A computer operation status monitoring and management system
By dividing the computer's running cycle into continuous intervals and using machine learning models to predict the computer's running state, the monitoring lag problem in the existing technology is solved, and the accurate and timely prediction of the computer's running state is achieved, and management efficiency is improved.
Patent Information
- Application Number
- CN202411123955.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-16
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-08-16
AI Technical Summary
When the existing computer operating status monitoring and management system monitors the computer operating status in real time, it cannot predict the status changes in advance, resulting in monitoring lag and reducing management efficiency.
The principle of interval continuity is used to divide the computer's running cycle into continuous monitoring intervals, collect historical status data, train machine learning models to predict the operating state of the next monitoring interval, and conduct early warning prompts and monitoring level selection based on real-time parameters.
By reducing the amount of data within the monitoring interval, increasing the monitoring frequency, achieving accurate prediction of the computer's operating status, discovering potential hazards in advance, avoiding monitoring lag, and improving management timeliness.
Smart Images

Figure CN118642922B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to computer monitoring and technical field, and more specifically, to a computer operation status monitoring and management system. Background Art
[0002] In the context of the rapid development of modern information technology, computers, as the core equipment for information processing, are widely used in all walks of life. With the complexity of computer systems and the diversity of application environments, the monitoring and management of computer operation status has become increasingly important. In order to timely and accurately schedule and operate computers efficiently and safely, it is necessary to monitor and manage the computer's operation status.
[0003] The patent application with reference publication number CN117539727A discloses a computer operation status monitoring method and monitoring system, which performs auxiliary detection on the hardware status of the computer through a hardware auxiliary detection module, not only realizing the extended analysis of the computer hardware status, but also providing information support for the analysis process of the hardware status detection module, and performing hardware status evaluation and analysis on the computer through the hardware status detection module, and generating a hardware status abnormality signal or a hardware status normal signal through analysis, so as to realize effective monitoring of the computer hardware status and reasonably judge its risk level, and perform detection and analysis on the application program running in the computer through the software status detection module, so as to realize effective monitoring of the software program running in the computer and reasonably judge its program abnormality, and the network detection and evaluation module obtains the network connected during the operation of the computer and marks it as the target network, and evaluates and analyzes the operation quality of the target network, so as to realize accurate evaluation of the quality of the network connected to the computer, and can reasonably analyze and accurately evaluate the hardware status, software status and network status of the computer, realize comprehensive monitoring of the computer and accurately judge its operation status, which is conducive to the management personnel being unable to make targeted improvement measures in time, and ensure the safe, stable and efficient operation of the computer.
[0004] The existing technology has the following deficiencies: when the existing management system monitors the operation status of the computer, it collects a large amount of diversified data at the current operation time of the computer, calculates and analyzes the data, and then determines the real-time operation status of the computer, so that the management system needs to calculate and analyze a large amount of multi-type data in a short time, which not only increases the workload of the management system, but also makes it easy for calculation and analysis errors to occur. In addition, the real-time calculation and analysis method can only analyze the real-time operation status of the computer, and cannot predict the upcoming changes in the computer's operation status in advance, resulting in a lag in the computer operation status monitoring, thereby reducing the management efficiency of the computer operation status monitoring.
[0005] In view of this, the present invention proposes a computer operation status monitoring and management system to solve the above problems. Summary of the invention
[0006] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a computer operation status monitoring and management system, applied to a monitoring and management server, comprising:
[0007] The monitoring interval module is used to divide the computer's operating cycle into continuous monitoring intervals according to the interval continuity principle; the data acquisition module is used to collect the computer's historical status data set in the continuous interval, the historical status data set includes comprehensive operating parameters and operating status values, the comprehensive operating parameters include working temperature value, intrusion detection trigger rate, update backup coverage rate and dangerous behavior ratio; the model training module is used to train a machine learning model that predicts the operating status value of the next monitoring interval based on the historical status data set; the prediction and judgment module is used to collect real-time comprehensive operating parameters, input them into the trained machine learning model, predict the operating status value of the next monitoring interval, and determine whether to issue an operating status warning prompt; the level selection module is used to compare the comprehensive operating parameters with the corresponding standard values, mark the parameters to be adjusted, and select the corresponding monitoring management level.
[0008] Furthermore, the interval continuity principle is that the duration of each monitoring interval is equal, and the last moment of the previous monitoring interval and the first moment of the next monitoring interval are adjacent moments.
[0009] Furthermore, the method for dividing the monitoring interval includes: taking the current time as the end time, working backwards to perform a complete operation status analysis to obtain the start time; recording the period between the start time and the end time as the operation cycle; querying the time taken by the data processor from receiving data to processing data and then outputting data i times through the timestamp to obtain i data processing times; marking the maximum and minimum values of the data processing time, and adding the maximum and minimum values of the data processing time to obtain the average to obtain the division time; using the division time as the division standard, marking the division time within the operation cycle The start time is the starting point of the first monitoring interval, and the end time is the end point of the last monitoring interval. The operation cycle is divided into equal time intervals to obtain monitoring interval.
[0010] Furthermore, the method for obtaining the intrusion detection trigger rate includes: In a monitoring interval, the w data to be processed are identified by keywords one by one through natural language processing technology to obtain w keywords; the corresponding processing permissions are respectively assigned to the w keywords through the permission management system, and the keywords assigned with processing permissions are respectively marked with processing orders to form processing events, and w processing events are obtained; according to the order of processing, the detection status of the w processing events when processing data is queried in sequence through the intrusion detection system; the processing events with the detection status of intrusion protection are recorded as intrusion events, and the number of intrusion events is counted. After comparing the number of intrusion events with the total number of processing events, the w processing events are obtained. The intrusion detection trigger rate is:
[0011] .
[0012] In the formula, For the The intrusion detection trigger rate of each monitoring interval, For the The number of intrusion events in the monitoring interval, For the The total number of processed events in a monitoring interval.
[0013] Further, the method for obtaining the updated backup coverage includes: In each monitoring interval, the number of data to be backed up in the backup center is counted one by one, and the data to be backed up are numbered in ascending order one by one; based on the preset backup time, Mark out the monitoring interval non-adjacent backup detection times; query them one by one by updating the backup system The number of the data to be backed up at the backup detection time; record the two data to be backed up with the same number as the target data, and count the number of target data; compare the number of target data with the number of data to be backed up in turn, and obtain Sub-coverage; the expression of sub-coverage is:
[0014] .
[0015] In the formula, For the The monitoring interval The sub-coverage rate of the backup detection time, For the The monitoring interval The number of target data at the backup detection time, For the The monitoring interval The number of data to be backed up at the backup detection time; after removing the maximum and minimum values of the sub-coverage rate, the remaining The coverage of each sub-unit is accumulated and averaged to obtain Update backup coverage; the expression of update backup coverage is:
[0016] .
[0017] In the formula, For the Update backup coverage for each monitoring interval, For the The monitoring interval The sub-coverage rate at the backup detection time.
[0018] Furthermore, the method for obtaining the proportion of dangerous behaviors includes: establishing two access points arranged inside and outside in the access management system, recording the access point located on the outside as the outer access point, and recording the access point located on the inside as the inner access point; marking the basic access level on the outer access point and the inner access point respectively, to obtain the outer basic level and the inner basic level; raising and lowering a preset extension level on the outer basic level and the inner basic level respectively, to obtain the outer level range and the inner level range; querying one by one through the access management system The access records received by the external access points and the internal access points in the monitoring interval are identified in turn, and the requested access levels of the access records are recorded in turn; the access records whose requested access levels are outside the external level range are recorded as the first dangerous behavior, and the number of the first dangerous behaviors is counted; the access records whose requested access levels are outside the internal level range are recorded as the second dangerous behavior, and the number of the second dangerous behaviors is counted; the number of the first dangerous behavior and the number of the second dangerous behavior are compared with the total number of access records, respectively, and then different weight factors are assigned and added to obtain The expression of the proportion of dangerous behaviors is:
[0019] .
[0020] In the formula, For the The proportion of risky behaviors in each monitoring interval, For the The number of first risky behaviors in each monitoring interval, For the The number of second risky behaviors in each monitoring interval, For the The total number of access records in the monitoring interval, , is the weight factor, , Both are greater than 0.
[0021] Furthermore, the operating status value includes a normal state and a dangerous state; the training method of the machine learning model for predicting the operating status value of the next monitoring interval includes: converting the comprehensive operating parameters into multiple feature vectors using a sliding window method, converting the operating status value into a label corresponding to the comprehensive operating parameters according to the sliding step, converting the normal state into 0, and converting the dangerous state into 1, one feature vector corresponds to one label, and constitutes a set of training data, and multiple sets of training data constitute a training set, arranging the comprehensive operating parameters in order of collection time, and presetting the prediction time step Z, sliding step Q and sliding window length N; using the feature vector as the input of the machine learning model, predicting the operating status value of the next monitoring interval after the time step Z as the output, and the subsequent operating status value of each training set as the prediction target, and taking the minimized sum of prediction errors as the training target, the machine learning model is trained to generate a machine learning model that predicts the operating status value of the next monitoring interval based on the comprehensive operating parameters of the previous monitoring interval.
[0022] Furthermore, the method for determining whether to issue an operating status warning prompt includes: when the machine learning model output is 0, the next operating status value is a normal state, and it is determined that no operating status warning prompt is issued; when the machine learning model output is 1, the next operating status value is a dangerous state, and it is determined that an operating status warning prompt is issued.
[0023] Furthermore, the method for marking the parameters to be adjusted includes: comparing the working temperature value with the preset temperature standard value, when the working temperature value is greater than the preset temperature standard value, marking the working temperature value as the parameter to be adjusted; comparing the intrusion detection trigger rate with the preset trigger standard value, when the intrusion detection trigger rate is greater than the preset trigger standard value, marking the intrusion detection trigger rate as the parameter to be adjusted; comparing the update backup coverage rate with the preset coverage standard value, when the update backup coverage rate is less than the preset coverage standard value, marking the update backup coverage rate as the parameter to be adjusted; comparing the dangerous behavior proportion with the preset proportion standard value, when the dangerous behavior proportion is greater than the preset proportion standard value, marking the dangerous behavior proportion as the parameter to be adjusted.
[0024] Furthermore, the monitoring management level includes the first-level management level and the second-level management level; the selection method of the first-level management level and the second-level management level includes: counting the number of parameters to be adjusted to obtain the adjustment value; when the adjustment value is 1 or 2, selecting the first-level management level; when the adjustment value is 3 or 4, selecting the second-level management level.
[0025] The technical effects and advantages of a computer operation status monitoring and management system of the present invention are as follows: the present invention divides the operation cycle of a computer into continuous monitoring intervals according to the interval continuity principle, collects a historical status data set of the computer in the continuous interval, the historical status data set includes comprehensive operation parameters and operation status values, trains a machine learning model for predicting the operation status value of the next monitoring interval according to the historical status data set, collects real-time comprehensive operation parameters, inputs them into the trained machine learning model, predicts the operation status value of the next monitoring interval, determines whether to issue an operation status early warning prompt, compares the comprehensive operation parameters with the corresponding standard values, marks the parameters to be adjusted, and selects the corresponding monitoring management level; compared with the prior art, on the basis of the interval continuity principle, the time span with a longer time span can be adjusted. The monitoring intervals are divided according to the large operating cycle, which reduces the amount of data in each monitoring interval and increases the monitoring frequency of the computer's operating status. At the same time, by collecting comprehensive operating parameters and based on the machine learning model, the operating status value of the next monitoring area can be accurately predicted according to the real-time status of the previous monitoring area, so as to make accurate and advance prediction and evaluation of the computer's operating status in the future, so as to discover the impending dangerous situation of the computer in advance, and mark the data that needs to be adjusted and optimized according to the current operating status, providing a targeted data basis for the adjustment and optimization operation of the computer's operating status, and then monitoring and processing can be carried out before the computer's operating status becomes dangerous, avoiding the lag caused by real-time monitoring management and improving the timeliness of computer operating status monitoring and management. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 A schematic diagram of a computer operation status monitoring and management system provided in Embodiment 1 of the present invention.
[0027] Figure 2 A flowchart of a computer operation status monitoring and management method provided in Embodiment 2 of the present invention.
[0028] Figure 3 A schematic diagram of a structural diagram of an electronic device provided in Embodiment 3 of the present invention.
[0029] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION
[0030] 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.
[0031] Example 1: Please refer to Figure 1 As shown, a computer operation status monitoring and management system described in this embodiment is applied to a monitoring and management server, including: a monitoring interval module, which divides the computer's operation cycle into continuous monitoring intervals according to the interval continuity principle; the operation cycle refers to the time required for the computer to perform a complete operation status monitoring and management, and serves as the basis for dividing the subsequent monitoring intervals, and the monitoring interval is the minimum length unit that constitutes the operation cycle, so that each monitoring interval can contain a sufficient amount of computer operation status change data, thereby representing the computer's operation status changes in a short period of time, and achieving an uninterrupted and refined segmentation effect of the computer operation status connection; and when dividing the operation cycle into monitoring intervals, it is necessary to ensure that the divided monitoring intervals can be in a continuous state on the timeline, and each monitoring interval can maintain independence, so it is necessary to use the interval continuity principle to divide the operation cycle; the interval continuity principle is: the duration of each monitoring interval is equal, and the last moment of the previous monitoring interval is equal to the time of the last moment ... The first moment of the next monitoring interval is an adjacent moment; through the interval continuity principle, it can ensure that the duration corresponding to each monitoring interval is consistent, and it can also ensure that the two adjacent monitoring intervals are in a continuous and uninterrupted manner on the timeline; the method of dividing the monitoring interval includes: taking the current moment as the end moment, and working backwards to perform a complete operation status analysis to obtain the start moment; the duration of a complete operation status analysis refers to the time taken from the first step to the last step in the steps of computer operation status monitoring and management, which can limit the overall time span of the computer operation status; the period from the start moment to the end moment is recorded as the operation cycle; query the time taken by the data processor from receiving data to processing data and then outputting data i times through the timestamp to obtain i data processing times; mark the maximum and minimum values of the data processing time, and add the maximum and minimum values of the data processing time and then calculate the average to obtain the division time; use the division time as the division standard to mark within the operation cycle The interval time is used as the dividing point between two adjacent monitoring intervals to ensure that the two adjacent monitoring intervals are in a continuous and uninterrupted state on the timeline; the start time is used as the starting point of the first monitoring interval, and the end time is used as the end point of the last monitoring interval. The operation cycle is divided into equal time intervals to obtain monitoring intervals; it should be noted that after the interval time is divided, the two adjacent monitoring intervals are in a continuous state on the timeline, so that the moments corresponding to the operation cycle can be divided and processed comprehensively and uninterruptedly, ensuring the continuity and stability of each monitoring interval on the timeline.
[0032] The data acquisition module collects the historical status data set of the computer in a continuous interval. The historical status data set includes comprehensive operating parameters and operating status values. The comprehensive operating parameters include operating temperature values, intrusion detection trigger rates, update backup coverage rates, and dangerous behavior ratios. Historical status data refers to the data set generated when the computer was running in the past monitoring interval, which can provide analysis of the good or bad computer operating status, and the comprehensive operating parameters are used to comprehensively represent the changes in the computer's operating status in the monitoring interval, so as to provide a data basis for the subsequent good or bad operating status, and also provide target data for subsequent optimization and control of the computer. The comprehensive operating parameters include operating temperature values, intrusion detection trigger rates, update backup coverage rates, and dangerous behavior ratios. The operating temperature value refers to the real-time operating temperature of the computer in each monitoring interval, which can be used to analyze the computer's hardware in the face of massive and complex data. The value after the working temperature rises during analysis, calculation and processing is expressed. When the working temperature value is larger, it means that the working temperature value of the computer hardware is larger when analyzing, calculating and processing massive and complex data. At this time, the excessively high working temperature will have a negative impact on the computer's operating state, and the computer's operating state is worse at this time; the working temperature value is obtained by querying the computer's built-in temperature monitoring software; the intrusion detection trigger rate refers to the ratio of the number of times the computer can trigger the intrusion detection system to perform intrusion detection processing when processing data to the total number of times in each monitoring interval, which can be used to express the number of times the computer's intrusion detection system is triggered. When the intrusion detection trigger rate is larger, it means that the number of times the computer can trigger the intrusion detection system to perform intrusion detection processing when processing data to the total number of times, and the computer's operating state is worse at this time; the method for obtaining the intrusion detection trigger rate includes: In a monitoring interval, w data to be processed are identified by keywords one by one through natural language processing technology to obtain w keywords; corresponding processing permissions are assigned to the w keywords through the permission management system, and the keywords assigned processing permissions are marked with processing orders to form processing events, and w processing events are obtained; processing permissions refer to the permissions corresponding to the data to be processed when processing data, and are used as the basis for keywords to perform intrusion detection during data processing, ensuring that each keyword can correspond to its own permissions; according to the order of processing, the detection status of w processing events when processing data is queried in turn through the intrusion detection system; the processing events with a detection status of intrusion protection are recorded as intrusion events, and the number of intrusion events is counted. After comparing the number of intrusion events with the total amount of processing events, the The intrusion detection trigger rate is:
[0033] .
[0034] In the formula, For the The intrusion detection trigger rate of each monitoring interval, For the The number of intrusion events in the monitoring interval, For the The total number of processed events in a monitoring interval.
[0035] The update backup coverage rate refers to the ratio of the data backed up by the computer's data backup center at each update to the total data within each monitoring interval, which can be used to indicate the amount of data backup when the computer is updated. The larger the update backup coverage rate, the larger the ratio of the data backed up by the computer's data backup center at each update to the total data, and the better the computer's operating status is at this time. The method for obtaining the update backup coverage rate includes: In each monitoring interval, the number of data to be backed up in the backup center is counted one by one, and the data to be backed up are numbered in ascending order one by one; based on the preset backup time, Mark out the monitoring interval The preset backup duration refers to the minimum span between two adjacent backup detection times, which can ensure that within the preset backup duration, the data to be backed up in the data backup center can complete at least one data update backup operation; query the backup system one by one by updating the backup system. The number of the data to be backed up at the backup detection time; record the two data to be backed up with the same number as the target data, and count the number of target data; compare the number of target data with the number of data to be backed up in turn, and obtain The expression of sub-coverage is:
[0036] .
[0037] In the formula, For the The monitoring interval The sub-coverage rate at the backup detection time, For the The monitoring interval The number of target data at the backup detection time, For the The monitoring interval The number of data to be backed up at the backup detection time; after removing the maximum and minimum values of the sub-coverage rate, the remaining The coverage of each sub-unit is accumulated and averaged to obtain Update backup coverage; the expression of update backup coverage is:
[0038] .
[0039] In the formula, For the Update backup coverage for each monitoring interval, For the The monitoring interval The sub-coverage rate at the backup detection time.
[0040] The proportion of dangerous behaviors refers to the proportion of the access records with dangerous behaviors identified by the computer to the total access records in each monitoring interval, which can be used to represent the number of dangerous accesses to the computer. The larger the proportion of dangerous behaviors, the larger the proportion of the access records with dangerous behaviors identified by the computer to the total access records, and the worse the operating state of the computer. The method for obtaining the proportion of dangerous behaviors includes: establishing two access points arranged inside and outside in the access management system, recording the access points located on the outside as the outer access points, and recording the access points located on the inside as the inner access points; the outer access points and the inner access points refer to the standard points in the access management system where the access records need to be queried and analyzed, and are used as the subsequent dangerous access points. The basis for identifying dangerous behaviors; the basic access level is noted on the external access point and the internal access point respectively to obtain the external basic level and the internal basic level; the basic access level refers to the initial level that can access the external access point and the internal access point normally and safely, so that the external access point and the internal access point can correspond to different access levels and permissions respectively; on the external basic level and the internal basic level, a preset extension level is raised and lowered respectively to obtain the external level range and the internal level range; the preset extension level is the minimum level unit for expanding the scope of the basic access level of a single point, so that the external basic level and the internal basic level can form a level range after expansion, thereby improving the efficiency of identifying dangerous behaviors; query one by one through the access management system The access records received by the external access points and the internal access points in the monitoring interval are collected, and the requested access levels of the access records are identified in turn; the requested access level refers to the access permission corresponding to the access record, and is used as the object for range comparison with the external level range and the internal level range, so as to improve the accuracy of identifying dangerous behaviors; the access records with the requested access level outside the external level range are recorded as the first dangerous behavior, and the number of the first dangerous behaviors is counted; the access records with the requested access level outside the internal level range are recorded as the second dangerous behavior, and the number of the second dangerous behaviors is counted; the number of the first dangerous behavior and the number of the second dangerous behavior are compared with the total number of access records, respectively, and then different weight factors are assigned and added to obtain The expression of the proportion of dangerous behaviors is:
[0041] .
[0042] In the formula, For the The proportion of risky behaviors in each monitoring interval, For the The number of first risky behaviors in each monitoring interval, For the The number of second risky behaviors in each monitoring interval, For the The total number of access records in the monitoring interval, , is the weight factor, , are all greater than 0; among them, , exemplary, is 0.522, is 0.478; , The size of is used to adjust the proportion of the number of the first dangerous behaviors and the number of the second dangerous behaviors in the total amount of access records, and to adaptively adjust the size of the proportion of dangerous behaviors, so that the change in the size of the number of the first dangerous behaviors and the number of the second dangerous behaviors can cause a relative change in the proportion of dangerous behaviors.
[0043] The operating status value is used to truly represent the actual operating status of the computer in each monitoring interval, and serves as the basis for judging whether the operating status of the computer has changed. The operating status value includes normal status and dangerous status. The normal status means that the computer's operating status in the monitoring interval is good and there are no abnormal conditions. The dangerous status means that the computer's operating status in the monitoring interval is poor and abnormal conditions have occurred. The operating status value is obtained by collecting the corresponding working temperature values, intrusion detection trigger rate, update backup coverage rate and dangerous behavior proportion rate under normal and dangerous conditions. The model training module trains a machine learning model to predict the operating status value of the next monitoring interval based on the historical status data set. After obtaining the historical status data set, the machine learning model can be trained based on the historical status data set, so that the machine learning model can predict the operating status value of the next monitoring interval based on the comprehensive operating parameters of the previous monitoring interval. The training method of a machine learning model for the operating status value of a monitoring interval includes: converting the comprehensive operating parameters into multiple feature vectors using a sliding window method, converting the operating status values into labels corresponding to the comprehensive operating parameters according to the sliding step, converting the normal state into 0, and converting the dangerous state into 1, one feature vector corresponds to one label, and constitutes a group of training data, and multiple groups of training data constitute a training set, arranging the comprehensive operating parameters in the order of collection time, and presetting the prediction time step Z, sliding step Q and sliding window length N; using the feature vector as the input of the machine learning model, predicting the operating status value of the next monitoring interval after the time step Z as the output, and the subsequent operating status value of each training set as the prediction target, taking the minimized sum of prediction errors as the training target, training the machine learning model, and generating a machine learning model that predicts the operating status value of the next monitoring interval based on the comprehensive operating parameters of the previous monitoring interval.
[0044] Exemplarily, the machine learning model is any one of the CNN neural network model or AlexNet; the calculation formula of the prediction error is:
[0045] .
[0046] In the formula, is the prediction error, is the group number of the eigenvector; For the The predicted state value corresponding to the group feature vector, For the The actual state value corresponding to the group training data; in the machine learning model, the feature vector is the comprehensive operating parameter, and the state value is the operating state value; other model parameters of the machine learning model, target loss value, optimization algorithm, training set test set validation set ratio and loss function optimization are all achieved through actual engineering implementation and continuous experimental tuning.
[0047] Specifically, a simple example of the sliding window method is as follows: assuming that you want to use comprehensive operating parameters (Q1, Q2, Q3, Q4, Q5, Q6) to train a machine learning model to predict the value of one time step in the future, you can use a sliding window of length 4 and a sliding step of length 1 to generate a predicted future training set and prediction target. For example: the training sets are (Q1, Q2, Q3, Q4) and (Q2, Q3, Q4, Q5), and the prediction targets are (E5) and (E6), E5 is the operating status value corresponding to the comprehensive operating parameters of the next monitoring interval of the monitoring interval where Q4 is located, and E6 is the operating status value corresponding to the comprehensive operating parameters of the next monitoring interval of the monitoring interval where Q5 is located.
[0048] The prediction and judgment module collects real-time comprehensive operating parameters, inputs them into the trained machine learning model, predicts the operating status value of the next monitoring interval, and determines whether to issue an operating status warning prompt; by inputting the collected real-time comprehensive operating parameters into the machine learning model, the operating status value of the next monitoring interval can be quickly and accurately predicted, so as to judge whether it is necessary to issue an operating status warning prompt for the computer's operating status based on the predicted operating status value; the determination method for whether to issue an operating status warning prompt includes: when the machine learning model output is 0, it means that the predicted next operating status value is normal, and at this time, there is no abnormality in the computer's operating status in the next monitoring interval, and it is determined that no operating status warning prompt is issued; when the machine learning model output is 1, it means that the predicted next operating status value is dangerous, and at this time, there is an abnormality in the computer's operating status in the next monitoring interval, and it is determined that an operating status warning prompt is issued.
[0049] The level selection module compares the comprehensive operating parameters with the corresponding standard values, marks the parameters to be adjusted, and selects the corresponding monitoring management level; the parameters to be adjusted refer to the specific parameters that will cause the operating status value of the computer in the next monitoring interval to become a dangerous state, and serve as data for subsequent adjustment and optimization of the computer operating status; the marking method of the parameters to be adjusted includes: comparing the operating temperature value with the preset temperature standard value; the preset temperature standard value refers to the maximum value of the operating temperature value when the operating status value of the computer in the next monitoring interval is in a normal state, which can limit the operating temperature value in the normal state; the preset temperature standard value is obtained by collecting a large number of historical maximum values of the operating temperature values when the operating status value of the computer in the next monitoring interval is in a normal state After that, the average value is calculated; when the working temperature value is greater than the preset temperature standard value, it means that the working temperature value at this time will cause the operating state value of the computer in the next monitoring interval to become a dangerous state, then the working temperature value is marked as a parameter to be adjusted; the intrusion detection trigger rate is compared with the preset trigger standard value; the preset trigger standard value refers to the maximum value of the intrusion detection trigger rate when the operating state value of the computer in the next monitoring interval is normal, which can limit the intrusion detection trigger rate in the normal state; the preset trigger standard value is obtained by collecting the maximum value of the intrusion detection trigger rate of a large number of computers in the history when the operating state value of the next monitoring interval is normal, and then calculating the average value; when the intrusion detection trigger rate is greater than the preset trigger standard value, it means that At this time, the intrusion detection trigger rate will cause the operating status value of the computer in the next monitoring interval to become a dangerous state, then the intrusion detection trigger rate will be marked as a parameter to be adjusted; the update backup coverage rate will be compared with the preset coverage standard value; the preset coverage standard value refers to the minimum value of the update backup coverage rate when the operating status value of the computer in the next monitoring interval is normal, which can limit the update backup coverage rate under normal conditions; the preset coverage standard value is obtained by collecting a large number of historical minimum values of the update backup coverage rate when the operating status value of the next monitoring interval of the computer is normal, and then calculating the average value; when the update backup coverage rate is less than the preset coverage standard value, it means that the update backup coverage rate at this time will cause the operating status of the computer in the next monitoring interval to be dangerous. If the status value becomes a dangerous state, the update backup coverage rate will be marked as a parameter to be adjusted; the dangerous behavior proportion is compared with the preset standard proportion value; the preset standard proportion value refers to the maximum value of the dangerous behavior proportion when the operating status value of the computer in the next monitoring interval is a normal state, which can limit the dangerous behavior proportion in the normal state; the preset standard proportion value is obtained by collecting a large number of historical maximum values of the dangerous behavior proportion when the operating status value of the computer in the next monitoring interval is a normal state, and then calculating the average value; when the dangerous behavior proportion is greater than the preset standard proportion value, it means that the dangerous behavior proportion at this time will cause the operating status value of the computer in the next monitoring interval to become a dangerous state, and the dangerous behavior proportion is marked as a parameter to be adjusted.
[0050] The monitoring management level refers to an expression of the severity of abnormal danger of the operating state value of the computer in the next monitoring interval, which is selected according to the number of parameters to be adjusted, so as to ensure that the computer can perform corresponding optimization processing before an abnormal situation occurs in the next monitoring interval; the monitoring management level includes the first-level management level and the second-level management level; the severity of abnormal danger of the operating state of the computer corresponding to the first-level management level is lower than the severity of abnormal danger of the operating state of the computer corresponding to the second-level management level; the selection method of the first-level management level and the second-level management level includes: counting the number of parameters to be adjusted to obtain the adjustment value; when the adjustment value is 1 or 2, the number of parameters marked as to be adjusted is 1 or 2, and the severity of abnormal danger of the operating state of the computer is low, then the first-level management level is selected; when the adjustment value is 3 or 4, the number of parameters marked as to be adjusted is 3 or 4, and the severity of abnormal danger of the operating state of the computer is high, then the second-level management level is selected.
[0051] After selecting the first-level management level or the second-level management level, the monitoring and management server can formulate corresponding optimization and adjustment measures for the computer's operating status based on the real-time monitoring and management level, thereby optimizing and adjusting the computer's operating status to prevent the computer from having a dangerous situation in the next monitoring interval, providing a good foundation for monitoring and management of the computer's operating status.
[0052] In this embodiment, the operation cycle of the computer is divided into continuous monitoring intervals according to the interval continuity principle, and the historical status data set of the computer in the continuous interval is collected. The historical status data set includes comprehensive operation parameters and operation status values. According to the historical status data set, a machine learning model for predicting the operation status value of the next monitoring interval is trained, and the real-time comprehensive operation parameters are collected and input into the trained machine learning model to predict the operation status value of the next monitoring interval, and determine whether to issue an operation status warning prompt, compare the comprehensive operation parameters with the corresponding standard values, mark the parameters to be adjusted, and select the corresponding monitoring management level. Compared with the prior art, on the basis of the interval continuity principle, the operation cycle with a larger time span can be monitored. The division reduces the amount of data in each monitoring interval and increases the monitoring frequency of the computer's operating status. At the same time, by collecting comprehensive operating parameters and based on the machine learning model, it can accurately predict the operating status value of the next monitoring area according to the real-time status of the previous monitoring area, thereby making accurate and advance predictions and evaluations of the computer's operating status in the future, so as to discover the impending dangerous situations of the computer in advance, and mark the data that needs to be adjusted and optimized according to the current operating status, providing a targeted data basis for the adjustment and optimization operations of the computer's operating status, and then monitoring and processing can be carried out before the computer's operating status becomes dangerous, avoiding the lag caused by real-time monitoring management and improving the timeliness of computer operating status monitoring and management.
[0053] Example 2: Please refer to Figure 2 As shown, the parts not described in detail in this embodiment are described in the first embodiment. A computer operation status monitoring and management method is provided, which is applied to a monitoring and management server and is implemented based on a computer operation status monitoring and management system, including: S1: According to the interval continuity principle, the computer operation cycle is divided into continuous monitoring intervals; S2: The historical status data set of the computer in the continuous interval is collected, and the historical status data set includes comprehensive operation parameters and operation status values. The comprehensive operation parameters include working temperature value, intrusion detection trigger rate, update backup coverage rate and dangerous behavior ratio; S3: According to the historical status data set, a machine learning model is trained to predict the operation status value of the next monitoring interval; S4: Real-time comprehensive operation parameters are collected and input into the trained machine learning model to predict the operation status value of the next monitoring interval, and determine whether to issue an operation status warning prompt; if an operation status warning prompt is issued, execute S5; if no operation status warning prompt is issued, repeat S4; S5: Compare the comprehensive operation parameters with the corresponding standard values, mark the parameters to be adjusted, and select the corresponding monitoring and management level.
[0054] Example 3: Please refer to Figure 3As shown, this embodiment discloses an electronic device, including a processor and a memory; wherein the memory stores a computer program that can be called by the processor; the processor executes and implements the computer operation status monitoring and management method by calling the computer program stored in the memory.
[0055] Since the electronic device introduced in this embodiment is an electronic device used to implement a computer operation status monitoring and management method in Embodiment 2 of this application, based on the computer operation status monitoring and management method introduced in the embodiment of this application, a person skilled in the art can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of this application is not described in detail here. As long as a person skilled in the art implements the electronic device used in a computer operation status monitoring and management method in the embodiment of this application, it belongs to the scope of protection of this application.
[0056] Example 4: Please refer to Figure 4 As shown, this embodiment discloses a computer-readable storage medium on which a rewritable computer program is stored; when the computer program is run, the computer operation status monitoring and management method described is executed.
[0057] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A computer operation status monitoring and management system, applied to a monitoring and management server, characterized in that: include: A monitoring interval module, used to divide the operation cycle of the computer into continuous monitoring intervals according to the interval continuity principle; A data collection module is used to collect a historical status data set of the computer in a continuous interval. The historical status data set includes comprehensive operating parameters and operating status values. The comprehensive operating parameters include operating temperature value, intrusion detection trigger rate, update backup coverage rate and dangerous behavior ratio; Methods for obtaining the intrusion detection trigger rate include: In p monitoring intervals, the w data to be processed are identified with keywords one by one through natural language processing technology to obtain w keywords; Through the authority management system, w keywords are respectively assigned corresponding processing authority, and the keywords assigned processing authority are respectively marked with processing order to form processing events, thereby obtaining w processing events; According to the processing sequence, the detection status of w processing events is queried in sequence through the intrusion detection system during data processing; Record the processing events with the detection status of intrusion protection as intrusion events, count the number of intrusion events, and compare the number of intrusion events with the total number of processing events to obtain p intrusion detection trigger rates; The expression of intrusion detection trigger rate is: Where QR jcp is the intrusion detection trigger rate of the pth monitoring interval, QR slp is the number of intrusion events in the pth monitoring interval, SJ zlp is the total number of processed events in the pth monitoring interval; Methods for obtaining updated backup coverage include: In p monitoring intervals, the number of data to be backed up in the backup center is counted one by one, and the data to be backed up are numbered in ascending order one by one; Based on the preset backup duration, s non-adjacent backup detection moments are marked in p monitoring intervals respectively; By updating the backup system, the numbers of the data to be backed up are queried one by one at the s backup detection moments; Record two to-be-backed data with the same number as target data, and count the number of target data; Compare the number of target data with the number of data to be backed up in turn to obtain s sub-coverages; The expression of sub-coverage is: In the formula, FG zps is the sub-coverage rate of the sth backup detection time in the pth monitoring interval, SL mbps is the number of target data at the sth backup detection time in the pth monitoring interval, SL fbps is the amount of data to be backed up at the sth backup detection time in the pth monitoring interval; After removing the maximum and minimum values of the sub-coverage, the remaining s-2 sub-coverages are accumulated and averaged to obtain p updated backup coverages; The expression for updating the backup coverage is: In the formula, GX fgp is the updated backup coverage rate of the pth monitoring interval, FG zpa is the sub-coverage rate of the a-th backup detection time in the p-th monitoring interval; Methods for obtaining the proportion of risky behaviors include: Two access points arranged inside and outside are respectively established in the access management system, and the access point located outside is recorded as the external access point, and the access point located inside is recorded as the internal access point; the external access point and the internal access point refer to the standard points in the access management system where access records need to be queried and analyzed; Respectively mark the basic access level on the external access point and the internal access point to obtain the external basic level and the internal basic level; On the outer basic level and the inner basic level, a preset extension level is raised and lowered respectively to obtain an outer level range and an inner level range; The access management system queries the access records received by the external access points and the internal access points in the p monitoring intervals one by one, and identifies the requested access levels of the access records in turn; Recording access records whose requested access level is outside the external level range as first risky behaviors, and counting the number of first risky behaviors; Recording access records whose requested access level is outside the inner level range as second dangerous behaviors, and counting the number of second dangerous behaviors; After comparing the number of the first dangerous behavior and the number of the second dangerous behavior with the total number of access records, they are assigned different weight factors and then added together to obtain the proportion of p dangerous behaviors; The expression of the proportion of dangerous behaviors is: Where WX zbp is the proportion of dangerous behaviors in the pth monitoring interval, SL w1p is the number of first dangerous behaviors in the pth monitoring interval, SL w2p is the number of the second dangerous behavior in the pth monitoring interval, ZL fwp is the total number of access records in the pth monitoring interval, σ1 and σ2 are weight factors, and both σ1 and σ2 are greater than 0; A model training module is used to train a machine learning model that predicts the operating status value of the next monitoring interval based on a set of historical status data; The prediction and judgment module is used to collect real-time comprehensive operating parameters, input them into the trained machine learning model, predict the operating status value of the next monitoring interval, and determine whether to issue an operating status warning prompt; The level selection module is used to compare the comprehensive operating parameters with the corresponding standard values, mark the parameters to be adjusted, and select the corresponding monitoring management level.
2. A computer operation status monitoring and management system according to claim 1, characterized in that: The interval continuity principle is that the duration of each monitoring interval is equal, and the last moment of the previous monitoring interval and the first moment of the next monitoring interval are adjacent moments.
3. A computer operation status monitoring and management system according to claim 2, characterized in that: The method for dividing the monitoring interval includes: Taking the current time as the end time, work backwards to calculate the time taken for a complete operation status analysis to obtain the start time; The period between the start time and the end time is recorded as the operation cycle; Query the time taken by the data processor from receiving data to processing data and then outputting data through the timestamp to obtain the i-th data processing time; Mark the maximum and minimum data processing time, add the maximum and minimum data processing time and calculate the average to obtain the partition time; Using the division duration as the division standard, p-1 interval moments are marked in the operation cycle; The start time is taken as the starting point of the first monitoring interval, the end time is taken as the end point of the last monitoring interval, and the operation cycle is divided into equal time lengths in combination with p-1 interval times to obtain p monitoring intervals.
4. A computer operation status monitoring and management system according to claim 3, characterized in that: The operating status value includes a normal state and a dangerous state; The training method of the machine learning model for predicting the operating status value of the next monitoring interval includes: The comprehensive operation parameters are converted into multiple feature vectors using the sliding window method. The operation state value is converted into a label corresponding to the comprehensive operation parameter according to the sliding step. The normal state is converted to 0, and the dangerous state is converted to 1. One feature vector corresponds to one label and constitutes a set of training data. Multiple sets of training data constitute a training set. The comprehensive operation parameters are arranged in the order of collection time, and the prediction time step Z, sliding step Q and sliding window length N are preset. The feature vector is used as the input of the machine learning model, and the operating status value of the next monitoring interval after the predicted time step Z is used as the output. The subsequent operating status value of each training set is used as the prediction target. The machine learning model is trained with the minimized sum of prediction errors as the training target to generate a machine learning model that predicts the operating status value of the next monitoring interval based on the comprehensive operating parameters of the previous monitoring interval.
5. A computer operation status monitoring and management system according to claim 4, characterized in that: The method for determining whether to issue an operation status warning prompt comprises: When the machine learning model output is 0, the next operating state value is normal, and it is determined that no operating state warning prompt will be issued; When the machine learning model output is 1, the next operating status value is a dangerous state, and an operating status warning prompt is issued.
6. A computer operation status monitoring and management system according to claim 5, characterized in that: The marking method of the parameter to be adjusted includes: Compare the working temperature value with a preset temperature standard value, and when the working temperature value is greater than the preset temperature standard value, mark the working temperature value as a parameter to be adjusted; The intrusion detection trigger rate is compared with a preset trigger standard value, and when the intrusion detection trigger rate is greater than the preset trigger standard value, the intrusion detection trigger rate is marked as a parameter to be adjusted; The update backup coverage ratio is compared with a preset coverage standard value, and when the update backup coverage ratio is less than the preset coverage standard value, the update backup coverage ratio is marked as a parameter to be adjusted; The risky behavior ratio is compared with a preset standard ratio value, and when the risky behavior ratio is greater than the preset standard ratio value, the risky behavior ratio is marked as a parameter to be adjusted.
7. A computer operation status monitoring and management system according to claim 6, characterized in that: The monitoring management level includes the primary management level and the secondary management level; The selection methods for the first-level management level and the second-level management level include: Count the number of parameters to be adjusted and obtain the adjustment value; When the adjustment value is 1 or 2, the first-level management level is selected; When the adjustment value is 3 or 4, the secondary management level is selected.
