Comprehensive monitoring SaaS service platform based on cloud computing architecture

Abstract

The application discloses a comprehensive monitoring SaaS service platform based on a cloud computing architecture, and relates to the technical field of hardware monitoring.The application solves the problem that when abnormal hardware is determined, various situations exist, which can easily cause misjudgment, thereby affecting the corresponding judgment of the corresponding operating personnel and the subsequent maintenance efficiency.The application is aimed at relevant physical machines with abnormal operation, and first performs numerical analysis to determine whether the abnormal misjudgment is caused by numerical fluctuation.Then, the load period of the cloud architecture is determined, the determined load period is comprehensively analyzed with the abnormal period of the physical machine to determine whether the relevant physical machine with abnormality exists in the load influence situation.Through step-by-step analysis, the relevant physical machine with real abnormality is determined and displayed, the comprehensiveness in the monitoring process is improved, the maintenance operation of external maintenance personnel is facilitated, and the maintenance efficiency is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hardware monitoring, in particular to a comprehensive monitoring SaaS service platform based on a cloud computing architecture. BACKGROUND

[0002] Cloud computing, at least as an extension of virtualization, has an increasingly wide range of influence, but cloud computing cannot support complex enterprise environments, so cloud computing architecture is emerging, experience shows that before cloud computing matures, we should pay more attention to the details of the system cloud computing architecture, based on the analysis of some existing cloud computing products and personal experience, a set of cloud computing architecture is summarized, the cloud computing architecture can be mainly divided into four layers.

[0003] The cloud computing architecture is composed of several hardware, according to the parameter cooperation between the several hardware, the data processing of the whole cloud computing architecture is realized, according to the running parameters in the running process of the hardware, the parameter monitoring is carried out through the set monitoring cloud platform, and the abnormal hardware is judged based on the abnormal parameters, but in the actual processing process, when the abnormal hardware is judged, there are many cases, which is easy to cause misjudgment, thereby affecting the corresponding judgment of the corresponding operating personnel and the subsequent maintenance efficiency. SUMMARY

[0004] In view of the defects of the prior art, the present application provides a comprehensive monitoring SaaS service platform based on a cloud computing architecture, which solves the problem that when the abnormal hardware is judged, there are many cases, which is easy to cause misjudgment, thereby affecting the corresponding judgment of the corresponding operating personnel and the subsequent maintenance efficiency.

[0005] To achieve the above purpose, the present application is realized by the following technical scheme: a comprehensive monitoring SaaS service platform based on a cloud computing architecture, comprising:

[0006] The data monitoring end monitors the physical machine working parameters associated with the cloud computing architecture, and transmits the real-time monitored physical machine working parameters to the abnormal device calibration end;

[0007] The abnormal device calibration end compares the specific working parameters monitored by different physical machines, and based on the specific comparison result, the abnormal physical machine is calibrated, including:

[0008] The real-time monitored specific working parameters are calibrated as G i , wherein i represents different physical machines, and the working parameters G i of the corresponding physical machine are compared with the comparison interval of the corresponding physical machine, wherein the comparison interval is a preset interval;

[0009] If G The corresponding time is marked as an abnormal time. If the duration of the abnormal time exceeds 5 seconds, the physical machine is marked as an abnormal physical machine. Otherwise, no marking is performed.

[0010] A set of monitoring periods T is determined, where T is a preset value. The number of abnormal physical machines generated in the current monitoring period T is determined. If there is only one set of physical machines in the current monitoring period T, a single set of abnormal analysis ends is executed. If there are multiple sets of physical machines in the current monitoring period T, a multiple set of abnormal analysis ends is executed.

[0011] The single set of abnormal analysis ends performs numerical analysis on the marked single set of abnormal physical machines to determine whether the abnormal physical machine is truly abnormal and performs re-marking. The specific method is as follows:

[0012] A set of monitoring micro-periods is determined, and the micro-period is a preset period. The output working parameters generated by the abnormal physical machine in the micro-period are recorded, and the output working interval is generated according to the maximum value or minimum value in the output working parameters.

[0013] The input parameters corresponding to the abnormal physical machine in the current micro-period are determined, and the output parameters corresponding to the input parameters are confirmed from the past parameter data. The minimum value and the maximum value are extracted from the extracted output parameters to generate the standard interval.

[0014] The output working interval and the standard interval are compared to determine the intersection range and the specific proportion ZB of the intersection range in the output working interval. If ZB≥Y1, the abnormal physical machine is marked as a fluctuating physical machine. If ZB

[0015] The multiple set of abnormal analysis ends performs numerical analysis on the marked multiple set of abnormal physical machines to determine whether there is an affected physical machine, and transmits the confirmed affected physical machine to the display end for display, including:

[0016] The specific abnormal time period corresponding to the abnormal physical machine in the current monitoring period T is determined, and the abnormal time period is the specific time period of numerical abnormality. The cross analysis of the confirmed specific abnormal time periods is performed to identify the cross time period. If there is only one set of cross time periods, the cross time period is marked as a pending time period. If there are multiple sets of cross time periods, the minimum value and the maximum value of the time period of the multiple sets of cross time periods are obtained to generate the pending time period belonging to the multiple sets of cross time periods.

[0017] The load period of the cloud computing architecture in the current monitoring period T is determined, and the load period is determined by the load state of the architecture. The cross analysis of the load period and the pending time period is performed to determine the intersection range, and then the specific proportion ZZ of the intersection range in the load period is determined.

[0018] If ZZ>=Y2, the affected signal is generated, otherwise, no signal is generated, and the numerical analysis of the multiple groups of abnormal physical machines is carried out through the single-group abnormality analysis end, to determine whether they are real abnormality, wherein Y2 is a preset value;

[0019] Based on the affected signal:

[0020] If there is only one group in the corresponding cross period, the several groups of abnormal physical machines associated with the cross period are marked as affected physical machines, and are displayed through the display end;

[0021] If there are multiple groups in the corresponding cross period, the cross period that has the cross range with the load period is marked as the affected period, and the several groups of abnormal physical machines associated with the affected period are marked as the affected physical machines, and are displayed through the display end, the cross period that does not have the cross range is marked as the unaffected period, and the numerical analysis of the abnormal physical machines associated with the unaffected period is carried out through the single-group abnormality analysis end, to determine whether they are real abnormality.

[0022] Preferably, the total processing center determines whether there is a related abnormal physical machine of data flow based on the determined abnormal physical machine, if there is, the linkage group is locked, and whether the corresponding abnormal period of the linkage group is cross abnormality is analyzed to determine whether there is a linkage abnormal group; comprising:

[0023] From the determined several groups of abnormal physical machines, the related abnormal physical machine of data flow is confirmed, and is marked as the linkage group;

[0024] The abnormal period of different abnormal physical machines of the linkage group in the monitoring period T is determined, and whether the corresponding abnormal period has a cross period is determined:

[0025] If there is a cross period, it is determined that the cross period is located at the maximum proportion value of different abnormal periods, if the maximum proportion value>=90%, the linkage group is marked as the linkage abnormal group, otherwise, no marking is performed;

[0026] If there is no cross period, no marking is performed;

[0027] The confirmed linkage abnormal group and abnormal physical machine are displayed through the display end.

[0028] The application provides a comprehensive monitoring SaaS service platform based on a cloud computing architecture.

[0029] The application identifies whether there is a related physical machine with running abnormality by monitoring different physical machines in the cloud computing architecture in real time;

[0030] For the related physical machine of abnormal operation, first, numerical analysis is carried out to determine whether the abnormal judgment is caused by numerical fluctuation, then the load period of the cloud architecture is determined, the determined load period and the abnormal period of the physical machine are comprehensively analyzed to determine whether the related physical machine of abnormality exists load influence, through step-by-step analysis, the related physical machine of real abnormality is determined and displayed, the comprehensiveness in the monitoring process is improved, the maintenance operation of external maintenance personnel is facilitated, and the maintenance efficiency is reduced.

[0031] Then, the related physical machine with abnormality is analyzed to determine whether the related physical machine has linkage abnormality, so as to facilitate subsequent operation personnel to maintain the unit abnormality. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 It is a schematic diagram of the principle framework of the application.

[0033] Figure 2 It is a schematic diagram of the abnormal equipment determination of the application. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the application will be described clearly and completely in the embodiments of the application combined with the drawings, obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0035] Embodiment one

[0036] Please refer to Figure 1 The application provides a comprehensive monitoring SaaS service platform based on cloud computing architecture, including a data monitoring end, an abnormal equipment calibration end, a single-group abnormality analysis end, a multi-group abnormality analysis end, a total processing center and a display end.

[0037] The data monitoring end and the abnormal equipment calibration end input node are electrically connected, wherein the abnormal equipment calibration end is respectively electrically connected with the single-group abnormality analysis end or the multi-group abnormality analysis end input node, wherein the multi-group abnormality analysis end is electrically connected with the single-group abnormality analysis end input node, wherein the single-group abnormality analysis end is electrically connected with the total processing center input node, and the total processing center is electrically connected with the display end input node.

[0038] Among them, the data monitoring end monitors the working parameters of the physical machine associated with the cloud computing architecture, and transmits the real-time monitored physical machine working parameters to the abnormal equipment calibration end, wherein the specified parameter monitoring is carried out by the specified numerical sensor, and the sensor is a high-sensitive sensor with specified model.

[0039] The abnormal equipment calibration end checks the specific working parameters monitored by different physical machines, and based on the specific checking result, the abnormal physical machine is initially calibrated. Specifically, each different physical machine is provided with a standard checking interval when running, and based on the checking interval, whether the running state of the corresponding physical machine is normal is identified.

[0040] The abnormal physical machine is initially calibrated by the abnormal physical machine calibration end. Figure 2 The specific way of initially calibrating the abnormal physical machine includes:

[0041] The specific working parameters monitored in real time are calibrated as G i Where i represents different physical machines, and the working parameters G i is checked with the checking interval of the corresponding physical machine, where the checking interval is a preset interval, which is determined by the operator according to past experience.

[0042] If G The checking interval is calibrated as an abnormal time, and if the duration of the abnormal time exceeds 5 seconds, the physical machine is calibrated as an abnormal physical machine, otherwise, no calibration is performed.

[0043] If G i ∈ checking interval, no calibration is performed.

[0044] A group of monitoring periods T are determined, where T is a preset value, generally determined by the operator according to experience, the number of abnormal physical machines generated in the monitoring period T is determined, if there is only a single group of physical machines in the monitoring period T, a single group of abnormal analysis end is executed, if there are multiple groups of physical machines in the monitoring period T, a multiple group of abnormal analysis end is executed.

[0045] Specifically, different devices are equipped with different checking intervals, and the values generated according to the checking results can determine whether the values are abnormal, and then based on the duration of the abnormal physical machine, the abnormal physical machine is locked.

[0046] The single group of abnormal analysis end analyzes the values of the calibrated single group of abnormal physical machines to determine whether the abnormal physical machine is truly abnormal, and performs re-calibration. Specifically, when the device parameters of the corresponding physical machine are abnormal, either the input parameters cause the physical machine to have a load, resulting in abnormality, or the internal operation of the physical machine is abnormal, resulting in value abnormality.

[0047] The specific way of analyzing the values of the abnormal physical machine includes:

[0048] Determine a set of monitoring micro-periods, whose micro-period is a preset period, generally 3 min, record the output working parameters generated by the abnormal physical machine in this micro-period, and generate its output working interval according to the maximum value or minimum value of its output working parameters;

[0049] Determine the input parameters of the abnormal physical machine in the micro-period, confirm the output parameters corresponding to the input parameters from the past parameter data, and extract the minimum value and the maximum value from the extracted several output parameters to generate its standard interval;

[0050] Compare the output working interval with the standard interval to determine the intersection range, and determine the specific proportion value ZB of the intersection range in the output working interval. If ZB≥Y1, the abnormal physical machine is marked as a fluctuation physical machine, otherwise, the marking of the abnormal physical machine remains unchanged, and is transmitted to the total processing center;

[0051] Specifically, when numerically analyzing the abnormal physical machine, the output parameters of the corresponding physical machine are determined first, and the normal output parameters in the past data are recorded. By comparing the output parameters of the two, it can be determined whether the physical machine is abnormal when outputting parameters, and whether there is parameter fluctuation can also be effectively analyzed, ensuring the accuracy of abnormal marking.

[0052] Embodiment two

[0053] Among them, a plurality of abnormal analysis ends perform numerical analysis on a plurality of marked abnormal physical machines to determine whether there is an affected physical machine, and transmit the confirmed affected physical machine to the display end and display it through the display end. The specific way to determine whether there is an affected physical machine includes:

[0054] Determine the specific abnormal period of the abnormal physical machine in the monitoring period T, which is the specific period of numerical anomaly, that is, the period corresponding to the numerical value not belonging to the corresponding comparison interval. Cross-analyze the confirmed several specific abnormal periods to identify the intersection period. If there is only one group of intersection periods, the intersection period is marked as a pending period. If there are multiple groups of intersection periods, obtain the period minimum value and the period maximum value of the multiple groups of intersection periods to generate a pending period belonging to multiple groups of intersection periods. For example, if there are multiple groups of intersection periods, [2, 3], [3.5, 4] and [4, 7], the determined pending period is [2, 7];

[0055] Determine the load period of the cloud computing architecture in the monitoring period T, which is determined by the load state of the architecture, that is, when the cloud computing architecture is in a load situation, determine the load period based on the initial time point and the end time point of the load, cross analyze the load period and the pending period to determine the intersection range, and then determine the specific proportion ZZ of the intersection range in the load period, for example: if the load period is [1, 3], and if the pending period is [2, 5], the intersection range generated is [2, 3], and the specific proportion ZZ of the load period is 1 / 2;

[0056] If ZZ≥Y2, generate an affected signal, otherwise, do not generate any signal, and perform numerical analysis on the multiple abnormal physical machines through the single abnormal analysis end to determine whether they are real exceptions, where Y2 is a preset value, and its specific value is determined by the operator according to experience;

[0057] Based on the affected signal:

[0058] If there is only one group in the corresponding intersection period, the several abnormal physical machines associated with the intersection period are marked as affected physical machines, and are displayed through the display end;

[0059] If there are multiple groups in the corresponding intersection period, the intersection period that has an intersection range with the load period is marked as an affected period, and the several abnormal physical machines associated with the affected period are marked as affected physical machines, and are displayed through the display end. The intersection period that does not have an intersection range is marked as an unaffected period, and the abnormal physical machines associated with the unaffected period are analyzed numerically through the single abnormal analysis end to determine whether they are real exceptions;

[0060] Specifically, the so-called existence of abnormal physical machines associated with the period means that the abnormal period generated by the corresponding abnormal physical machine intersects with the period, that is, the period is obtained by intersecting the abnormal period of the associated physical machine, so there is a corresponding association;

[0061] This numerical analysis method can fully identify whether the generated abnormal physical machine is caused by the load of the cloud computing architecture, and improve the comprehensive monitoring effect of abnormal equipment.

[0062] Embodiment three

[0063] Among them, the total processing center determines whether there is a related abnormal physical machine of data flow based on the determined abnormal physical machine, if there is, lock the linkage group, and analyze whether the corresponding abnormal period of the linkage group is intersected, to determine whether there is a linkage abnormal machine group, wherein the specific way to determine includes:

[0064] From the determined several groups of abnormal physical machines, it is confirmed that there is a related abnormal physical machine for data flow, and it is marked as a linkage unit, and a single linkage unit only includes two groups of abnormal physical machines;

[0065] Determine the abnormal period of different abnormal physical machines in the linkage unit in the monitoring period T, and determine whether there is a cross period in the corresponding abnormal period:

[0066] If there is a cross period, determine the maximum proportion value of the cross period in different abnormal periods, if the maximum proportion value is greater than or equal to 90%, mark this linkage unit as a linkage abnormal unit, otherwise, do not mark;

[0067] If there is no cross period, do not mark;

[0068] The confirmed linkage abnormal unit and abnormal physical machine are displayed through the display end for external operators to view to determine the abnormal situation of the physical machine unit and take corresponding processing measures.

[0069] Specifically, after determining the abnormal physical machine unit, many abnormal units will cause synchronous abnormality of the next group due to the relationship of data flow, thereby causing linkage abnormality. In view of this kind of linkage abnormal unit, in order to facilitate the subsequent operation personnel to maintain the abnormal unit, it is necessary to determine the linkage abnormal unit and take corresponding control measures in time to improve the overall monitoring effect of the platform.

[0070] Embodiment four

[0071] In the specific implementation process of this embodiment, all the implementation processes of the above three embodiments are included;

[0072] As described above, the different physical machines in the cloud computing architecture are monitored in real time to identify whether there is a related physical machine with running abnormality;

[0073] For the related physical machine with running abnormality, first, numerical analysis is performed to determine whether the abnormality is caused by numerical fluctuation. Then, the load period of the cloud architecture is determined, and the determined load period and the abnormal period of the physical machine are comprehensively analyzed to determine whether the related physical machine with abnormality exists load influence. Through step-by-step analysis, the related physical machine with real abnormality is determined and displayed to improve the comprehensiveness in the monitoring process, facilitate the maintenance operation of external maintenance personnel, and reduce the maintenance efficiency;

[0074] Then, the related physical machine with abnormality is analyzed to determine whether it has linkage abnormality, which facilitates the subsequent operation personnel to maintain the abnormal unit.

[0075] Some data in the above formula are dimensionless for numerical calculation, and the contents not described in detail in the specification are all prior art known by those skilled in the art.

[0076] The above examples are only used to illustrate the technical method of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. A comprehensive monitoring SaaS service platform based on cloud computing architecture, characterized in that: include: The data monitoring terminal monitors the operating parameters of the physical machines associated with the cloud computing architecture and transmits the real-time monitored operating parameters of the physical machines to the abnormal device calibration terminal. The malfunctioning equipment calibration terminal performs numerical verification of the specific operating parameters monitored by different physical machines. Based on the verification results, it performs initial calibration of the malfunctioning physical machine. Specific methods include: The specific working parameters of real-time monitoring are calibrated as G. i Where i represents different physical machines, and the corresponding physical machine's operating parameters G i The check interval is compared with the corresponding physical machine's check interval, which is a preset interval; If G i ∉ Check the interval and mark the corresponding time as an abnormal time. If the duration of the abnormal time exceeds 5 seconds, mark this physical machine as an abnormal physical machine; otherwise, do not mark it. A set of monitoring periods T is determined, where T is a preset value. The number of abnormal physical machines generated in this monitoring period T is determined. If there is only a single group of physical machines in this monitoring period T, then the single group of abnormal analysis terminal is executed. If there are multiple groups of physical machines in this monitoring period T, then the multiple groups of abnormal analysis terminal are executed. The single-group anomaly analysis terminal performs numerical analysis on the calibrated single-group anomalous physical machines to determine whether the anomalous physical machine is truly anomalous, and then recalibrates it. The specific method is as follows: A set of monitoring microcycles is determined, and the microcycle is a preset period. The output working parameters generated by this abnormal physical machine within this microcycle are recorded, and the output working range is generated based on the maximum or minimum value within the output working parameters. Determine the input parameters of the corresponding abnormal physical machine within this micro-cycle, and confirm the output parameters corresponding to this input parameter from past parameter data. Then, extract the minimum and maximum values ​​from the extracted output parameters to generate its standard range. The output working range is compared with the standard range to determine the intersection range, and the specific proportion ZB of this intersection range in the output working range is determined. If ZB≥Y1, this abnormal physical machine is labeled as a fluctuating physical machine. Multiple anomaly analysis terminals perform numerical analysis on the identified multiple sets of anomalous physical machines to determine whether any physical machines are affected. The confirmed affected physical machines are then transmitted to the display terminal for presentation. Specific methods include: Determine the specific abnormal time period corresponding to the abnormal physical machine within this monitoring period T. The abnormal time period is the specific time period of numerical abnormality. Cross-analyze the confirmed specific abnormal time periods to identify the cross-time periods. If there is only one set of cross-time periods, mark this cross-time period as the undetermined time period. If there are multiple sets of cross-time periods, obtain the minimum and maximum time periods of multiple sets of cross-time periods to generate the undetermined time periods belonging to multiple sets of cross-time periods. The load period of this cloud computing architecture in this monitoring period T is determined. The load period is determined by the state of this architecture under load. The load period is cross-analyzed with the undetermined period to determine the cross-range. Then, the specific proportion ZZ of the cross-range is determined to be in the load period. If ZZ≥Y2, an affected signal is generated; otherwise, no signal is generated. Numerical analysis is performed on multiple abnormal physical machines through a single abnormal analysis terminal to determine whether they are truly abnormal, where Y2 is a preset value. Based on the affected signals: If there is only one corresponding cross-period, then the several abnormal physical machines associated with this cross-period will be marked as affected physical machines and displayed through the display terminal. If there are multiple sets of corresponding cross-period periods, the cross-period periods that have an overlap with the load period are marked as affected periods, and the several sets of abnormal physical machines associated with the affected periods are marked as affected physical machines and displayed through the display terminal. The cross-period periods that do not have an overlap are marked as unaffected periods. The abnormal physical machines associated with the unaffected periods are numerically analyzed through the single-group anomaly analysis terminal to determine whether they are truly abnormal.

2. The comprehensive monitoring SaaS service platform based on cloud computing architecture according to claim 1, characterized in that, The G i When the interval is checked, no calibration is performed.

3. The comprehensive monitoring SaaS service platform based on cloud computing architecture according to claim 1, characterized in that, When the specific percentage value ZB < Y1, the calibration of this abnormal physical machine remains unchanged and is transmitted to the central processing center.

4. The comprehensive monitoring SaaS service platform based on cloud computing architecture according to claim 3, characterized in that, The central processing center, based on the identified abnormal physical machines, determines whether there are related abnormal physical machines for data flow. If so, it locks the linked units and analyzes whether the corresponding abnormal time periods of the linked units overlap to determine whether there are linked abnormal units.

5. The comprehensive monitoring SaaS service platform based on cloud computing architecture according to claim 4, characterized in that, The central processing center determines whether there are any abnormally linked units using the following methods: From the identified groups of abnormal physical machines, identify the relevant abnormal physical machines that have data flow and mark them as linked units; Determine the abnormal time periods for different abnormal physical units of the linked units within this monitoring cycle T, and determine whether there are overlapping time periods among the corresponding abnormal time periods: If there are overlapping time periods, determine the maximum percentage of the overlapping time periods that fall within different abnormal time periods. If the maximum percentage is ≥90%, mark this linked unit as an abnormal linked unit; otherwise, do not mark it. If there is no overlap in time periods, no calibration is required; The confirmed abnormal linkage units and abnormal physical machines are displayed through the display terminal.

Images