data health status of each device and the preset coefficient set, the comprehensive fault impact value of each device is determined;
[0184] Based on the comprehensive fault impact value of each device and the dependency matrix, the data flow direction of each device is determined;
[0185] A fault link diagram is generated based on the comprehensive fault impact value of each device and the data flow direction.
[0186] In an optional embodiment, before determining the health status of the operating data of each device based on the current operating data of each device in the device set and the corresponding abnormal threshold of each device, the method further includes:
[0187] Determine the device type of the devices in the device set, and obtain the historical operation data corresponding to the device type;
[0188] Filter out abnormal operation data from the historical operation data;
[0189] Calculate the standard deviation of the abnormal operating data to obtain the abnormal threshold corresponding to the device.
[0190] In an optional embodiment, determining the data flow direction of each device based on the comprehensive fault impact value of each device and the dependency matrix includes:
[0191] Based on the comprehensive fault impact value of each device and the dependency matrix, the source device and dependent device corresponding to each device are determined respectively.
[0192] Determine the first task processing type of the source device and the second task processing type of the corresponding dependent device for each of the aforementioned devices;
[0193] Based on the first task processing type and the second task processing type corresponding to each device, the data flow direction of each device is determined.
[0194] Each module in the aforementioned equipment cluster fault analysis device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.
[0195] It should be noted that this embodiment constructs an initial topology matrix based on the device set and the connection relationships between devices, and initializes the initial topology matrix to obtain an initial topology matrix; calculates the propagation parameters corresponding to each topology element in the initial topology matrix to obtain a propagation parameter set; constructs a dependency matrix based on the initial topology matrix, the propagation parameter set, and preset inter-device dependencies; and generates a fault link diagram based on the current operating data of each device in the device set and the dependency matrix. Through the above implementation, by calculating the propagation parameters between each device in the device set, the degree of dependency between each device is first determined, and the degree of dependency is displayed through the dependency matrix. Then, the current operating data of the devices is used to determine whether a device has failed. When a failure occurs, the current operating data and the dependency matrix are used to determine the fault link diagram formed by each failed device. This fault link diagram can clearly show the device that first failed, as well as other related devices affected by that device, thus effectively improving the accuracy and efficiency of fault analysis of the device cluster.
[0196] Example 9
[0197] In this embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows. Figure 5 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for analyzing device cluster faults.
[0198] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present disclosure and does not constitute a limitation on the computer device to which the present disclosure is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0199] Example 10
[0200] In this embodiment, a computer-readable storage medium is provided, such as... Figure 6 As shown, a computer program is stored thereon, and when the computer program is executed by the processor, it implements the steps in the above-described method embodiments.
[0201] Example 11
[0202] In this embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.
[0203] It should be noted that the information collected is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data all comply with the relevant laws, regulations and standards of the relevant countries and regions, necessary confidentiality measures have been taken, and it does not violate public order and good morals. Corresponding operation portals are provided for users to choose to authorize or refuse.
[0204] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this disclosure can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this disclosure may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this disclosure may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.
[0205] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0206] The embodiments described above are merely illustrative of several implementations of this disclosure, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent disclosure. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this disclosure, and these all fall within the protection scope of this disclosure. Therefore, the protection scope of this disclosure should be determined by the appended claims.
Claims
1. A method of device cluster failure analysis, the method comprising: The method comprises the following steps: constructing an original topology matrix based on a device set and a connection relationship between devices, initializing the original topology matrix to obtain an initialized topology matrix; calculating a propagation parameter corresponding to each topology element in the initialized topology matrix to obtain a propagation parameter set; constructing a dependency relationship matrix based on the initialized topology matrix, the propagation parameter set, and a preset dependency relationship between devices; generating a fault link graph based on current running data of each device in the device set and the dependency relationship matrix.
2. The method of claim 1, wherein, The method comprises the following steps: determining a target device corresponding to each topology element in the initialized topology matrix and an element corresponding device; calculating a business correlation degree and a communication quality between the target device and the element corresponding device corresponding to each topology element, respectively; determining a propagation parameter corresponding to each topology element based on the business correlation degree and the communication quality corresponding to each topology element to obtain a propagation parameter set.
3. The method of claim 1, wherein, The method comprises the following steps: determining a device with abnormal current running data in the device set to obtain a fault device; wherein the current running data at least includes current network traffic, current CPU utilization and current memory usage; determining a dependency relationship element corresponding to the fault device in the dependency relationship matrix; generating a fault link graph based on the current running data of the fault device and the dependency relationship element.
4. The method of claim 3, wherein, After determining the device with abnormal current running data in the device set to obtain the fault device, the method further comprises the following steps: fitting real-time running data and previous running data of the fault device to obtain a performance change curve; determining a target instantaneous running data based on the performance change curve and a preset warning value; analyzing the target instantaneous running data to obtain a fault analysis result.
5. The method of claim 1, wherein, The method comprises the following steps: determining a running data health degree of each device in the device set based on current running data of each device in the device set and an abnormal threshold value corresponding to each device; determining a comprehensive fault influence value of each device based on the running data health degree of each device and a preset coefficient group; determining a data flow direction of each device based on the comprehensive fault influence value of each device and the dependency relationship matrix; generating a fault link graph based on the comprehensive fault influence value of each device and the data flow direction.
6. The method of claim 5, wherein, Before determining the running data health degree of each device in the device set based on the current running data of each device in the device set and the abnormal threshold value corresponding to each device, the method further comprises the following steps: determining a device type of a device in the device set to obtain historical running data corresponding to the device type; filtering abnormal running data from the historical running data; calculating a standard deviation of the abnormal running data to obtain an abnormal threshold value corresponding to the device.
7. The method of claim 5, wherein, The data flow direction of each device is determined based on the comprehensive failure influence value of each device and the dependency relationship matrix, including: The source device and the dependent device corresponding to each device are determined respectively based on the comprehensive failure influence value of each device and the dependency relationship matrix; The first task processing type of the source device corresponding to each device and the second task processing type of the dependent device corresponding to each device are determined respectively; The data flow direction of each device is determined based on the first task processing type and the second task processing type corresponding to each device.
8. An apparatus cluster failure analysis device, comprising: The device comprises: A matrix generation module configured to construct an original topology matrix based on a device set and a connection relationship between devices, and initialize the original topology matrix to obtain an initialized topology matrix; A parameter calculation module configured to calculate a propagation parameter corresponding to each topology element in the initialized topology matrix to obtain a propagation parameter set; A matrix construction module configured to construct a dependency relationship matrix based on the initialized topology matrix, the propagation parameter set, and a preset dependency relationship between devices; A link determination module configured to generate a failure link diagram based on current running data of each device in the device set and the dependency relationship matrix. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 7.
10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 7.