Data center operation monitoring system and method based on out-of-band technology and medium

Abstract

The invention discloses a data center operation monitoring system and method based on an out-of-band technology and a medium, and the system comprises an out-of-band data collection module which is used for obtaining hardware state information, performance parameters and system logs of data center equipment through an out-of-band communication channel, and a data processing and analysis module which is connected with the out-of-band data collection module and is used for processing and analyzing the data center equipment. The system comprises a data processing and analyzing module for cleaning, converting and analyzing collected data, a monitoring alarm module for sending alarm information when a preset threshold value is exceeded according to an analysis result of the data processing and analyzing module, and a visual display module for displaying the analysis result to a user in a visual chart, graph and report form mode. And the configuration management module is used for configuring and managing various parameters of the monitoring system. Through innovative designs of out-of-band data acquisition, intelligent modules and the like, comprehensive data acquisition and accurate alarm and prediction are realized, the monitoring level of a data center is improved, the fault risk is reduced, the cost is saved, and stable and efficient operation is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of data center monitoring systems, and in particular to a data center operation monitoring system, method and medium based on out-of-band technology. Background Art

[0002] With the rapid development of information technology, data centers are becoming increasingly important in modern society. As key facilities for storing, processing and distributing large amounts of data, the stable operation of data centers is essential for the normal operation of enterprises, government agencies and various industries. However, traditional data center operation monitoring methods face many challenges and are difficult to meet the growing business needs and high reliability requirements.

[0003] Traditional data center monitoring mainly relies on in-band communication to obtain information through the device's own operating system and management interface. This method cannot obtain the device's underlying hardware status information and detailed performance parameters, such as the server's power status, fan speed, and hardware fault code. This results in the inability to detect potential problems in the device in a timely manner, increasing the risk of system failure.

[0004] Traditional monitoring systems are inefficient when processing large amounts of data. As the scale of data centers continues to expand, the number of devices and the amount of data are growing exponentially. Traditional data processing methods have difficulty cleaning, converting, and analyzing massive amounts of data in a short period of time. This affects the real-time and accuracy of monitoring data and makes it impossible to reflect the actual operating status of the data center in a timely manner. For example, when a sudden performance problem occurs in a data center, it may take a long time for traditional monitoring systems to analyze the root cause of the problem, delaying the best time to troubleshoot.

[0005] Traditional monitoring systems are usually only capable of simple data statistics and report generation, and lack the ability to conduct in-depth data analysis and intelligent prediction. They are unable to predict the future operating trends of the equipment based on historical data and current status, and are also difficult to discover potential performance bottlenecks and potential failures. For example, traditional monitoring systems often cannot provide effective support for hard disk life prediction and network traffic trend analysis.

[0006] The alarm mechanism of traditional monitoring systems is imperfect. The alarm accuracy is low. It is usually based on simple threshold settings. When a parameter exceeds the preset threshold, the alarm is triggered, which is prone to false alarms and missed alarms. It cannot accurately reflect the actual problems of the data center, and it is difficult to distinguish the severity of the alarm. All alarms are presented in the same way, making it difficult for administrators to quickly judge the urgency and scope of the problem. In addition, the traditional alarm method lacks correlation analysis capabilities and can only alarm a single device or parameter, and cannot automatically determine the correlation between multiple alarms. When complex fault conditions occur in the data center, it increases the difficulty of troubleshooting.

[0007] The traditional monitoring system has a simple visualization function and can only provide basic charts and reports, which cannot intuitively reflect the overall operation of the data center. Administrators need to spend a lot of time and energy analyzing complex data tables and graphs to understand the operation status. At the same time, it lacks flexibility and customization and cannot meet the personalized needs of different users.

[0008] Configuration management of traditional monitoring systems is difficult. Manual configuration is inefficient, including device information entry, monitoring threshold setting, and alarm policy customization, which all need to be done manually by the administrator. This is not only inefficient but also prone to errors. With the continuous increase of data center equipment and changes in business needs, manual configuration management has become increasingly difficult and cannot meet the requirements of rapid response and flexible adjustment. In addition, traditional monitoring systems lack the functions of importing and exporting configuration templates and configuration version management. When the data center environment changes or the system is upgraded, the administrator needs to reconfigure and cannot quickly migrate the existing configuration. The lack of configuration version management also makes it difficult for administrators to roll back to historical configuration versions, increasing the risk of configuration management.

[0009] With the continuous expansion of the scale of data centers and the increasing complexity of business needs, a more efficient, intelligent and comprehensive operation monitoring system is urgently needed. First of all, it is necessary to have comprehensive data collection capabilities, and be able to obtain the hardware status information, performance parameters and system logs of data center equipment through out-of-band communication technology to achieve all-round monitoring of equipment. For example, real-time monitoring of the server's power status, fan speed, hard disk temperature underlying hardware information, as well as the port status and traffic statistics performance parameters of network equipment. Secondly, it is necessary to have strong data processing and analysis capabilities, using big data processing technology and machine learning algorithms to quickly process massive monitoring data, and conduct in-depth analysis and intelligent prediction. For example, real-time monitoring of abnormal fluctuations and trend changes in data, predicting equipment failure risks and performance bottlenecks, and providing a basis for preventive maintenance. Furthermore, a complete alarm mechanism is indispensable, supporting alarm level settings and alarm history query, and can perform alarm correlation analysis based on the time, type and device association of alarm events, providing alarm root cause analysis functions, and helping administrators quickly locate the root cause of the problem. At the same time, intuitive visual display is also very important, providing multi-dimensional view switching and custom display functions, and presenting the operating status and analysis results of data center equipment to users in the form of intuitive charts, graphics and reports. Finally, efficient configuration management is the key. It supports remote configuration and batch configuration operations, and has configuration template import and export functions and configuration version management functions, making it convenient for administrators to quickly configure and manage various parameters of the monitoring system. Summary of the invention

[0010] The present invention proposes a data center operation monitoring system, method and medium based on out-of-band technology to solve the problems mentioned in the above-mentioned prior art.

[0011] In order to achieve the above object, the present invention adopts the following technical solution: a data center operation monitoring system based on out-of-band technology, comprising:

[0012] Out-of-band data acquisition module: used to obtain hardware status information, performance parameters and system logs of data center equipment through out-of-band communication channels. The out-of-band data acquisition module includes out-of-band communication interfaces adapted to different types of equipment to ensure seamless connection with various types of servers, storage devices and network devices;

[0013] Data processing and analysis module: connected to the out-of-band data collection module, cleans, converts and analyzes the collected data to extract valuable operation information. This module uses big data processing technology and machine learning algorithms to monitor abnormal fluctuations and trend changes in data in real time;

[0014] Monitoring and alarm module: Based on the analysis results of the data processing and analysis module, an alarm message is issued for situations that exceed the preset threshold. The alarm methods include but are not limited to sound and light alarms, email notifications, and SMS reminders. At the same time, this module supports alarm level settings and alarm history query. The monitoring and alarm module uses the following formula to calculate the comprehensive alarm index AI:

[0015]

[0016] Where n is the number of different types of alarm parameters (such as high CPU usage, high memory usage, and hard disk read and write error rate), A i is the actual value of the ith alarm parameter, w i is the weight coefficient of the ith alarm parameter, f(A i ) is the alarm function of the i-th alarm parameter (when the CPU usage exceeds 80%, f(CPU usage )=1, otherwise 0); m is the number of alarms in different time periods, T j is the number of alarms in the jth time period, v j is the weight coefficient of the jth time period (determined according to the importance of the time period or the frequency of problems in the time period in historical data), g(T j ) is the conversion function of the number of alarms in a time period, and a logarithmic function is used to smooth the impact of the number of alarms. g(T j )=log(T j +1). When AI exceeds the preset alarm threshold, the alarm mechanism is triggered.

[0017] Visualization display module: displays the operating status and analysis results of data center equipment to users in the form of intuitive charts, graphs and reports, so that users can understand the overall operation of the data center in real time. The visualization display module provides multi-dimensional view switching and customized display functions;

[0018] Configuration management module: used to configure and manage various parameters of the monitoring system, including device information entry, monitoring threshold setting, alarm strategy customization and user authority management. The configuration management module supports remote configuration and batch configuration operations. The configuration management module uses the following formula to evaluate the configuration rationality index CI:

[0019] Where k is the number of configuration parameter types (such as the number of monitoring devices, the number of alarm threshold settings, and the number of user permission settings), P l is the actual value of the lth configuration parameter, Q l is the ideal value or industry standard value of the lth configuration parameter, h(P l , Q l ) is the rationality function of the lth configuration parameter, and the absolute value function is used to measure the deviation between the actual value and the ideal value. l , Q l )=|P l -Q l |,S l is the importance weight of the lth configuration parameter (determined according to the influence of the configuration parameter on the system operation). The smaller the CI value, the more reasonable the configuration. When the CI exceeds a certain threshold, the system can prompt the user to optimize the configuration.

[0020] Intelligent prediction module: Use historical data and machine learning models to predict the future operating status of data center equipment, including hardware failure prediction and performance bottleneck prediction, to provide a basis for preventive maintenance. This module uses time series analysis and deep learning algorithms, combined with the operating rules of the equipment and historical failure data for prediction. The prediction results are presented in the form of probability and provide a detailed prediction report. The intelligent prediction module uses the following fault prediction probability formula FP:

[0021]

[0022] Where t is the current time, t0 is the time when the equipment last operated normally, σ is the time standard deviation calculated based on the historical failure time interval data of the equipment, which reflects the discrete degree of equipment failure time; n is the number of factors that affect equipment failure (such as equipment operating temperature, humidity, load), p i is the probability of the ith factor causing equipment failure (obtained through historical data and machine learning algorithm training), x iis the current state value of the i-th factor (e.g., the degree to which the temperature exceeds the normal range). The larger the FP value, the higher the probability that the device will fail in the future.

[0023] Energy consumption management module: monitors the energy consumption of data center equipment in real time, and provides energy consumption optimization suggestions by analyzing the power consumption and load data of the equipment. This module has the function of analyzing energy consumption data, can identify peak energy consumption periods and high-energy consumption equipment, and supports the formulation of energy-saving strategies, such as dynamically adjusting equipment power and optimizing equipment operation modes to reduce energy consumption and operating costs of the data center. The energy consumption management module uses the following energy consumption evaluation formula:

[0024]

[0025] Among them, E actual is the actual energy consumption of the equipment, E rated is the rated energy consumption of the equipment, reflecting the energy efficiency of the equipment; m is the number of different load states, L j is the operating time proportion under the jth load state, u j is the energy consumption weight coefficient under the jth load state (determined according to the energy consumption characteristics of the equipment under different loads); C peak is the energy consumption during the peak period, C average is the average energy consumption, and v is the adjustment coefficient for peak energy consumption periods (considering that peak energy consumption periods have a greater impact on energy costs). The EE value can be used to evaluate the energy consumption of the device. When the EE value exceeds the preset energy-saving threshold, the system can trigger energy-saving strategy adjustment recommendations.

[0026] Furthermore, the out-of-band data acquisition module has the function of automatically identifying the type and protocol of the out-of-band interface of the device, and can dynamically adjust the acquisition frequency according to the working status of the device to reduce the impact on the normal operation of the device.

[0027] Furthermore, the data processing and analysis module includes a data cleaning submodule, a feature extraction submodule and an anomaly detection submodule. The data cleaning submodule is used to remove noise and duplicate data in the collected data. The feature extraction submodule is used to extract key features from the cleaned data. The anomaly detection submodule adopts an anomaly detection algorithm based on deep learning to monitor the operating status of the equipment in real time and judge anomalies.

[0028] Furthermore, the monitoring alarm module supports alarm correlation analysis, and can automatically determine whether there is a chain alarm situation based on the time, type and device association of the alarm event, and provide an alarm root cause analysis function to help users quickly locate the root cause of the problem.

[0029] Furthermore, the visualization display module provides real-time data update and dynamic chart display functions. Users can access monitoring data anytime and anywhere through a web interface or a mobile client, and can customize a personalized monitoring dashboard according to their needs.

[0030] Furthermore, the configuration management module has the function of importing and exporting configuration templates, which is convenient for users to quickly migrate configurations in different data center environments or when the system is upgraded. It also supports configuration version management, allowing users to roll back to historical configuration versions.

[0031] Furthermore, a data center operation monitoring method based on out-of-band technology includes the following steps:

[0032] Data collection steps: Collect hardware status information, performance parameters, and system logs of data center equipment through out-of-band communication channels to ensure that the collection process does not affect the normal business operation of the equipment;

[0033] Data processing steps: clean, convert and analyze the collected data, remove noise and duplicate data, extract key features, and use machine learning algorithms to monitor the data in real time and judge anomalies;

[0034] Alarm judgment step: compare the analyzed data according to the preset threshold and alarm strategy. When the data is found to be beyond the normal range, the alarm mechanism is triggered and the corresponding alarm information is sent to relevant personnel;

[0035] Visualization step: Visualize the operation status and analysis results of the data center equipment in the form of charts, graphs and reports, providing users with an intuitive monitoring interface, so that users can understand the overall operation of the data center in real time;

[0036] Configuration management steps: configure and manage various parameters of the monitoring system, including adding, modifying and deleting device information, setting monitoring thresholds, customizing alarm strategies, and allocating and managing user permissions;

[0037] Intelligent prediction step: Use historical data and trained machine learning models to predict the future operating status of data center equipment, generate preventive maintenance plans and recommendations based on the prediction results, and regularly update the prediction report;

[0038] Energy consumption management steps: Real-time monitoring of energy consumption data of data center equipment, analysis of energy consumption trends and equipment load relationships, formulation and adjustment of energy-saving strategies based on analysis results, and evaluation of the implementation effects of energy-saving strategies;

[0039] Security audit steps: record various operation logs during the operation of the data center, use security audit rules to analyze the logs, detect security threats and violations, issue security alerts in a timely manner and generate audit reports.

[0040] Furthermore, before the data collection step, a device initialization step is also included for automatically identifying the type and out-of-band interface information of the data center device and establishing an out-of-band communication connection with the device;

[0041] In the data processing step, real-time stream processing technology is used to analyze the collected data in real time, and batch processing technology is combined to conduct in-depth analysis of historical data to explore the potential patterns and trends of the data;

[0042] In the alarm judgment step, alarms are graded according to the severity and impact scope of the alarm event. Different alarm methods and response processes are used for alarms of different levels.

[0043] In the visualization step, users are supported to customize the visualization content and layout according to their own focus and business needs. At the same time, data drilling function is provided, and users can obtain detailed equipment information and operation data by clicking on data points in the chart.

[0044] In the configuration management step, a configuration audit function is provided to record the user's modification operations and modification time of configuration parameters, so as to facilitate tracing and analysis when problems arise.

[0045] Furthermore, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a data center operation monitoring method based on out-of-band technology is implemented.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The present invention uses an out-of-band data acquisition module and an out-of-band communication interface that adapts to different types of equipment to ensure seamless connection with various data center equipment, accurately obtain information such as hardware status, overcome the problem of incomplete data collection in traditional monitoring methods, and lay the foundation for comprehensive and accurate monitoring.

[0048] The data processing and analysis module uses big data processing technology and machine learning algorithms to efficiently clean, convert and analyze collected data, monitor abnormal fluctuations and trend changes in real time, improve the speed and accuracy of data processing, and provide strong support for timely problem discovery and scientific decision-making.

[0049] The monitoring alarm module uses a unique comprehensive alarm index calculation formula to accurately judge the alarm situation, supports alarm level setting and correlation analysis, can quickly locate the root cause of the problem, greatly improve the fault response speed, and reduce the impact of faults.

[0050] The visualization display module presents the data center operation status and analysis results in an intuitive and diverse manner to meet the personalized needs of users, allowing users to grasp the overall situation in real time and improve management efficiency.

[0051] The configuration management module evaluates configurations through a rationality index, supports remote and batch configuration operations, as well as configuration template import and export and version management, simplifies the configuration management process, and improves the rationality and flexibility of configuration.

[0052] The intelligent prediction module uses time series analysis and deep learning algorithms to predict the future operating status of equipment, provide a scientific basis for preventive maintenance, and reduce the risk of equipment failure. The energy consumption management module monitors energy consumption in real time and provides energy-saving strategies to effectively reduce data center operating costs.

[0053] In summary, the present invention comprehensively improves the level of data center operation monitoring and provides a solid guarantee for the stable, efficient and energy-saving operation of the data center. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a schematic block diagram of a data center operation monitoring system based on out-of-band technology proposed by the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise" and "counterclockwise" indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0057] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. In addition, the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, and it can be the internal connection of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0058] Reference Figure 1 :A data center operation monitoring system based on out-of-band technology, comprising:

[0059] Out-of-band data collection module: used to obtain hardware status information, performance parameters and system logs of data center equipment through out-of-band communication channels. The out-of-band data collection module includes out-of-band communication interfaces adapted to different types of equipment to ensure seamless connection with various servers, storage devices and network devices;

[0060] Data processing and analysis module: connected to the out-of-band data collection module, cleans, converts and analyzes the collected data to extract valuable operation information. This module uses big data processing technology and machine learning algorithms to monitor abnormal fluctuations and trend changes in data in real time;

[0061] Monitoring and alarm module: Based on the analysis results of the data processing and analysis module, an alarm message is issued for situations that exceed the preset threshold. The alarm methods include but are not limited to sound and light alarms, email notifications, and SMS reminders. At the same time, this module supports alarm level settings and alarm history query. The monitoring and alarm module uses the following formula to calculate the comprehensive alarm index AI:

[0062]

[0063] Where n is the number of different types of alarm parameters (such as high CPU usage, high memory usage, and hard disk read and write error rate), A i is the actual value of the ith alarm parameter, w i is the weight coefficient of the ith alarm parameter, f(A i ) is the alarm function of the i-th alarm parameter (when the CPU usage exceeds 80%, f(CPU usage )=1, otherwise 0); m is the number of alarms in different time periods, Tj is the number of alarms in the jth time period, v j is the weight coefficient of the jth time period (determined according to the importance of the time period or the frequency of problems in the time period in historical data), g(T j ) is the conversion function of the number of alarms in a time period, and a logarithmic function is used to smooth the impact of the number of alarms. g(T j )=log(T j +1). When AI exceeds the preset alarm threshold, the alarm mechanism is triggered.

[0064] Visualization display module: displays the operating status and analysis results of data center equipment to users in the form of intuitive charts, graphs and reports, so that users can understand the overall operation of the data center in real time. The visualization display module provides multi-dimensional view switching and customized display functions;

[0065] Configuration management module: used to configure and manage various parameters of the monitoring system, including device information entry, monitoring threshold setting, alarm strategy customization and user authority management. The configuration management module supports remote configuration and batch configuration operations. The configuration management module uses the following formula to evaluate the configuration rationality index CI:

[0066] Where k is the number of configuration parameter types (such as the number of monitoring devices, the number of alarm threshold settings, and the number of user permission settings), P l is the actual value of the lth configuration parameter, Q l is the ideal value or industry standard value of the lth configuration parameter, h(P l , Q l ) is the rationality function of the lth configuration parameter, and the absolute value function is used to measure the deviation between the actual value and the ideal value. l , Q l )=|P l -Q l |,S l is the importance weight of the lth configuration parameter (determined according to the influence of the configuration parameter on the system operation). The smaller the CI value, the more reasonable the configuration. When the CI exceeds a certain threshold, the system can prompt the user to optimize the configuration.

[0067] Intelligent prediction module: Use historical data and machine learning models to predict the future operating status of data center equipment, including hardware failure prediction and performance bottleneck prediction, to provide a basis for preventive maintenance. This module uses time series analysis and deep learning algorithms, combined with the operating rules of the equipment and historical failure data for prediction. The prediction results are presented in the form of probability and provide a detailed prediction report. The intelligent prediction module uses the following fault prediction probability formula FP:

[0068]

[0069] Where t is the current time, t0 is the time when the equipment last operated normally, σ is the time standard deviation calculated based on the historical failure time interval data of the equipment, which reflects the discrete degree of equipment failure time; n is the number of factors that affect equipment failure (such as equipment operating temperature, humidity, load), p i is the probability of the ith factor causing equipment failure (obtained through historical data and machine learning algorithm training), x i is the current state value of the i-th factor (e.g., the degree to which the temperature exceeds the normal range). The larger the FP value, the higher the probability that the device will fail in the future.

[0070] Energy consumption management module: monitors the energy consumption of data center equipment in real time, and provides energy consumption optimization suggestions by analyzing the power consumption and load data of the equipment. This module has the function of analyzing energy consumption data, can identify peak energy consumption periods and high-energy consumption equipment, and supports the formulation of energy-saving strategies, such as dynamically adjusting equipment power and optimizing equipment operation modes to reduce energy consumption and operating costs of the data center. The energy consumption management module uses the following energy consumption evaluation formula:

[0071]

[0072] Among them, E actual is the actual energy consumption of the equipment, E rated is the rated energy consumption of the equipment, reflecting the energy efficiency of the equipment, m is the number of different load states, L j is the operating time proportion under the jth load state, u j is the energy consumption weight coefficient under the jth load state (determined according to the energy consumption characteristics of the equipment under different loads); C peak is the energy consumption during the peak period, C average is the average energy consumption, and v is the adjustment coefficient for peak energy consumption periods (considering that peak energy consumption periods have a greater impact on energy costs). The EE value can be used to evaluate the energy consumption of the device. When the EE value exceeds the preset energy-saving threshold, the system can trigger energy-saving strategy adjustment recommendations.

[0073] In the present invention, the out-of-band data acquisition module demonstrates powerful functions and intelligent features. The module has the advanced function of automatically identifying the type and protocol of the out-of-band interface of the device, which means that no matter what type of data center equipment it faces, it can quickly and accurately determine the specific specifications of its out-of-band interface and the communication protocol it follows. Through this automatic identification mechanism, the need for manual intervention is greatly reduced, and the automation level and deployment efficiency of the system are improved.

[0074] At the same time, the module can also dynamically adjust the collection frequency according to the working status of the device. When the device is in a high-load operation state, the module will intelligently reduce the collection frequency to reduce the impact on the normal operation of the device and avoid increasing the burden on the device due to excessive data collection, thereby ensuring that the device can focus on processing key business. On the contrary, when the device is in a relatively idle state, the module can appropriately increase the collection frequency to obtain richer and more timely data, thereby achieving more detailed monitoring of the device's operating status.

[0075] In the present invention, the data processing and analysis module consists of multiple submodules working together to provide strong support for in-depth data analysis and equipment status monitoring. The data cleaning submodule is responsible for removing noise and duplicate data from the collected data. It uses advanced filtering algorithms and data deduplication technology to effectively identify and remove interference factors and duplicate information in the data, ensuring that the data input to the subsequent processing links has high quality and accuracy.

[0076] The feature extraction submodule extracts key features from the cleaned data. By using various feature extraction algorithms, such as principal component analysis and independent component analysis, this module can extract the key features that best reflect the operating status and performance of the equipment from a large amount of data, providing a strong basis for subsequent anomaly detection and analysis.

[0077] The anomaly detection submodule uses an anomaly detection algorithm based on deep learning to monitor the operating status of the equipment in real time and judge anomalies. After being trained with a large amount of data, the deep learning model can learn the characteristics of the normal operating mode of the equipment, so that it can accurately identify abnormal situations that are different from the normal mode. Once an anomaly is detected, the module will immediately issue an alarm to provide a guarantee for timely measures.

[0078] In the present invention, the monitoring alarm module is highly intelligent and practical. It supports alarm correlation analysis and can automatically determine whether there is a chain alarm situation based on the time, type and device correlation of the alarm event. When multiple alarm events are closely related in time or involve devices with correlation, the module will automatically perform correlation analysis to determine whether there is a potential systemic problem.

[0079] In addition, the module also provides alarm root cause analysis function to help users quickly locate the root cause of the problem. Through in-depth analysis of alarm events and mining of correlation relationships, the module can provide users with detailed alarm root cause analysis reports, pointing out the possible links and causes of the problem, greatly shortening the troubleshooting time.

[0080] In the present invention, the visualization display module provides users with a convenient and intuitive monitoring experience. The module provides real-time data update and dynamic chart display functions to ensure that users can understand the latest operating status of data center equipment at any time. Whether it is the performance indicators, operating status or alarm information of the equipment, it can be displayed in the form of intuitive charts, graphics, etc., so that users can see it at a glance.

[0081] Users can access monitoring data anytime and anywhere through the web interface or mobile client, breaking the limitations of time and space. No matter where the user is, as long as there is a network connection, they can view the operation of the data center in real time through the web page or mobile client, which is convenient for users to conduct remote monitoring and management.

[0082] At the same time, users can also customize personalized monitoring dashboards according to their needs. Users can select the devices and indicators they are concerned about and display them on the personalized dashboard, which facilitates quick access to key information and improves monitoring efficiency.

[0083] In the present invention, the configuration management module provides users with convenient and efficient configuration management functions. The module has the function of importing and exporting configuration templates, which facilitates users to quickly migrate configurations in different data center environments or when the system is upgraded. Users can export existing configuration templates and then import them in new environments or upgraded systems, greatly reducing the time and workload of configuration work.

[0084] At the same time, the module supports configuration version management, allowing users to roll back to historical configuration versions. When users modify the configuration, if they find a problem or need to restore to the previous configuration state, they can easily roll back to the historical configuration version through the version management function, which improves the stability and reliability of the system.

[0085] The present invention also discloses a data center operation monitoring method based on out-of-band technology, comprising the following steps:

[0086] Data collection steps: Through advanced out-of-band communication channels, the hardware status information, performance parameters and system logs of data center equipment are accurately collected. During the collection process, special technical means are used to ensure that there will be no interference with the normal business operation of the equipment. For example, through intelligent flow control and data caching mechanisms, the resource occupancy of the collection task is reasonably allocated, so that the collection process is like silently carried out behind the operation of the equipment, without affecting the key business processing of the equipment at all. At the same time, the collection algorithm is continuously optimized, and the collection frequency and method are dynamically adjusted according to the characteristics and workload of different devices to achieve the best collection effect and the minimum impact.

[0087] Data processing steps: Comprehensive and detailed cleaning, conversion and analysis of the collected data. First, use efficient data cleaning algorithms to remove noise and duplicate data to ensure the accuracy and purity of the data. Then, use intelligent conversion algorithms to convert the data into a unified format and standard to facilitate subsequent analysis and processing. Next, use advanced feature extraction technology to extract key features from the cleaned data. These features can accurately reflect the operating status and performance of the equipment. Finally, use powerful machine learning algorithms to monitor the data in real time and judge anomalies. After being trained with a large amount of data, the machine learning model can quickly and accurately identify abnormal patterns in the data and promptly detect possible problems with the equipment.

[0088] Alarm judgment steps: According to the preset threshold and alarm strategy, the analyzed data is strictly compared. When the data is found to be beyond the normal range, the alarm mechanism is triggered immediately. The system will send corresponding alarm information to relevant personnel according to different situations to ensure that the problem can be handled in time. For example, for urgent and serious problems, sound and light alarms and SMS notifications can be used to ensure that relevant personnel can be informed and take action at the first time. At the same time, the alarm strategy can be customized according to the actual situation, and the triggering conditions and methods of the alarm can be flexibly adjusted to meet the needs of different scenarios.

[0089] Visual display steps: Display the operating status and analysis results of the data center equipment in the form of intuitive charts, graphs and reports. Through the carefully designed visual interface, a clear and intuitive monitoring interface is provided for users. Users can understand the overall operation of the data center at a glance, including the performance indicators, fault status, energy consumption, etc. of the equipment. At the same time, the visual display supports multi-dimensional view switching and customized display functions. Users can choose different display methods and contents according to their needs and preferences, and obtain the required information conveniently and quickly.

[0090] Configuration management steps: Comprehensively configure and manage various parameters of the monitoring system. This includes adding, modifying, and deleting device information to ensure that the system can accurately identify and monitor each device. The monitoring threshold settings can be adjusted according to the characteristics and business needs of different devices to improve the accuracy and effectiveness of alarms. The customization of alarm strategies can flexibly select different alarm methods and trigger conditions according to actual conditions to meet the needs of different scenarios. The allocation and management of user permissions can ensure the security and stability of the system and prevent unauthorized operations and access.

[0091] Intelligent prediction steps: Use rich historical data and trained machine learning models to accurately predict the future operating status of data center equipment. Through in-depth analysis and mining of historical data, machine learning models can learn the laws and trends of equipment operation, thereby making reasonable predictions about future operating status. Based on the prediction results, the system will generate detailed preventive maintenance plans and suggestions to help users maintain and manage equipment in advance and avoid potential failures and problems. At the same time, the prediction report is updated regularly to ensure the accuracy and timeliness of the prediction results.

[0092] Energy consumption management steps: Real-time monitoring of energy consumption data of data center equipment, and analysis of energy consumption trends and equipment load relationships through advanced data analysis technology. Understanding the energy consumption performance of different equipment under different loads provides a scientific basis for formulating energy-saving strategies. Based on the analysis results, formulate and adjust reasonable energy-saving strategies, such as dynamically adjusting equipment power, optimizing equipment operation modes, etc. At the same time, evaluate the implementation effect of energy-saving strategies, continuously optimize energy-saving strategies, and reduce energy consumption and operating costs of data centers.

[0093] Security audit steps: record various operation logs during the operation of the data center, and conduct in-depth analysis of the logs using strict security audit rules. Detect security threats and violations, issue security alerts in a timely manner, and ensure the safe and stable operation of the data center. Generate detailed audit reports to provide strong support and basis for security management. Through comprehensive recording and analysis of operation logs, potential security risks and problems can be discovered in a timely manner, and effective measures can be taken to prevent and deal with them;

[0094] Security audit steps: record various operation logs during the operation of the data center, use security audit rules to analyze the logs, detect security threats and violations, issue security alerts in a timely manner and generate audit reports.

[0095] In the present invention, a device initialization step is set before the data collection step, which plays a vital role. It can automatically identify the type of data center equipment and out-of-band interface information, and accurately determine the characteristics and connection requirements of different devices through an intelligent identification mechanism. After determining the device type and interface information, the system will quickly establish an out-of-band communication connection with the device to ensure the stability and efficiency of data transmission, laying a solid foundation for subsequent data collection work.

[0096] In the data processing step, advanced real-time stream processing technology is used to analyze the collected data in real time. This technology can process the data at the moment it is generated, ensuring timely monitoring of the operating status of the data center equipment. At the same time, in order to explore the potential laws and trends of the data, batch processing technology is also combined to conduct in-depth analysis of historical data. Through the comprehensive analysis of a large amount of historical data, long-term trends and potential problems in equipment operation can be discovered, providing a scientific basis for the optimization and maintenance of the data center.

[0097] In the alarm judgment step, the severity and impact scope of the alarm event are fully considered. The system will classify the alarms according to different situations. For alarms with higher severity and larger impact scope, more urgent and obvious alarm methods will be used, such as sound and light alarms, and a rapid response process will be initiated at the same time to ensure that relevant personnel can learn about the problem and take measures as soon as possible. For relatively minor alarms, a milder alarm method can be used, such as email notification, and the response process can also be relatively slow, so as to reasonably allocate resources and improve processing efficiency.

[0098] In the visualization step, the personalized needs of users are fully considered. Users are supported to customize the visualization content and layout according to their focus and business needs. Users can choose which equipment information to display and how to display it according to their work habits and focus. At the same time, a data drilling function is also provided. When users click on the data points in the chart, they can obtain detailed equipment information and operating data, which is convenient for users to deeply understand the specific situation of the equipment and provide a more accurate basis for decision-making.

[0099] In the configuration management step, a configuration audit function is provided. This function can record the user's modification operations and modification time of configuration parameters. When problems occur, they can be traced and analyzed by viewing the audit records. This can clarify responsibilities and help quickly locate the root cause of the problem, improve the efficiency of problem solving, and ensure the stable operation of the data center.

[0100] The present invention also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a data center operation monitoring method based on out-of-band technology is implemented.

[0101] The above are only preferred specific implementation modes of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can replace or change the technical solution and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A data center operation monitoring system based on out-of-band technology, characterized in that: include: Out-of-band data acquisition module: used to obtain hardware status information, performance parameters and system logs of data center equipment through an out-of-band communication channel, and the out-of-band data acquisition module is adapted to the out-of-band communication interface; Data processing and analysis module: connected with the out-of-band data collection module, cleans, converts and analyzes the collected data, and uses big data processing technology and machine learning algorithms to monitor abnormal fluctuations and trend changes of data in real time; Monitoring and alarm module: Based on the analysis results of the data processing and analysis module, an alarm message is issued for situations that exceed the preset threshold. The alarm methods include sound and light alarms, email notifications, and SMS reminders. It supports alarm level settings and alarm history query. The monitoring and alarm module uses the following formula to calculate the comprehensive alarm index AI: Where n is the number of different types of alarm parameters, A i is the actual value of the ith alarm parameter, w i is the weight coefficient of the ith alarm parameter, f(A i ) is the alarm function of the ith alarm parameter; m is the number of alarms in different time periods, T j is the number of alarms in the jth time period, v j is the weight coefficient of the jth time period, g(T j ) is the conversion function of the number of alarms in a time period, and a logarithmic function is used to smooth the impact of the number of alarms. g(T j )=log(T j +1), when AI exceeds the preset alarm threshold, the alarm mechanism is triggered. Visualization display module: displays the operating status and analysis results of data center equipment to users in the form of charts, graphs and reports. The visualization display module includes multi-dimensional view switching and custom display functions; Configuration management module: used to configure and manage various parameters of the monitoring system, including device information entry, monitoring threshold setting, alarm strategy customization and user authority management. The configuration management module supports remote configuration and batch configuration operations. The configuration management module uses the following formula to evaluate the configuration rationality index CI: Where k is the number of configuration parameter types (such as the number of monitoring devices, the number of alarm threshold settings, and the number of user permission settings), P l is the actual value of the lth configuration parameter, Q l is the ideal value or industry standard value of the lth configuration parameter, h(P l , Q l ) is the rationality function of the lth configuration parameter, and the absolute value function is used to measure the deviation between the actual value and the ideal value. l , Q l )=|P l -Q l |,S l is the importance weight of the lth configuration parameter. The smaller the CI value, the more reasonable the configuration. When the CI exceeds a certain threshold, the system prompts the user to optimize the configuration. Intelligent prediction module: Use historical data and machine learning models to predict the future operating status of data center equipment, including hardware failure prediction and performance bottleneck prediction. Use time series analysis and deep learning algorithms to combine the operating rules of the equipment and historical failure data for prediction. The prediction results are presented in the form of probability and a prediction report is provided. The intelligent prediction module uses the following fault prediction probability formula FP: Where t is the current time, t0 is the time when the equipment last operated normally, σ is the time standard deviation calculated based on the historical failure time interval data of the equipment, which reflects the discrete degree of equipment failure time; n is the number of factors that affect equipment failure, p i is the probability that the ith factor causes equipment failure, x i is the current state value of the i-th factor. The larger the FP value, the higher the probability of equipment failure. Energy consumption management module: It has the function of analyzing energy consumption data, identifying peak energy consumption periods and high-energy consumption equipment, supporting dynamic adjustment of equipment power and optimizing equipment operation mode, and is used to reduce energy consumption and operating costs of data centers. The energy consumption management module uses the following energy consumption evaluation formula: Among them, E actual is the actual energy consumption of the equipment, E rated is the rated energy consumption of the equipment, reflecting the energy efficiency of the equipment; m is the number of different load states, L j is the operating time proportion under the jth load state, u j is the energy consumption weight coefficient under the jth load state; C peak is the energy consumption during the peak period, C average is the average energy consumption, v is the adjustment coefficient during the peak energy consumption period, and the energy consumption of the equipment is evaluated by the EE value. When the EE value exceeds the preset energy-saving threshold, the system triggers the energy-saving strategy adjustment suggestion.

2. The data center operation monitoring system based on out-of-band technology according to claim 1 is characterized in that: The out-of-band data acquisition module has the function of automatically identifying the type and protocol of the out-of-band interface of the device, and dynamically adjusting the acquisition frequency according to the working status of the device, so as to reduce the impact on the normal operation of the device.

3. The data center operation monitoring system based on out-of-band technology according to claim 1, characterized in that: The data processing and analysis module includes a data cleaning submodule, a feature extraction submodule and an anomaly detection submodule. The data cleaning submodule is used to remove noise and duplicate data in the collected data. The feature extraction submodule is used to extract key features from the cleaned data. The anomaly detection submodule adopts an anomaly detection algorithm based on deep learning to monitor the operating status of the equipment in real time and judge anomalies.

4. The data center operation monitoring system based on out-of-band technology according to claim 1, characterized in that: The monitoring alarm module supports alarm correlation analysis, automatically determines whether there is a chain alarm situation based on the time, type and device correlation of the alarm event, and provides an alarm root cause analysis function.

5. The data center operation monitoring system based on out-of-band technology according to claim 1, characterized in that: The visualization display module provides real-time data update and dynamic chart display functions. Users can access monitoring data anytime and anywhere through a web interface or a mobile client, and customize a personalized monitoring dashboard according to their needs.

6. The data center operation monitoring system based on out-of-band technology according to claim 1, characterized in that: The configuration management module has the functions of importing and exporting configuration templates, so that users can quickly migrate configurations in different data center environments or when upgrading the system. It also supports configuration version management, allowing users to roll back to historical configuration versions.

7. A data center operation monitoring method based on out-of-band technology, using the data center operation monitoring system based on out-of-band technology according to any one of claims 1 to 6, characterized in that: The following steps are involved: Data collection steps: Collect hardware status information, performance parameters, and system logs of data center equipment through out-of-band communication channels to ensure that the collection process does not affect the normal business operation of the equipment; Data processing steps: clean, convert and analyze the collected data, remove noise and duplicate data, extract key features, and use machine learning algorithms to monitor the data in real time and judge anomalies; Alarm judgment step: compare the analyzed data according to the preset threshold and alarm strategy. When the data is found to be beyond the normal range, the alarm mechanism is triggered and the corresponding alarm information is sent to relevant personnel; Visualization step: Visualize the operation status and analysis results of the data center equipment in the form of charts, graphs and reports, providing users with an intuitive monitoring interface, so that users can understand the overall operation of the data center in real time; Configuration management steps: configure and manage various parameters of the monitoring system, including adding, modifying and deleting device information, setting monitoring thresholds, customizing alarm strategies, and allocating and managing user permissions; Intelligent prediction step: Use historical data and trained machine learning models to predict the future operating status of data center equipment, generate preventive maintenance plans and recommendations based on the prediction results, and update the prediction report; Energy consumption management steps: Real-time monitoring of energy consumption data of data center equipment, analysis of energy consumption trends and equipment load relationships, formulation and adjustment of energy-saving strategies based on analysis results, and evaluation of the implementation effects of energy-saving strategies; Security audit steps: record various operation logs during the operation of the data center, use security audit rules to analyze the logs, detect security threats and violations, issue security alerts and generate audit reports.

8. The data center operation monitoring method based on out-of-band technology according to claim 7 is characterized in that: Before the data collection step, a device initialization step is also included, which is used to automatically identify the type and out-of-band interface information of the data center device and establish an out-of-band communication connection with the device; In the data processing step, real-time stream processing technology is used to analyze the collected data in real time, and batch processing technology is combined to conduct in-depth analysis of historical data to explore the potential patterns and trends of the data; In the alarm judgment step, alarms are graded according to the severity and impact scope of the alarm event. Different alarm methods and response processes are used for alarms of different levels. In the visualization step, users are supported to customize the visualization content and layout according to their own focus and business needs. At the same time, a data drilling function is provided, and users can obtain detailed equipment information and operating data by clicking on data points in the chart. In the configuration management step, a configuration audit function is provided to record the user's modification operations and modification time on configuration parameters.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the data center operation monitoring method based on out-of-band technology according to any one of claims 7 to 8 is implemented.

Images