Data center storage backup fault monitoring method, device, equipment and medium

Abstract

The invention relates to the field of data processing, in particular to a data center storage backup fault monitoring method and device, equipment and a medium. The method comprises the steps that equipment state data and current monitoring data of a storage backup system are collected in real time, whether the system is in a normal state or not is dynamically judged, if abnormity is detected, a fault type is distinguished as an equipment fault or a data fault, and for the data fault, a fault frame message is automatically captured and recorded; and determining a repair mode corresponding to the fault frame message, executing a repair operation based on the repair mode, and verifying a repair result. And if the repair is successful, generating a standardized report containing fault time, fault type, repair mode and result, and pushing the standardized report to the operation and maintenance terminal to realize closed-loop management and traceability of fault processing, and meanwhile, through automatic monitoring, intelligent diagnosis and standardized repair, the fault processing efficiency is improved. And the fault response efficiency of the data center storage backup system and the reliability of backup data are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, and in particular to a data center storage backup fault monitoring method, device, equipment and medium. BACKGROUND

[0002] With the acceleration of digital transformation, data centers have become the core of enterprise business data storage and processing. In order to protect data security and business continuity, storage backup systems need to periodically backup massive data in full / incremental mode, and support fast recovery (such as RTO<4 hours, RPO<1 hour). Backup scenarios cover structured data (such as databases), unstructured data (such as videos, logs), and hybrid cloud / local architecture, and need to meet compliance requirements (such as long-term data retention in the financial and medical industries). Under this background, the stability and fault handling capability of the storage backup system become a key challenge.

[0003] Currently, the fault monitoring of the data center storage backup system mainly relies on manual inspection. Specifically, the operation and maintenance personnel need to manually check the device status indicator, analyze backup logs, and troubleshoot network transmission records, etc., to determine the fault type by experience and take repair measures. However, this manual monitoring method mainly relies on human experience, and the fault misjudgment rate is high, resulting in low reliability of backup data. SUMMARY

[0004] In order to improve the reliability of backup data, the present application provides a data center storage backup fault monitoring method, device, equipment and medium.

[0005] In a first aspect, the present application provides a data center storage backup fault monitoring method, which adopts the following technical solution: A data center storage backup fault monitoring method, comprising: Collecting device state data and obtaining current monitoring data, wherein the device state data is the state data of the devices corresponding to the data center storage backup system; Determining whether the data center storage backup system is in a normal state based on the device state data and the current monitoring data; If the data center storage backup system is not in a normal state, determining the fault type of the data center storage backup system, wherein the fault type is a device fault or a data fault; When the fault type is a data fault, recording the fault frame packet corresponding to the current monitoring data, and determining the repair method corresponding to the fault frame packet; Repairing the fault frame packet based on the repair method, and determining whether the fault frame packet is repaired successfully; If the repair is successful, a repair report corresponding to the current monitoring data is generated and sent to the corresponding operation and maintenance terminal.

[0006] By adopting the technical solutions, the device state data and the current monitoring data of the storage backup system are collected in real time, and it is dynamically determined whether the system is in a normal state. If an exception is detected, the fault type is distinguished as a device fault or a data fault. For a data fault, the fault frame packet is automatically captured and recorded, and the repair method corresponding to the fault frame packet is determined to perform a repair operation based on the repair method and verify the repair result. If the repair is successful, a standardized report containing the fault time, type, repair method and result is generated and pushed to the operation and maintenance terminal, realizing closed-loop management and traceability of fault processing. At the same time, through automatic monitoring, intelligent diagnosis and standardized repair, the fault response efficiency of the data center storage backup system and the reliability of the backup data are significantly improved.

[0007] In a possible implementation, the device state data includes device temperature, device rotation speed and device vibration signal, and the current monitoring data includes running log of backup software and network communication data. The method comprises: The device vibration signal is processed by a CNN subnetwork to obtain a spatial feature vector of the device, and the device temperature and the device rotation speed are processed by an LSTM subnetwork to obtain a time sequence feature vector; The spatial feature vector and the time sequence feature vector are input into an encoder, the inter-modal correlation weight is calculated by a multi-head attention mechanism, and a fusion feature vector is generated; The fusion feature vector is input into a fully connected classification network, and a device health score is obtained as output; Based on the device health score, it is determined whether the data center storage backup system is in a normal state. If it is determined that the data center storage backup system is not in a normal state, it is determined that the data center storage backup system is not in a normal state. If it is determined that the data center storage backup system is in a normal state, it is determined whether the data center storage backup system is in a normal state based on the current monitoring data.

[0008] In a possible implementation, based on the current monitoring data, it is determined whether the data center storage backup system is in a normal state, comprising: The historical fault data includes historical fault frame packets and historical network communication data corresponding to each historical fault frame. performing time domain analysis and frequency domain analysis on the historical failure data to obtain a first time domain relationship between historical failure frame packets and abnormal failure forms, a second time domain relationship between historical failure frames and abnormal network communication data, and a frequency domain feature vector matrix corresponding to the historical failure data; determining abnormal packet forms and abnormal network communication data forms corresponding to the historical failure frames based on the first time domain relationship, the second time domain relationship, and the frequency domain feature vector matrix; comparing the current monitoring data with the abnormal packet forms and the abnormal network communication data forms respectively to determine whether the current monitoring data has abnormal packet forms or / and abnormal network communication data forms; determining that the data center storage backup system is not in a normal state when the current monitoring data has abnormal packet forms or / and abnormal network communication data forms.

[0009] In a possible implementation, performing time domain analysis and frequency domain analysis on the historical failure data to obtain a first time domain relationship between historical failure frame packets and abnormal failure forms, a second time domain relationship between historical failure frames and abnormal network communication data, and a frequency domain feature vector matrix corresponding to the historical failure data includes: performing sliding window segmentation on the historical failure data, and analyzing abnormal failure forms corresponding to each historical failure packet in each historical window; determining a first time domain relationship between historical failure packets and abnormal failure forms through correlation analysis based on abnormal failure forms corresponding to each historical failure packet in each historical window; determining abnormal network communication data in each historical window, calculating cross-correlation of historical window sequences, identifying time sequence coupling relationship between historical failure frames and abnormal network communication data, and forming a second time domain relationship; performing frequency domain analysis on the abnormal failure forms and the abnormal network communication data in the historical failure data respectively to obtain a first frequency domain feature vector corresponding to the abnormal failure forms and a second frequency domain feature vector corresponding to the abnormal network communication data; integrating the first frequency domain feature vector and the second frequency domain feature vector to obtain a frequency domain feature vector matrix corresponding to the historical failure data.

[0010] In a possible implementation, the historical failure data further includes historical failure frame data, and the determining of the repair mode corresponding to the failure frame packet includes: analyzing the fault frame packet, extracting a key feature vector, and performing similarity comparison between the key feature vector and a historical feature vector corresponding to historical fault frame data to obtain a comprehensive similarity, the historical fault frame data including the historical feature vector and a historical repair method; When there is a comprehensive similarity greater than a first similarity threshold, the comprehensive similarity greater than the first similarity threshold is determined as a first target similarity, and a historical repair method corresponding to the first target similarity is obtained as a repair method corresponding to the fault frame packet; When there is no comprehensive similarity greater than a first similarity threshold, and there are multiple comprehensive similarities greater than a second similarity threshold, the comprehensive similarity greater than the second similarity threshold is determined as a second target similarity, and based on a random forest algorithm, a plurality of historical repair methods corresponding to the second target similarities are combined to obtain an optimal strategy combination, and the optimal strategy combination is determined as the repair method corresponding to the fault frame packet, wherein the second similarity threshold is less than the first similarity threshold; When none of the comprehensive similarities is greater than the second similarity threshold, a fault propagation graph model is constructed based on each historical repair method, the key feature vector is input into the fault propagation graph model, and an expected success degree of each repair method output by the fault propagation graph model is obtained, and a repair method with the highest expected success degree is selected as the repair method corresponding to the fault frame packet.

[0011] In a possible implementation, repairing the fault frame packet based on the repair method includes: determining whether the repair method meets an automatic repair condition; If the repair method meets the automatic repair condition, a repair instruction is generated based on the repair method, and the repair instruction is sent to a repair end corresponding to the fault frame packet; If the repair method does not meet the automatic repair condition, a basic repair method is obtained, a basic repair instruction corresponding to the basic repair method is generated, and the basic repair instruction is sent to the repair end corresponding to the fault frame packet.

[0012] In a possible implementation, when the repair method meets the automatic repair condition, the determination of whether the fault frame packet is successfully repaired includes: an automatic collection of a repair completion signal, and after the repair completion signal is collected, repaired repair monitoring data and a repaired fault frame packet are obtained; comparison of the repair monitoring data with the abnormal packet form and the abnormal network communication data form to determine whether the repair monitoring data has an abnormal packet form or / and an abnormal network communication data form; If it is determined that the repair monitoring data does not exist in the form of an abnormal message or / and the form of abnormal network communication data, the repair fault frame message is parsed to obtain a repair key feature vector, and based on the historical feature vector, it is determined whether the fault frame message is successfully repaired.

[0013] In a second aspect, the present application provides a data center storage backup fault monitoring device, which adopts the following technical solution: A data center storage backup fault monitoring device comprises: A collection module is configured to collect device state data and obtain current monitoring data, wherein the device state data is state data of a device corresponding to a data center storage backup system. A state determination module is configured to determine whether the data center storage backup system is in a normal state based on the device state data and the current monitoring data. A type determination module is configured to determine a fault type of the data center storage backup system if the data center storage backup system is not in the normal state, wherein the fault type is a device fault or a data fault. A recording module is configured to record a fault frame message corresponding to the current monitoring data and determine a repair method corresponding to the fault frame message if the fault type is the data fault. A repair module is configured to repair the fault frame message based on the repair method and determine whether the fault frame message is successfully repaired. A sending module is configured to generate a repair report corresponding to the current monitoring data and send the repair report to a corresponding operation and maintenance terminal if the repair is successful.

[0014] In a third aspect, the present application provides an electronic device, which adopts the following technical solution: An electronic device comprises: At least one processor; A memory; At least one application program, wherein the at least one application program is stored in the memory and is configured to be executed by the at least one processor, and the at least one application program is configured to execute the method of any one of the first aspect.

[0015] In a fourth aspect, the present application provides a computer readable storage medium, which adopts the following technical solution: A computer readable storage medium comprises a computer program capable of being loaded by a processor and executing the method of any one of the first aspect.

[0016] In summary, the present application has the following beneficial technical effects: By collecting and storing the device state data and the current monitoring data of the storage backup system in real time, and dynamically judging whether the system is in a normal state, if an exception is detected, the fault type is distinguished as a device fault or a data fault, for the data fault, the fault frame packet is automatically captured and recorded, and the repair mode corresponding to the fault frame packet is determined, so as to perform a repair operation based on the repair mode, and verify the repair result. If the repair is successful, a standardized report containing the fault time, type, repair mode and result is generated and pushed to the operation and maintenance terminal, realizing closed-loop management and traceability of fault handling, and through automatic monitoring, intelligent diagnosis and standardized repair, the fault response efficiency of the storage backup system of the data center and the reliability of the backup data are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a flowchart of a data center storage backup fault monitoring method provided by an embodiment of the present application; Figure 2 is a block diagram of a data center storage backup fault monitoring device provided by an embodiment of the present application; Figure 3 is a schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be described in detail with reference to the accompanying drawings of the embodiments of the present application. Figure 1 to Figure 3 The present application will be described in further detail.

[0019] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme of the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0020] The present application provides a data center storage backup fault monitoring method, as shown in Figure 1 The method provided in the embodiments of the present application is executed by an electronic device, which can be a server or a terminal device. The server can be a standalone physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc., but is not limited thereto. The terminal device and the server can be directly or indirectly connected through wired or wireless communication, which is not limited in the embodiments of the present application. The method includes steps S101-S106, wherein: Step S101, collect device state data and obtain current monitoring data.

[0021] The device state data is the state data of the devices (such as hard disks, tape drives, switches, etc.) corresponding to the data center storage backup system. Specifically, the device state data can include device temperature, rotation speed, vibration signal, voltage and current, etc. physical indicators, reflecting the health of the device hardware.

[0022] The current monitoring data is related to the operation of the storage backup system, covering the operation log of the backup software (recording the execution of the backup task, error information, etc.), network communication data (such as transmission bandwidth, delay, packet loss rate), and data verification results, etc. for monitoring system software and data transmission state.

[0023] Specifically, the electronic device collects device state data in real time through sensors (such as temperature sensors, vibration sensors) deployed on the storage backup device; at the same time, it interacts with the backup software and network equipment through data collection programs to obtain real-time operation logs and network communication data to obtain current monitoring data.

[0024] Step S102, based on the device state data and the current monitoring data, determine whether the data center storage backup system is in a normal state.

[0025] The normal state means that the devices, software and data transmission of the data center storage backup system are running normally as expected, without hardware failure, software error or data loss, etc.

[0026] After obtaining the device state data and the current monitoring data, a pre-trained machine learning model can be used to analyze the device state data and the current monitoring data. Specifically, the convolutional neural network (CNN) can be used to process the device vibration signal and extract spatial features; the long short-term memory network (LSTM) can be used to analyze the time sequence change of the device temperature and rotation speed. The extracted features are fused and input into a classification model to calculate the health score of the system, which is compared with the preset normal threshold to determine whether the system is in a normal state.

[0027] Step S103, if the data center storage backup system is not in a normal state, determine the fault type of the data center storage backup system.

[0028] The fault type is a device fault or a data fault. The device fault is an abnormality of the hardware device (such as hard disk damage, fan stall, controller failure) in the storage backup system, which causes the system to fail to work normally. The data fault is a problem in the storage, transmission or processing of the backup data, such as data loss, damage, verification error, or incomplete execution of the backup task resulting in incomplete data.

[0029] When the electronic device determines that the system is in an abnormal state, the characteristics of the abnormal data are further analyzed. If the hardware-related indicators (such as excessive temperature and abnormal speed) in the device state data exceed the normal range, it is determined that the device is faulty; if there are backup data verification errors, backup task failure records, and other situations in the current monitoring data, it is determined that the data is faulty.

[0030] Further, if it is determined that the data center storage backup system is in a normal state, no further operation is performed.

[0031] Step S104, when the fault type is a data fault, the fault frame message corresponding to the current monitoring data is recorded, and the repair method corresponding to the fault frame message is determined.

[0032] The fault frame message is a data packet or data segment containing error information generated due to data faults during data transmission or storage, which records the data state and related parameters at the time of fault occurrence.

[0033] After determining that the fault type is a data fault, the complete data frame message related to the fault can be extracted from the current monitoring data, including error code, data content, transmission time, and other information, and stored in the fault log. According to the characteristics of the fault frame message (such as error type and data range involved), the corresponding repair method is matched in the pre-defined repair strategy library. For example, if it is a verification error, the data is re-acquired from the nearest complete backup copy.

[0034] Further, when the fault type is a device fault, an alarm message containing device fault information is generated, and the alarm message is sent to the terminal device of the operation and maintenance personnel corresponding to the data center storage backup system to remind the operation and maintenance personnel to repair the device.

[0035] Step S105, based on the repair method, the fault frame message is repaired, and it is determined whether the repair of the fault frame message is successful.

[0036] After obtaining the repair method corresponding to the fault frame message, the repair operation can be performed according to the determined repair method. Specifically, a data retransmission request is sent to the data source, or the correct data is read from the backup storage to overwrite the faulty data. After the repair is completed, the repaired data is verified again (such as comparing the hash value with the hash value of the original correct data), if the verification is passed, it is determined that the repair is successful; if the verification is not passed, it is considered that the repair fails.

[0037] Step S106, if the repair is successful, a repair report corresponding to the current monitoring data is generated, and the repair report is sent to the corresponding operation and maintenance terminal.

[0038] The repair report is a document recording the data fault repair process and result, containing information such as fault occurrence time, fault type, repair method, repair time consumption, data state after repair, and the like, and is used for the operation and maintenance personnel to understand the fault processing situation.

[0039] The operation and maintenance terminal is a device used by the operation and maintenance personnel to receive fault information and manage the system, such as an operation and maintenance workstation, a mobile terminal, and the like, and facilitates the operation and maintenance personnel to timely master the system state and perform processing.

[0040] After confirming the successful repair, a structured repair report is automatically generated, which summarizes the key information of the fault processing. Then, the repair report is sent to the pre-set operation and maintenance terminal through a network communication protocol (such as HTTP, SMTP), such as being sent to the mailbox of the operation and maintenance personnel in the form of an email, or being pushed to the operation and maintenance management platform to remind the operation and maintenance personnel that the fault has been successfully repaired.

[0041] Further, if the repair fails, an alarm signal of the current monitoring data is generated, and the alarm signal is sent to the corresponding operation and maintenance terminal.

[0042] The embodiment of the present application provides a data center storage backup fault monitoring method, by collecting device state data and current monitoring data of a storage backup system in real time, and dynamically judging whether the system is in a normal state, if an exception is detected, distinguishing the fault type as a device fault or a data fault, for the data fault, automatically capturing and recording the fault frame packet, and determining the repair method corresponding to the fault frame packet, to perform a repair operation based on the repair method, and verify the repair result. If the repair is successful, a standardized report containing the fault time, type, repair method and result is generated, and is pushed to the operation and maintenance terminal, realizing closed-loop management and traceability of fault processing, at the same time, through automatic monitoring, intelligent diagnosis and standardized repair, the fault response efficiency of the data center storage backup system and the reliability of the backup data are significantly improved.

[0043] In a possible implementation manner of the embodiment of the present application, in the step S102, the device state data includes device temperature, device rotating speed and device vibration signal, and the current monitoring data includes running log of backup software and network communication data, The determination of whether the data center storage backup system is in a normal state based on the device state data and the current monitoring data includes: The device vibration signal is processed through the CNN sub-network to obtain a spatial feature vector of the device, and the device temperature and the device rotating speed are processed through the LSTM sub-network to obtain a time sequence feature vector; The spatial feature vector and the time sequence feature vector are input into an encoder, the inter-modal correlation weight is calculated through a multi-head attention mechanism, and a fusion feature vector is generated; Input the fusion feature vector into a fully connected classification network, and obtain a device health score output; Based on the device health score, determine whether the data center storage backup system is in a normal state; If it is determined that the data center storage backup system is not in a normal state, it is determined that the data center storage backup system is not in a normal state; If it is determined that the data center storage backup system is in a normal state, based on the current monitoring data, determine whether the data center storage backup system is in a normal state.

[0044] Among them, the equipment vibration signal is the signal formed by the mechanical vibration generated by the equipment in the running process, and different equipment failures (such as bearing wear and poor gear engagement) will produce specific mode vibration signals, which are important basis for reflecting the running state of the equipment.

[0045] After obtaining the equipment vibration signal collected from the sensor, the equipment vibration signal can be preprocessed, such as filtering, normalization, etc., to remove noise interference and unify the data format. Subsequently, the processed vibration signal is input into the pre-trained CNN sub-network model, and the CNN sub-network model automatically extracts the local features in the vibration signal through the convolution layer, and then reduces the data dimension through the pooling layer, and finally outputs a spatial feature vector that can reflect the spatial structure of the vibration signal. At the same time, the device temperature sequence and the device speed sequence corresponding to the current period are obtained, and the time sequence data of the device temperature and speed are arranged and input into the LSTM sub-network in time sequence. The LSTM sub-network learns and extracts the change trend, periodicity and other features of the temperature and speed data in the time dimension by virtue of its unique memory unit, and outputs a vector containing time sequence features.

[0046] Further, the obtained spatial feature vector and time sequence feature vector are input into the encoder module. The encoder uses a multi-head attention mechanism to analyze the two feature vectors from multiple different angles. In the calculation process, the correlation degree between the device vibration signal features and the device temperature and speed features is calculated respectively through matrix operation and similarity measurement, and multiple attention weights are obtained. These weights reflect the relative importance of different modal data in describing the state of the equipment. Finally, merging and weighted summation are performed to generate a fusion feature vector that combines spatial and time sequence information, so that the vector can more comprehensively reflect the running state of the equipment.

[0047] Further, the generated fusion feature vector is directly input into a fully connected classification network. The neurons in each layer of the fully connected classification network sequentially calculate and transform the input fusion feature vector, and through multiple layers of linear and nonlinear operations, map the high-dimensional feature vector to a specific-dimensional output space. Finally, the network outputs a numerical value, i.e., a device health score. The score is based on the device multi-aspect state information contained in the fusion feature vector, and is obtained through the mapping relationship learned by the network, and can directly reflect the current health degree of the device. A normal threshold range of the device health score is obtained in advance, for example, the score is in the normal state when the score is 80-100. After obtaining the device health score, it is compared with the set threshold. If the score is within the normal threshold range, it is determined that the data center storage backup system is in a normal state; if the score is lower than the lower limit of the normal threshold, it is determined that the system is not in a normal state.

[0048] When it is determined that the data center storage backup system is not in a normal state, it is directly and explicitly determined that the system has an exception. If it is determined that the system is in a normal state, further analysis is performed on the current monitoring data such as the running log of the backup software and network communication data, so as to determine whether the data center storage backup system is truly in a normal running state.

[0049] In one possible implementation of the embodiment of the present application, based on the current monitoring data, it is determined whether the data center storage backup system is in a normal state, comprising: Obtaining historical fault data, the historical fault data including historical fault frame packets and historical network communication data corresponding to each historical fault frame; Performing time domain analysis and frequency domain analysis on the historical fault data to obtain a first time domain relationship between the historical fault frame packets and abnormal fault forms, a second time domain relationship between the historical fault frames and abnormal network communication data, and a frequency domain feature vector matrix corresponding to the historical fault data; Based on the first time domain relationship, the second time domain relationship, and the frequency domain feature vector matrix, determining the abnormal packet form and the abnormal network communication data form corresponding to the historical fault frame; Comparing the current monitoring data with the abnormal packet form and the abnormal network communication data form respectively to determine whether the current monitoring data has the abnormal packet form or / and the abnormal network communication data form; When the current monitoring data has the abnormal packet form or / and the abnormal network communication data form, it is determined that the data center storage backup system is not in a normal state.

[0050] The historical fault records corresponding to the current data center storage backup system are retrieved from the data center storage backup database, the historical fault frame packets and historical network communication data corresponding to each historical fault event are extracted, and the historical fault data set is constructed by classifying and arranging according to the timestamp and fault type. Further, the historical fault frame packets are segmented by sliding window, the statistical features (such as error code frequency, peak intensity) in each window are calculated, and the time sequence association with the fault form is analyzed. For example, if a certain type of error code frequently appears within 5 minutes before the fault, a corresponding relationship is established. The historical network communication data (such as delay curve) is subjected to fast Fourier transform (FFT), the main frequency components and energy distribution are extracted, and a frequency domain feature vector matrix is constructed. For example, the characteristic frequency of periodic delay fluctuation is identified.

[0051] Based on the first time domain relationship and the second time domain relationship, the historical fault frame packets are subjected to cluster analysis, common error code combinations, data damage patterns, etc. are identified, and abnormal packet form templates are formed. At the same time, through the frequency domain feature vector matrix, the frequency feature mode (such as high-frequency jitter, low-frequency oscillation) of abnormal network communication data is determined, and it is associated with a specific fault type.

[0052] The fault frame packets in the current monitoring data are subjected to pattern matching with the abnormal packet form templates, for example, whether a specific error code sequence or data segment damage feature is checked. At the same time, the current network communication data is subjected to time domain and frequency domain analysis, the features are extracted and compared with the frequency feature mode of the abnormal network communication data form, and the similarity is calculated based on cosine similarity or Euclidean distance. If there is a fault frame in the current monitoring data that matches the abnormal packet form, or the frequency domain features of the network communication data are highly similar to the abnormal network communication data form (the similarity exceeds a preset threshold, such as 80%), it is determined that the data center storage backup system is not in a normal state.

[0053] One possible implementation of the embodiments of the present application, time domain analysis and frequency domain analysis are performed on the historical fault data, the first time domain relationship between the historical fault frame packets and the abnormal fault form, the second time domain relationship between the historical fault frame and the abnormal network communication data, and the frequency domain feature vector matrix corresponding to the historical fault data are obtained, including: The historical fault data is segmented by sliding window, and the abnormal fault form corresponding to each historical fault packet in each historical window is parsed; Based on the abnormal fault form corresponding to each historical fault packet in each historical window, the first time domain relationship between the historical fault packet and the abnormal fault form is determined through association analysis; The abnormal network communication data in each historical window is determined, and the cross-correlation of the historical window sequence is calculated to identify the time sequence coupling relationship between the historical fault frame and the abnormal network communication data, and the second time domain relationship is formed; The abnormal failure form in the historical failure data and the abnormal network communication data are respectively subjected to frequency domain analysis to obtain a first frequency domain feature vector corresponding to the abnormal failure form and a second frequency domain feature vector corresponding to the abnormal network communication data. The first frequency domain feature vector and the second frequency domain feature vector are integrated to obtain a frequency domain feature vector matrix corresponding to the historical failure data.

[0054] Specifically, the size (such as 30 seconds) and the step (such as 10 seconds) of a set sliding window can be acquired, and the historical failure data is subjected to time sequence cutting. In each window, the key fields (such as error code, timestamp, and device ID) of the failure message are analyzed and matched with a preset failure classification dictionary to automatically label the corresponding abnormal failure form (for example, the error code "ERR-003" corresponds to "disk read / write error") of each message. Further, the failure messages in each historical window are aggregated and counted to calculate the occurrence frequency and time distribution characteristics (such as the first occurrence time and peak time) of different abnormal failure forms, and these statistical rules are sorted into a rule library to form a first time domain relationship.

[0055] The network communication data (such as the delay value per second) is extracted from the historical window and compared with a preset threshold to mark abnormal points. Then, for each failure frame, the time difference with the network abnormality within 5 minutes before and after is calculated, the pattern is identified through cross-correlation analysis, the time sequence coupling relationship of "network jitter before failure" is formed, and a second time domain relationship is obtained. Further, the abnormal failure form in the historical failure data is subjected to time domain signal reconstruction (such as converting the error code sequence into a numerical sequence), and the fast Fourier transform (FFT) is applied to convert to the frequency domain, the main frequency component and its energy are extracted, and a first frequency domain feature vector is formed. At the same time, the same processing is performed on the abnormal network communication data (such as the delay curve) to obtain a second frequency domain feature vector. The first frequency domain feature vector and the second frequency domain feature vector are aligned according to the fault type and the timestamp, and are merged into a unified feature matrix. For example, for each historical failure sample, the front half of the matrix stores the frequency domain features of the failure form, and the back half stores the frequency domain features of the network abnormality. Each row of the frequency domain feature vector matrix corresponds to a failure sample, and each column corresponds to a frequency domain feature dimension.

[0056] In a possible implementation of the embodiment of the application, the historical failure data further includes historical failure frame data, and the repair mode corresponding to the failure frame message is determined, including: The failure frame message is analyzed, the key feature vector is extracted, and the key feature vector is compared with the historical feature vector corresponding to the historical failure frame data in terms of similarity to obtain a comprehensive similarity, the historical failure frame data including the historical feature vector and the historical repair mode; When the comprehensive similarity greater than the first similarity threshold exists, the comprehensive similarity greater than the first similarity threshold is determined as the first target similarity, and the historical repair mode corresponding to the first target similarity is obtained as the repair mode corresponding to the fault frame packet; When the comprehensive similarity greater than the first similarity threshold does not exist, and the comprehensive similarity greater than the second similarity threshold exists, the comprehensive similarity greater than the second similarity threshold is determined as the second target similarity, and the historical repair modes corresponding to the second target similarity are combined based on the random forest algorithm to obtain an optimal strategy combination, and the optimal strategy combination is determined as the repair mode corresponding to the fault frame packet, wherein the second similarity threshold is less than the first similarity threshold. When the comprehensive similarity is not greater than the second similarity threshold, a fault propagation graph model is constructed based on each historical repair mode, the key feature vector is input into the fault propagation graph model, and the expected success degrees of different repair modes output by the fault propagation graph model are obtained, and the repair mode with the highest expected success degree is selected as the repair mode corresponding to the fault frame packet.

[0057] The key feature vector is a feature set capable of representing the fault type, such as an error code, a data segment damage position, a timestamp, etc.

[0058] Specifically, the fault frame packet is parsed, and key features (such as an error code sequence, a data check value, and a transmission protocol version) are extracted and converted into a standardized vector. Further, the historical fault frame database is traversed, and the similarity between the feature vector of each historical record and the current vector is calculated. For example, the cosine value of the vector angle is calculated to obtain a similarity score between 0 and 1, and finally a similarity list of the current fault and all historical faults is generated. It should be noted that the manner of extracting and parsing the packet features is well known, and the embodiments of the present application are not limited thereto.

[0059] Further, after obtaining the comprehensive similarity, records greater than the first threshold in the similarity list are screened. If such records exist, the repair mode of the corresponding fault is directly extracted from the historical database (for example, when the comprehensive similarity is 0.95, the historical repair mode is “re-synchronize backup task”), and the repair mode is determined as the repair scheme of the current fault.

[0060] When there is no high similarity match, the historical fault records with a comprehensive similarity between the first similarity threshold and the second similarity threshold are screened out. For each record, its repair method (such as "cleaning cache" and "adjusting network parameters") is extracted, and these methods are used as candidate strategies. Further, a random forest algorithm can be used to train a model based on historical data, learn the effectiveness weight of different repair methods, and generate a combined strategy (such as cleaning cache first, and adjusting parameters if it is ineffective) as a repair solution for the current fault. Further, when the historical fault record with a comprehensive similarity between the first similarity threshold and the second similarity threshold is only one, the repair method corresponding to the fault is directly extracted from the historical database, and it is determined as the repair solution for the current fault.

[0061] When all historical similarities are lower than the second similarity threshold, a fault propagation graph model can be started. The fault propagation graph model is constructed based on historical data, and records the conversion relationship between different fault states and the effect of repair methods. Specifically, the key feature vector of the current fault is input into the fault propagation graph model, the fault propagation graph model predicts the success probability of each possible repair method by simulating the fault evolution path, and selects the repair method with the highest expected success rate (such as "restoring from backup") as the solution to the current fault. For example, for a "data check error" fault, the model may predict that the success probability of "restoring from backup" is 85%, and the success probability of "recalculating checksum" is 30%.

[0062] One possible implementation of an embodiment of the present application is to repair the fault frame message based on the repair method, which includes: determining whether the repair method meets the automatic repair condition; if the repair method meets the automatic repair condition, generating a repair instruction based on the repair method and sending the repair instruction to the repair end corresponding to the fault frame message; if the repair method does not meet the automatic repair condition, obtaining a basic repair method and generating a basic repair instruction corresponding to the basic repair method, and sending the basic repair instruction to the repair end corresponding to the fault frame message.

[0063] The automatic repair condition refers to a set of pre-set rules and standards used by electronic devices in a data center storage backup system to evaluate whether automatic repair can be directly performed without human intervention when determining the fault repair method, including operation risk assessment (such as success rate > 90%), time window limit (such as non-peak business hours), and system resource state (such as CPU usage < 80%).

[0064] The basic repair method is a pre-defined general operation suitable for various fault scenarios, such as log collection, service state checking, and temporary data isolation.

[0065] Specifically, the attributes of the current repair method (such as operation type, impact range, and historical success rate) are obtained from the data center storage backup system knowledge base and compared with the automatic repair condition library. For example, if the repair method is "restart the backup service", check if the historical restart success rate of the current service is greater than 95%; if the repair method is "data retransmission", check if the network bandwidth is sufficient (current usage rate < 70%) and it is not a peak period (such as 2 am). When all conditions are met, it is determined that the "automatic repair condition is met".

[0066] The repair method is converted into an executable format instruction. After the instruction is generated, the electronic device sends it to the corresponding repair end through a secure channel (such as a TLS encrypted RESTAPI) and starts timeout monitoring (such as triggering an alarm if there is no response for 120 seconds). The repair end is the hardware device or software component responsible for performing the actual repair operation (such as a backup server or a storage controller).

[0067] Further, when the repair method does not meet the automatic repair condition, an operation that best matches the current fault type is selected from the basic repair library (such as "collect system logs") to obtain a basic repair method, and a basic repair instruction corresponding to the basic repair method is generated. Further, after the instruction is sent, an artificial intervention process is triggered: an alarm information is sent to the operation and maintenance team, and a fault work order is generated in the management interface with the basic repair result.

[0068] In one possible implementation of the embodiment of the application, when the repair method meets the automatic repair condition, it is determined whether the fault frame packet is successfully repaired, including: An automatic repair completion signal is collected, and after the repair completion signal is collected, repair monitoring data and a repaired fault frame packet are obtained; The repair monitoring data are compared with the abnormal packet form and the abnormal network communication data form respectively to determine whether the repair monitoring data have the abnormal packet form or / and the abnormal network communication data form; If it is determined that the repair monitoring data do not have the abnormal packet form or / and the abnormal network communication data form, the repaired fault frame packet is parsed to obtain a repair key feature vector, and based on a historical feature vector, it is determined whether the fault frame packet is successfully repaired.

[0069] The repaired fault frame packet is the data frame after repair, which contains the correction content of the original fault position.

[0070] Specifically, after sending the repair instruction, a timeout timer (such as 120 seconds) is started to listen for an HTTP 200 response or an MQTT message from the repair end. After receiving the success signal, the network bandwidth, delay, packet loss rate, and other indicators after repair are obtained from the monitoring system (such as Prometheus); the latest operation log of the backup software is extracted from the log system (such as ELKStack); and the fault frame packet after repair (such as the retransmitted data block) is obtained from the storage system.

[0071] Further, the repair monitoring data is compared with a preset abnormal pattern library. Specifically, the repaired packet header is parsed to check whether it contains the error code sequence in the historical fault (such as matched by a regular expression), and frequency domain analysis (such as FFT transform) is performed on the current network traffic to check whether there is a frequency feature (such as 50Hz jitter) matching the historical abnormality. If no matching item is found, the next feature verification is entered. If a matching item is found, an alarm information is sent to the operation and maintenance team.

[0072] Further, the fault frame packet after repair is deeply parsed to determine whether the fault frame packet is successfully repaired. Specifically, the version number, sequence number, and other key fields in the protocol header are extracted, the extracted feature vector is compared with the normal template in the historical feature library by cosine similarity calculation (threshold value is 0.95), and multi-dimensional verification is performed. More specifically, the multi-dimensional verification includes checking whether the timestamp of the repaired packet conforms to the normal timing, and verifying whether the length and the number of segments of the data frame are consistent with the standard template. If the similarity exceeds the threshold value, it is determined that the repair is successful; otherwise, an alarm information is sent to the operation and maintenance team.

[0073] Further, when the repair method does not meet the automatic repair condition, it is determined that the fault frame packet is not successfully repaired, and an alarm information is sent to the operation and maintenance team.

[0074] The above embodiment introduces a data center storage backup fault monitoring method from the perspective of method flow, and the following embodiment introduces a data center storage backup fault monitoring device from the perspective of virtual module or virtual unit. For details, see the following embodiment.

[0075] Referring to Figure 2 , the data center storage backup fault monitoring device 20 can specifically include: a collection module 201, a state determination module 202, a type determination module 203, a recording module 204, a repair module 205, and a sending module 206, wherein: A data center storage backup fault monitoring device 20 includes: The collection module 201 is configured to collect device state data and obtain current monitoring data, wherein the device state data is state data of a device corresponding to a data center storage backup system. The state determination module 202 is configured to determine whether the data center storage backup system is in a normal state based on the device state data and the current monitoring data. The type determination module 203 is configured to determine a fault type of the data center storage backup system if the data center storage backup system is not in the normal state, the fault type being a device fault or a data fault. The recording module 204 is configured to record a fault frame packet corresponding to the current monitoring data and determine a repair method corresponding to the fault frame packet if the fault type is the data fault. The repair module 205 is configured to repair the fault frame packet based on the repair method and determine whether the fault frame packet is successfully repaired. The sending module 206 is configured to generate a repair report corresponding to the current monitoring data and send the repair report to a corresponding operation and maintenance terminal if the repair is successful.

[0076] In one possible implementation of the embodiment, the device state data includes device temperature, device rotation speed and device vibration signal, and the current monitoring data includes running logs of backup software and network communication data. The state determination module 202 is configured to, when determining whether the data center storage backup system is in the normal state based on the device state data and the current monitoring data, specifically: process the device vibration signal through a CNN subnetwork to obtain a spatial feature vector of the device, and process the device temperature and the device rotation speed through an LSTM subnetwork to obtain a time sequence feature vector; input the spatial feature vector and the time sequence feature vector into an encoder, calculate inter-modal correlation weights through a multi-head attention mechanism, and generate a fusion feature vector; input the fusion feature vector into a fully connected classification network, and obtain an output device health score; determine whether the data center storage backup system is in the normal state based on the device health score; if it is determined that the data center storage backup system is not in the normal state, determine that the data center storage backup system is not in the normal state; if it is determined that the data center storage backup system is in the normal state, determine whether the data center storage backup system is in the normal state based on the current monitoring data.

[0077] In one possible implementation of the embodiment, when the state determination module 202 determines whether the data center storage backup system is in the normal state based on the current monitoring data, the state determination module 202 is specifically configured to: obtain historical fault data, the historical fault data including historical fault frame packets and historical network communication data corresponding to each historical fault frame; The historical failure data is analyzed in time domain and in frequency domain to obtain a first time domain relationship of the historical failure frame packet and the abnormal failure form, a second time domain relationship of the historical failure frame and abnormal network communication data, and a frequency domain feature vector matrix corresponding to the historical failure data; Based on the first time domain relationship, the second time domain relationship and the frequency domain feature vector matrix, the abnormal packet form and the abnormal network communication data form corresponding to the historical failure frame are determined. The current monitoring data are compared with the abnormal packet form and the abnormal network communication data form to determine whether the current monitoring data exist the abnormal packet form or / and the abnormal network communication data form. When the current monitoring data exist the abnormal packet form or / and the abnormal network communication data form, it is determined that the data center storage backup system is not in a normal state.

[0078] In one possible implementation of the embodiment, when the historical failure data is analyzed in time domain and in frequency domain to obtain a first time domain relationship of the historical failure frame packet and the abnormal failure form, a second time domain relationship of the historical failure frame and abnormal network communication data, and a frequency domain feature vector matrix corresponding to the historical failure data, the state determination module 202 is specifically configured to: The historical failure data is divided by a sliding window, and the abnormal failure form corresponding to each historical failure packet in each historical window is analyzed. Based on the abnormal failure form corresponding to each historical failure packet in each historical window, the first time domain relationship of the historical failure packet and the abnormal failure form is determined through correlation analysis. The abnormal network communication data in each historical window is determined, and the cross correlation of the historical window sequence is calculated to identify the time sequence coupling relationship of the historical failure frame and the abnormal network communication data, and form the second time domain relationship. The abnormal failure form and the abnormal network communication data in the historical failure data are analyzed in frequency domain to obtain a first frequency domain feature vector corresponding to the abnormal failure form and a second frequency domain feature vector corresponding to the abnormal network communication data. The first frequency domain feature vector and the second frequency domain feature vector are integrated to obtain a frequency domain feature vector matrix corresponding to the historical failure data.

[0079] In one possible implementation of the embodiment, the historical failure data further includes historical failure frame data, and when the repair mode corresponding to the failure frame packet is determined, the recording module 204 is specifically configured to: The failure frame packet is analyzed to extract a key feature vector, and the key feature vector is compared with a historical feature vector corresponding to the historical failure frame data to obtain a comprehensive similarity, the historical failure frame data including the historical feature vector and a historical repair mode. When the comprehensive similarity greater than the first similarity threshold exists, the comprehensive similarity greater than the first similarity threshold is determined as the first target similarity, and the historical repair mode corresponding to the first target similarity is obtained as the repair mode corresponding to the fault frame packet; When the comprehensive similarity greater than the first similarity threshold does not exist, and the comprehensive similarities greater than the second similarity threshold exist, the comprehensive similarity greater than the second similarity threshold is determined as the second target similarity, and the historical repair modes corresponding to the second target similarities are combined and applied based on the random forest algorithm to obtain an optimal strategy combination, and the optimal strategy combination is determined as the repair mode corresponding to the fault frame packet, wherein the second similarity threshold is less than the first similarity threshold; When the comprehensive similarities are all not greater than the second similarity threshold, a fault propagation graph model is constructed based on each historical repair mode, the key feature vector is input into the fault propagation graph model, and the expected success degrees of different repair modes output by the fault propagation graph model are obtained, and the repair mode with the highest expected success degree is selected as the repair mode corresponding to the fault frame packet.

[0080] In an embodiment of the present application, the repair module 205 is specifically configured to: determine whether the repair mode meets an automatic repair condition; if the repair mode meets the automatic repair condition, generate a repair instruction based on the repair mode, and send the repair instruction to a repair end corresponding to the fault frame packet; if the repair mode does not meet the automatic repair condition, obtain a basic repair mode, generate a basic repair instruction corresponding to the basic repair mode, and send the basic repair instruction to the repair end corresponding to the fault frame packet.

[0081] In an embodiment of the present application, when the repair mode meets the automatic repair condition, the repair module 205 is specifically configured to: automatically collect a repair completion signal, and after the repair completion signal is collected, obtain repaired repair monitoring data and a repaired fault frame packet; compare the repair monitoring data with the abnormal packet form and the abnormal network communication data form respectively to determine whether the repair monitoring data has the abnormal packet form or / and the abnormal network communication data form; if it is determined that the repair monitoring data does not have the abnormal packet form or / and the abnormal network communication data form, analyze the repaired fault frame packet to obtain a repaired key feature vector, and determine whether the fault frame packet is repaired successfully based on the historical feature vector.

[0082] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the system, device and unit described above can refer to the corresponding processes in the foregoing method embodiments, and will not be described here.

[0083] Referring to Figure 3 , the embodiments of the present application also introduce an electronic device from the perspective of an entity device, such as Figure 3 , as shown in Figure 3 The electronic device 300 shown in the figure includes a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, such as through a bus 302. Optionally, the electronic device 300 can also include a transceiver 304. It should be noted that the transceiver 304 is not limited to one in actual application, and the structure of the electronic device 300 does not constitute a limitation on the embodiments of the present application.

[0084] The processor 301 can be a CPU (Central Processing Unit, central processor), a general-purpose processor, a DSP (Digital Signal Processor, data signal processor), an ASIC (Application Specific Integrated Circuit, application specific integrated circuit), an FPGA (Field Programmable Gate Array, field programmable gate array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute various exemplary logical blocks, modules and circuits described in combination with the disclosure content of the present application. The processor 301 can also be a combination of computing functions, such as one or more microprocessor combinations, combinations of DSP and microprocessor, etc.

[0085] The bus 302 can include a channel for transmitting information between the above-mentioned components. The bus 302 can be a PCI (Peripheral Component Interconnect, peripheral component interconnect) bus or an EISA (Extended Industry Standard Architecture, extended industry standard architecture) bus, etc. The bus 302 can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, Figure 3 only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0086] The memory 303 can be a ROM (Read Only Memory) or other type of static storage device that can store static information and instructions; a RAM (Random Access Memory) or other type of dynamic storage device that can store information and instructions; an EEPROM (Electrically Erasable Programmable Read-Only Memory), a CD-ROM (Compact Disc Read-Only Memory) or other optical disk storage, a magnetic disk storage or other magnetic storage devices or any other medium capable of storing desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0087] The memory 303 is configured to store application program codes for implementing the solutions of the present application, and the processor 301 is configured to control the execution of the application program codes. The processor 301 is configured to execute the application program codes stored in the memory 303 to implement the content shown in the foregoing method embodiments.

[0088] The electronic device includes, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (Personal Digital Assistant), a PAD (Tablet Personal Computer), a PMP (Portable Multimedia Player), a vehicle terminal (e.g., a car navigation terminal), and the like, and a fixed terminal such as a digital TV, a desktop computer, and the like, and can also be a server or the like. Figure 3 The electronic device shown is merely an example and should not impose any limitation on the functions and use range of the embodiments of the present application.

[0089] The embodiments of the present application provide a computer readable storage medium having computer programs stored thereon, which, when executed on a computer, enable the computer to perform the corresponding content in the foregoing method embodiments.

[0090] It should be understood that, although the steps in the flowcharts of the drawings are shown in sequence according to the direction of the arrows, these steps are not necessarily executed in sequence according to the direction of the arrows. Unless otherwise specified herein, the execution of these steps is not limited in sequence, and they can be executed in other sequences. Moreover, at least some of the steps in the flowcharts of the drawings can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence is not necessarily sequential, but can be executed in rotation or alternation with at least some of the other steps or sub-steps or stages of other steps.

[0091] The above merely provides part of the embodiments of the present application, and it should be pointed out that, for those skilled in the art, some improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method for monitoring data center storage backup failures, the method comprising: The method comprises the following steps: Collecting device state data and obtaining current monitoring data, wherein the device state data is the state data of the device corresponding to the data center storage backup system; Based on the device state data and the current monitoring data, it is determined whether the data center storage backup system is in a normal state; If the data center storage backup system is not in a normal state, the fault type of the data center storage backup system is determined, which is device failure or data failure; When the fault type is data failure, the current monitoring data corresponding to the fault frame packet is recorded, and the repair method corresponding to the fault frame packet is determined; Based on the repair method, the fault frame packet is repaired, and it is determined whether the repair is successful; If the repair is successful, a repair report corresponding to the current monitoring data is generated, and the repair report is sent to the corresponding operation and maintenance terminal.

2. The data center storage backup failure monitoring method of claim 1, wherein, The device state data includes device temperature, device speed and device vibration signal, and the current monitoring data includes backup software running log and network communication data, Wherein, based on the device state data and the current monitoring data, it is determined whether the data center storage backup system is in a normal state, which comprises: The device vibration signal is processed by the CNN subnetwork to obtain the spatial feature vector of the device, and the device temperature and the device speed are processed by the LSTM subnetwork to obtain the time sequence feature vector; The spatial feature vector and the time sequence feature vector are input into the encoder, the inter-modal correlation weight is calculated by the multi-head attention mechanism, and the fusion feature vector is generated; The fusion feature vector is input into the full connection classification network, and the output device health score is obtained; Based on the device health score, it is determined whether the data center storage backup system is in a normal state; If it is determined that the data center storage backup system is not in a normal state, it is determined that the data center storage backup system is not in a normal state; If it is determined that the data center storage backup system is in a normal state, it is determined whether the data center storage backup system is in a normal state based on the current monitoring data.

3. The method of claim 2, wherein, Based on the current monitoring data, it is determined whether the data center storage backup system is in a normal state, which comprises: Obtain historical fault data, including historical fault frame packets and historical network communication data corresponding to each historical fault frame; Temporal analysis and frequency domain analysis are performed on the historical fault data to obtain the first time domain relationship between the historical fault frame packet and the abnormal fault form, the second time domain relationship between the historical fault frame and the abnormal network communication data, and the frequency domain feature vector matrix corresponding to the historical fault data; Based on the first time domain relationship, the second time domain relationship and the frequency domain feature vector matrix, the abnormal packet form and the abnormal network communication data form corresponding to the historical fault frame are determined; comparing the current monitoring data with the abnormal packet form and the abnormal network communication data form respectively to determine whether the current monitoring data has the abnormal packet form or / and the abnormal network communication data form; determining that the data center storage backup system is not in a normal state when the current monitoring data has the abnormal packet form or / and the abnormal network communication data form.

4. The method of claim 3, wherein, The time domain analysis and the frequency domain analysis of the historical fault data obtain a first time domain relationship between the historical fault frame packet and the abnormal fault form, a second time domain relationship between the historical fault frame and the abnormal network communication data, and a frequency domain feature vector matrix corresponding to the historical fault data, and the time domain analysis and the frequency domain analysis of the historical fault data include: performing sliding window segmentation on the historical fault data, and analyzing the abnormal fault form corresponding to each historical fault packet in each historical window; determining the first time domain relationship between the historical fault frame packet and the abnormal fault form through correlation analysis based on the abnormal fault form corresponding to each historical fault packet in each historical window; determining the abnormal network communication data in each historical window, calculating the cross-correlation of the historical window sequence, identifying the time sequence coupling relationship between the historical fault frame and the abnormal network communication data, and forming the second time domain relationship; performing frequency domain analysis on the abnormal fault form and the abnormal network communication data in the historical fault data respectively to obtain a first frequency domain feature vector corresponding to the abnormal fault form and a second frequency domain feature vector corresponding to the abnormal network communication data; integrating the first frequency domain feature vector and the second frequency domain feature vector to obtain a frequency domain feature vector matrix corresponding to the historical fault data.

5. The data center storage backup failure monitoring method of claim 3 or 4, wherein, The historical fault data further includes historical fault frame data, and the determination of the repair mode corresponding to the fault frame packet includes: analyzing the fault frame packet, extracting a key feature vector, and comparing the key feature vector with a historical feature vector corresponding to the historical fault frame data to obtain a comprehensive similarity, wherein the historical fault frame data includes a historical feature vector and a historical repair mode; when there is a comprehensive similarity greater than a first similarity threshold, determining the comprehensive similarity greater than the first similarity threshold as a first target similarity, and obtaining the historical repair mode corresponding to the first target similarity as the repair mode corresponding to the fault frame packet; when there is no comprehensive similarity greater than the first similarity threshold and there are multiple comprehensive similarities greater than a second similarity threshold, determining the comprehensive similarity greater than the second similarity threshold as a second target similarity, and combining the historical repair modes corresponding to the multiple second target similarities based on a random forest algorithm to obtain an optimal strategy combination, and determining the optimal strategy combination as the repair mode corresponding to the fault frame packet, wherein the second similarity threshold is less than the first similarity threshold. When the comprehensive similarity is all not greater than the second similarity threshold, a fault propagation graph model is constructed based on each historical repair method, the key feature vector is input into the fault propagation graph model, and expected success degrees of different repair methods output by the fault propagation graph model are obtained, and a repair method with the highest expected success degree is selected as a repair method corresponding to the fault frame packet.

6. The data center storage backup failure monitoring method of claim 5, wherein, The repairing the fault frame packet based on the repair method comprises: determining whether the repair method meets an automatic repair condition; if the repair method meets the automatic repair condition, generating a repair instruction based on the repair method and sending the repair instruction to a repair end corresponding to the fault frame packet; if the repair method does not meet the automatic repair condition, obtaining a basic repair method, generating a basic repair instruction corresponding to the basic repair method, and sending the basic repair instruction to the repair end corresponding to the fault frame packet.

7. The method of claim 6, wherein, When the repair method meets the automatic repair condition, the determining whether the fault frame packet is successfully repaired comprises: automatically collecting a repair completion signal, obtaining repaired repair monitoring data and a repaired fault frame packet after the repair completion signal is collected; comparing the repair monitoring data with the abnormal packet form and the abnormal network communication data form respectively to determine whether the repair monitoring data has the abnormal packet form or / and the abnormal network communication data form; if it is determined that the repair monitoring data does not have the abnormal packet form or / and the abnormal network communication data form, analyzing the repaired fault frame packet to obtain a repair key feature vector, and determining whether the fault frame packet is successfully repaired based on the historical feature vector.

8. A data center storage backup failure monitoring apparatus characterized by comprising: comprise: a collection module configured to collect device state data and obtain current monitoring data, the device state data being state data of a device corresponding to a data center storage backup system; a state determination module configured to determine whether the data center storage backup system is in a normal state based on the device state data and the current monitoring data; a type determination module configured to determine a fault type of the data center storage backup system if the data center storage backup system is not in the normal state, the fault type being a device fault or a data fault; a record module configured to record a fault frame packet corresponding to the current monitoring data and determine a repair method corresponding to the fault frame packet if the fault type is the data fault; a repair module configured to repair the fault frame packet based on the repair method and determine whether the fault frame packet is successfully repaired; a sending module configured to generate a repair report corresponding to the current monitoring data and send the repair report to a corresponding operation and maintenance terminal if the repair is successful.

9. An electronic device, comprising: The electronic device comprises: at least one processor; a memory; at least one application program, wherein the at least one application program is stored in the memory and configured to be executed by the at least one processor, and the at least one application program is configured to execute the data center storage backup fault monitoring method of any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program product, when executed in the computer, causes the computer to perform the data center storage backup failure monitoring method of any one of claims 1-7.

Images