Cloud database health assessment method and device, equipment, storage medium and computer program product

Abstract

The invention discloses a cloud database health assessment method, device and equipment, a storage medium and a computer program product, and the method comprises the steps: obtaining a first real-time parameter of a cloud server, and generating a health index of the cloud server according to the first real-time parameter; obtaining a second real-time parameter of the cloud database, and generating a performance index of the cloud database according to the second real-time parameter; adjusting the performance index based on the business historical data to obtain a corrected performance index; and determining the health state of the cloud database based on the health index and the correction performance index. Through the quantitative evaluation method, the health state of the cloud database can be monitored and evaluated more accurately, so that the stability and reliability of the database are ensured. Therefore, service interruption or delay caused by database performance problems can be reduced, and the beneficial effects of optimizing resource configuration and improving fault prevention capability are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud technology, and in particular to a cloud database health assessment method, device, equipment, storage medium and computer program product. BACKGROUND

[0002] In a cloud environment, ensuring the health status of database services is a core task for maintaining system stability and providing high-quality services. Health status detection of cloud databases is a critical step in this process, which involves comprehensive assessment and monitoring of the running state, performance indicators, error logs, and backup integrity of database services to ensure their normal operation and optimal performance. Many cloud provider database products rely on third-party monitoring tools to detect product health status. For example, open-source monitoring systems can be used to monitor the performance and status of databases through plugins and extensions; powerful database monitoring functions can be achieved by combining visualization and dashboard tools with open-source monitoring and alerting tools.

[0003] However, the current mainstream cloud database health status exploration scheme generally relies on index visualization dashboards, which are designed to help operations personnel intuitively understand various indicators of cloud products. Through the cloud monitoring platform developed based on mainstream tools, operations personnel manually assess resources based on visual data to identify potential optimization spaces and take appropriate optimization measures accordingly. However, this method still faces some challenges, such as weak visualization of cloud database health status and failure to consider the impact of business cycle behavior, resulting in unnecessary monitoring alarms. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a cloud database health assessment method, device, equipment, storage medium and computer program product, which solves the problem of high false alarm rate and dependence on manual health status assessment in the prior art, and realizes automatic and precise cloud database health assessment.

[0005] To achieve the above purpose, the embodiments of the present application provide a cloud database health assessment method, comprising: obtaining a first real-time parameter of a cloud server, and generating a health index of the cloud server based on the first real-time parameter; obtaining a second real-time parameter of a cloud database, and generating a performance index of the cloud database based on the second real-time parameter; adjusting the performance index based on business historical data to obtain a corrected performance index; determining the health status of the cloud database based on the health index and the corrected performance index.

[0006] As an improvement of the above-mentioned scheme, the first real-time parameter of the cloud server is obtained, and the health index of the cloud server is generated according to the first real-time parameter, comprising: The first real-time parameter of the cloud server is obtained; wherein the first real-time parameter includes at least one of the CPU usage, memory usage and storage space usage of the cloud server; According to the comparison result of the first real-time parameter and the first threshold set, the health index of the cloud server is generated; When any of the first real-time parameters exceeds the first threshold set, a first discriminant value is outputted; or, When all the first real-time parameters do not exceed the first threshold set, a second discriminant value is outputted.

[0007] As an improvement of the above-mentioned scheme, the second real-time parameter of the cloud database is obtained, and the performance index of the cloud database is generated according to the second real-time parameter, comprising: The second real-time parameter of the cloud database is obtained; wherein the second real-time parameter includes at least one of the CPU usage, memory usage, IOPS usage, connection number, slow query number and thread running number of the cloud database; Based on the second real-time parameter and a preset adjustment strategy, a second threshold set is generated; Based on the second threshold set, the second real-time parameter is normalized to generate the performance index of the cloud server.

[0008] As an improvement of the above-mentioned scheme, the performance index is adjusted based on the business historical data to obtain a corrected performance index, comprising: A periodic correction factor is generated based on the business historical data; The performance index is dynamically corrected according to the correction factor to obtain the corrected performance index.

[0009] As an improvement of the above-mentioned scheme, the periodic correction factor is generated based on the business historical data, comprising: According to the business historical data, the periodic fluctuation characteristics of business load are analyzed; Based on the periodic fluctuation characteristics, a correction function synchronized with the periodicity of the business load is constructed.

[0010] As an improvement of the above-mentioned scheme, the health state of the cloud database is determined based on the health index and the corrected performance index, comprising: The health index and the corrected performance index are fused to generate a health state quantization value; According to the numerical interval of the health state quantization value, it is mapped to a health state level system. The health state level system at least includes an inactivation state and a health state.

[0011] The embodiment of the present application also provides a cloud database health evaluation device, which comprises: a health calculation module configured to acquire a first real-time parameter of a cloud server, and generate a health index of the cloud server according to the first real-time parameter; a performance calculation module configured to acquire a second real-time parameter of a cloud database, and generate a performance index of the cloud database according to the second real-time parameter; a dynamic correction module configured to adjust the performance index based on business historical data to obtain a corrected performance index; a state confirmation module configured to determine a health state of the cloud database based on the health index and the corrected performance index.

[0012] The embodiment of the present application also provides a terminal device, which comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, and the processor executes the computer program to implement the cloud database health evaluation method.

[0013] The embodiment of the present application also provides a computer readable storage medium, which comprises a stored computer program, wherein the computer program controls a device where the computer readable storage medium is located to execute the cloud database health evaluation method when the computer program runs.

[0014] The embodiment of the present application also provides a computer program product, which comprises a computer program or computer instructions, and the computer program or the computer instructions implement the cloud database health evaluation method when executed by a processor.

[0015] Compared with the prior art, the cloud database health evaluation method, device, equipment, storage medium and computer program product provided by the embodiment of the present application have the beneficial effects that the health index of the cloud server is calculated based on the real-time parameter of the cloud server to determine the health state of the cloud server, then the performance index of the cloud database is determined based on the real-time parameter of the cloud database, the corrected performance index is obtained by optimizing according to the influence of the periodic characteristics in the customer business historical data, and finally the safety level of the cloud database health state is divided by combining the cloud database performance parameter standard requirement and the cloud server health state. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 is a flowchart of a preferred embodiment of the cloud database health evaluation method provided by the present application; Figure 2 is a structural schematic diagram of a preferred embodiment of a cloud database health assessment device provided by the present application; Figure 3 is a structural schematic diagram of a preferred embodiment of a terminal device provided by the present application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0018] Please refer to Figure 1 , Figure 1 is a flow schematic diagram of a preferred embodiment of a cloud database health assessment method provided by the present application. The cloud database health assessment method comprises: Step S100, acquiring a first real-time parameter of a cloud server, and generating a health index of the cloud server according to the first real-time parameter; Step S200, acquiring a second real-time parameter of a cloud database, and generating a performance index of the cloud database according to the second real-time parameter; Step S300, adjusting the performance index based on business historical data to obtain a corrected performance index; Step S400, determining a health state of the cloud database based on the health index and the corrected performance index.

[0019] In actual implementation, the present application determines the state of the nodes of the cloud service machine based on the detection mode of the health state of the cloud server nodes and the index. Then, based on the performance parameters of the cloud database, the threshold value of the performance index is dynamically adjusted according to real-time data by using a machine learning algorithm, so as to reduce the false positive rate and improve the detection accuracy. The cloud database health state index calculation method is proposed, and the abnormal fluctuation of the performance index is effectively weakened. The future performance index trend is predicted based on time series analysis, so as to predict in advance possible failures such as similar high-availability switching failures. The health state index calculation method is optimized in view of the influence of the periodic characteristics of the customer business. Finally, the cloud database health state is divided into security levels in combination with the performance parameter standard requirements of the cloud database and the health state of the cloud server.

[0020] Specifically, the application is based on cloud server node performance index to build a health state detection method discriminant index, to express the health status of the cloud server node in the form of a single index, so as to facilitate the operation and maintenance personnel to quickly locate the server environment abnormal problem. Then, based on the CPU usage, memory usage, IOPS and other performance parameters, the machine learning algorithm is used to dynamically adjust the threshold value of the performance index according to the real-time data; the database performance quantization model is constructed to obtain the database performance index PI, and the future performance index trend is predicted based on time series analysis, so as to predict in advance the possible failure such as high availability switching failure; the health state detection method discriminant index and the customer business characteristic index CI are combined to propose a cloud database health state calculation method, to obtain the health state quantization index MS, and the corresponding health state evaluation grade is given.

[0021] In step S100, the first real-time parameter of the cloud server is obtained, and the health index of the cloud server is generated according to the first real-time parameter.

[0022] In the above embodiment, the application uses at least one of the CPU usage, memory usage and space usage of the cloud server as the first real-time parameter, and combines the server job standard performance threshold value as the first threshold set, for example, the comparison parameter including one or more real-time parameters, to construct the health state detection method discriminant index, and express the health status of the cloud server node in the form of a single index, so as to facilitate the operation and maintenance personnel to quickly locate the server environment abnormal problem.

[0023] Specifically, the health status of the cloud server is the basis for the normal operation of the cloud database, and the application uses the CPU usage, memory usage and space usage as the first real-time parameter to construct the health state detection method discriminant index (i.e. the above health index), to directly describe the health status of the cloud server in the form of index.

[0024] The control range of the cloud server CPU usage, memory usage and space usage is expressed, as shown in the following formula 1: (Formula 1) In the formula, represents the performance range of the cloud server in the normal operation state; , , z c are the job threshold values of the cloud server in the aspects of CPU usage, memory usage and space usage, respectively.

[0025] The satisfaction degree between the actual usage of the cloud server CPU, memory and space and the maximum control range is expressed in the form of index, as shown in the following formula 2: (Formula 2) In the formula, is the satisfaction index between the real-time index of the cloud node performance and the threshold of the large job, represents the actual usage rate of CPU, memory, and space of the cloud node, and the position of the ship in the fixed coordinate system. When , is 0.

[0026] Based on the satisfaction index , the system control mode discrimination index is constructed, as shown in the following formula 3: (Formula 3) The relationship between the health status of the cloud server node and the cloud database health state discrimination method is shown in Table 1. Through the relationship between the actual usage rate of CPU, memory, and space of the cloud server and the maximum control range (i.e. the first threshold set), it is determined whether the node performance index is within the allowed maximum threshold range. The satisfaction degree between the two is expressed by an index. When the cloud node is in a healthy state, i.e. the node performance index is within the threshold range allowed by the system normal work, the satisfaction degree index between the two is calculated -1, and the cloud database health state detection focuses on the performance index of the database. The control mode index value is 1; otherwise, when the cloud node state is abnormal, the satisfaction degree index is greater than -1, and the cloud database health state detection focuses on the cloud server environment problem, and is -1.

[0027] Table 1 Relationship between cloud server health state and cloud database health state discrimination method

[0028] In step S200, the second real-time parameter of the cloud database is obtained, and the performance index of the cloud database is generated according to the second real-time parameter.

[0029] In this embodiment, the present application proposes a cloud database performance quantification model (taking cloud MySQL as a specific type of cloud database). The cloud database performance quantification model obtains a performance quantification index by using six performance parameters as second real-time parameters. First, the input parameters of the performance quantification model are determined by combining the actual usage rate of the database job and the job performance index threshold requirement. Second, according to the relationship between the input parameters and the cloud database health state, the performance parameter index is calculated, and the six input performance parameters are normalized to the cloud database performance quantification index PI.

[0030] Specifically, this embodiment aims to achieve quantitative analysis of database performance based on various typical database performance parameters by constructing a cloud database performance quantification model. Combining the constraints of database operations and typical performance parameter standards, appropriate and effective performance parameters are selected as input parameters for the quantification model. Taking cloud MySQL as an example, the input module of the quantification model includes six database performance-related indicators such as CPU utilization and IOPS utilization, used to construct the cloud database performance index. The cloud database performance evaluation standards and performance parameter information are shown in Table 2 below.

[0031] Table 2 Performance Evaluation Criteria and Second Real-Time Parameters for Cloud Database (Cloud MySQL)

[0032] Table 2 introduces the specified values ​​of the six input parameters in the cloud database performance quantification model. For example, these can be the range of commonly used performance standard values ​​adopted by cloud vendors, the initial values ​​of the performance parameters involved in the quantification model, and the real-time thresholds.

[0033] To improve the accuracy of anomaly detection in database performance metrics, an adaptive threshold adjustment method can be used to dynamically adjust the thresholds of performance metrics. This method primarily utilizes reinforcement learning algorithms, adjusting the performance metric thresholds based on the current state and a Q-value table (which stores the value of each state-action pair). Bayesian optimization is employed, using a Gaussian process model and an acquisition function to optimize the performance metric thresholds.

[0034] Q-learning is used to obtain the threshold adjustment strategy. Define the state. s This represents the current values ​​of various performance metrics, such as CPU utilization, memory utilization, disk I / O, and network latency; actions. a This represents an operation that adjusts the threshold, such as increasing or decreasing the threshold; initializing the Q-table to store the value of each state-action pair; at each time step, the agent observes the current state. s Select Action a And execute, observe immediate rewards r and the next state Update the Q-value table and iterate until convergence; select the optimal action based on the Q-value table. a The threshold for performance indicators is dynamically adjusted, as shown in Formula 4. (Formula 4) In the formula, :state s Take action below a of Q The value represents the worth of the current strategy; Learning rate controls the step size for updates; r: reward value, take action a : immediate reward after : discount factor, indicating the degree of attenuation of future rewards : take action a : next state after : next action in state : next action in state

[0035] By using a reinforcement learning algorithm, the threshold of the performance indicator is adjusted according to the current state and the Q-value table to obtain a second threshold set. In order to further optimize the threshold of the performance indicator, a Bayesian optimization method is adopted, that is, a Gaussian process model and an acquisition function are used. The threshold of the performance indicator to be optimized is selected, such as the CPU usage threshold, the memory usage threshold, etc.; the objective function f ( x) is defined, which represents the detection accuracy or system performance under a given threshold; a set of thresholds and corresponding objective function values are initialized by using random sampling or existing data; the Gaussian process model is trained using the initial data set to obtain the surrogate function f ( x) and uncertainty estimation; the acquisition function (such as expected improvement, probability improvement, or confidence upper bound) is selected to find the optimal threshold on the surrogate function; experiments are performed according to the optimal threshold to obtain a new objective function value, and the data set is updated; the Gaussian process model is repeatedly trained and the acquisition function is optimized until convergence. Respectively as shown in formula 5 and formula 6.

[0036] (Formula 5) In the formula, GP (Gaussian Process) Gaussian process, is a mean function, is a kernel function, which represents the similarity between any two points x and .

[0037] (Formula 6) In the formula, represents the current optimal value; is an indicator function, which takes 1 when , otherwise 0.

[0038] In another embodiment, based on the relationship between each performance number and the health state of the cloud database, the input parameters are normalized to obtain a stability parameter index . In order to make different performance indicators comparable and reduce the influence of anomaly detection, the negative correlation between the input parameters and the health state of the cloud database is considered, and the job expectation standard value interval and dynamic adjustment threshold of each performance parameter as shown in Table 2 are adopted Thi , the normalized method is constructed as shown in Equation 7.

[0039] (Equation 7) In the formula, represents the input parameter normalization index, k is the mooring ship stability index coefficient, the value is taken as 0.5, indicating that 50% of the performance parameters meet the requirement standard. represents the corresponding input parameter maximum operation threshold, represents the threshold taken by the proposed quantization model Th i , refers to the minimum operation threshold when all input parameters meet the performance requirements, and the specific value is shown in Table 2 above.

[0040] To improve the sensitivity of fault perception, the trend of unopened performance indicators is predicted based on time series analysis, so as to predict similar high-availability switching failures and other faults in advance. ARIMA (AutoRegressive Integrated Moving Average) combined with AR (AutoRegressive) and MA (Moving Average) is used for time series analysis, and the data is made stationary through difference processing. Collect historical data of various performance indicators, such as CPU usage, memory usage, disk I / O, network latency, etc.; difference processing is performed on the data to eliminate trends and seasonality, making it stationary; train the ARIMA model according to the historical data. The three parameters p , d and q of the ARIMA model can be selected using automated methods (such as AIC, BIC criteria); input the latest performance indicator data into the ARIMA model to predict the performance indicator values in the future period. For example, predict the values k at the next time points (the actual performance parameter indices ). The core formula of ARIMA is shown in Equation 8.

[0041] (Equation 8) In the formula, X t is the predicted value of the performance parameter t at the current time ; B is the lag operator; d is the difference number, making the time series stationary, when d =1, ; c is the constant term; represents the influence of the value at the past n time point on the current value. represents the influence of the error at the past n time point on the current value. is a random error term at the current time point. p represents the order of the autoregressive part. q represents the order of the moving average part.

[0042] The performance parameter index is normalized according to the weighted average method to obtain the cloud database performance index PI , as shown in formula 9.

[0043] (Formula 9) In the formula, N is the total number of input parameters involved in the ship stability quantitative model, and the value is 6; represents the influence of the input parameter on the health status of the cloud database, .

[0044] Combined with the prediction algorithm, the cloud database performance index at the future t+k time point is as shown in the following formula 10: (Formula 10) It can be understood that the adaptive threshold adjustment algorithm in the above embodiment can dynamically adjust the threshold of the performance index according to real-time data, thereby improving the accuracy and timeliness of the early warning, and reducing false positives and false negatives. In addition, using time series analysis to predict future performance trends can also help developers to deploy solutions in advance.

[0045] In step S300, the performance index is adjusted based on the business historical data to obtain a corrected performance index. That is, a periodic correction factor is generated based on the business historical data, and the performance index is dynamically corrected according to the correction factor to obtain the corrected performance index.

[0046] Specifically, according to the above business historical data, the periodic fluctuation characteristics of the business load are analyzed; and based on the above periodic fluctuation characteristics, a correction function synchronized with the business load period is constructed.

[0047] It can be understood that the periodic characteristics of customer business can directly cause the performance index of the database to fluctuate, thereby affecting the evaluation of the health status of the database. In order to reduce the performance index fluctuation caused by the periodic characteristics of the business and thereby cause the abnormal alarm of the health status of the database, the application proposes an automatic correction index based on a time period CIai , to correct the cloud database performance index, realize the health state quantitative evaluation of the cloud database based on customer business cycle behavior and performance index. The present application uses a composite periodic function to represent the correction index CI ai , as shown in formula 11.

[0048] CI ai =T ai sin( tω + )+R ai cos( tω + θ ) (formula 11) In the above formula 11, the correction index CI ai Input parameters based on the business cycle of the correction index; T ai and R ai is the weight coefficient related to the performance parameter , defined by data-driven and expert knowledge fusion; ω、 , θ is a constant to represent the periodic characteristics of the performance parameter . Combined with the customer business cycle characteristics index, the cloud database performance index PI , as shown in formula 12.

[0049] (formula 12) In step S400, the health state of the cloud database is determined based on the health index and the corrected performance index.

[0050] Among them, the present application generates a health state quantitative value by fusing the above health index and the above corrected performance index; and then according to the numerical interval of the health state quantitative value, it is mapped to a health state level system; wherein the health state level system at least includes: inactivation state and health state.

[0051] Specifically, according to the cloud database health state evaluation standard, it is primarily ensured that the cloud node where the cloud database is located is available; secondly, control each performance index of the cloud database within the appropriate working range. Therefore, the health state quantitative evaluation of the cloud database of the ship mooring state needs to consider the environment (cloud server) where the cloud database is located and its each performance index. In the health state detection mode based on the cloud node each performance index, the correction index Based on the research, combined with the cloud database performance index PI, construct a cloud database health state quantitative evaluation method. Cloud database health state index MS The calculation formula is shown in formula 13.

[0052] (Formula 13) The above health state index MS can be Distinguish the availability of the cloud server, when MS is negative , indicating that the server node state of the cloud database is abnormal; when MS is positive , indicating that the cloud server node is normal operation; MS Can measure the pros and cons of cloud database performance, on the basis of normal operation of cloud server, health state index MS And performance index PI There is a linear mapping relationship between them. Therefore, the cloud database health state index can reflect the availability information of the cloud server and the performance information of the cloud database.

[0053] Further, in this step, in order to more intuitively understand the health state of the cloud database, the health state detection method discriminant index And cloud database performance index PI , the health state evaluation of cloud database is divided into 2 categories and 6 state levels, as shown in the following table 3.

[0054] Table 3 Health state evaluation level of cloud database

[0055] The first type: for cloud server node state, when the cloud server node state is abnormal, it cannot guarantee the normal operation of the cloud database, the health state detection method discriminant index , only contains one health state level: cloud database health state index , the health state is rated as the first level-"inactivated".

[0056] The second type: on the basis of meeting the requirements of normal operation of cloud server node, consider the performance of cloud database. In this case, the health state detection method discriminant index , based on formula 13, there is MS = PI , the health state index ( ) represents the performance of the cloud database; according to the number of performance parameters conforming to the specified value shown in table 2 above, and introducing the concept of performance margin (according to the existing data analysis, the performance margin can be set to 20%-30%), the health of the cloud database is divided into 5 evaluation levels, and the health state level is from the worst "danger" to the best "quite stable".

[0057] When the cloud database performance index , all performance parameters meet the specified standards shown in Table 2, and the database performance meets the minimum requirements. To ensure that the database has sufficient performance, the cloud database requires a certain margin. According to the existing data analysis, the performance margin can be set to 30%.

[0058] Therefore, the detection method provided by the present application has good scalability and can combine various performance indicators of the cloud database, including CPU usage, memory usage, disk I / O, and database connection number. The health status of the cloud database is represented in a unique exponential form, which is more intuitive. In addition, the present application quickly evaluates the health status of the database and server node by combining multiple performance parameters, provides a quantitative index of the health status, and thus facilitates the operation and maintenance personnel to quickly locate the problem and identify the key problem affecting the primary and backup switching, and take prompt measures.

[0059] Correspondingly, the present application also provides a cloud database health evaluation device capable of realizing all processes of the cloud database health evaluation method in the above embodiments.

[0060] Please refer to Figure 2 , Figure 2 is a structural schematic diagram of a preferred embodiment of a cloud database health evaluation device provided by the present application. The cloud database health evaluation device comprises: a health calculation module 201 configured to obtain a first real-time parameter of a cloud server and generate a health index of the cloud server according to the first real-time parameter; a performance calculation module 202 configured to obtain a second real-time parameter of a cloud database and generate a performance index of the cloud database according to the second real-time parameter; a dynamic correction module 203 configured to adjust the performance index based on business historical data to obtain a corrected performance index; a state confirmation module 204 configured to determine the health status of the cloud database based on the health index and the corrected performance index.

[0061] Preferably, the health calculation module 201 is specifically configured to: obtain the first real-time parameter of the cloud server; wherein the first real-time parameter comprises at least one of the CPU usage, the memory usage and the storage space usage of the cloud server; generate the health index of the cloud server according to the comparison result of the first real-time parameter and a first threshold set; wherein when any of the first real-time parameters exceeds the first threshold set, a first discrimination value is output; or output a second discriminant value when all the first real-time parameters do not exceed the first threshold set.

[0062] Preferably, the performance calculation module 202 is specifically configured to: obtain the second real-time parameter of the cloud database; wherein the second real-time parameter comprises at least one of CPU usage, memory usage, IOPS usage, connection number, slow query number and thread running number of the cloud database; generate a second threshold set based on the second real-time parameter and a preset adjustment strategy; perform normalization processing on the second real-time parameter based on the second threshold set, to generate the performance index of the cloud server.

[0063] Preferably, the dynamic correction module 203 is specifically configured to: generate a periodic correction factor based on the business historical data; perform dynamic correction on the performance index according to the correction factor, to obtain the corrected performance index.

[0064] Preferably, the generating a periodic correction factor based on the business historical data comprises: analyze periodic fluctuation characteristics of business load according to the business historical data; construct a correction function synchronized with the periodicity of the business load based on the periodic fluctuation characteristics.

[0065] Preferably, the state confirmation module 204 is specifically configured to: fuse the health index and the corrected performance index, to generate a health state quantization value; map to a health state level system according to a numerical interval of the health state quantization value; wherein the health state level system at least comprises: inactivated state and healthy state.

[0066] In specific implementation, the working principle, control flow and technical effects of the cloud database health evaluation device provided by the embodiments of the present application correspond to the cloud database health evaluation method in the above embodiments, and will not be repeated here.

[0067] Please refer to Figure 3 , Figure 3 is a structure schematic diagram of a preferred embodiment of a terminal device provided by the present application. The terminal device comprises a processor 301, a memory 302, and a computer program stored in the memory 302 and configured to be executed by the processor 301, and the processor 301 implements the cloud database health evaluation method described in any of the above embodiments when executing the computer program.

[0068] Preferably, the computer program can be split into one or more modules / units (such as computer program 1, computer program 2, …), which are stored in the memory 302 and executed by the processor 301 to accomplish the present application. The one or more modules / units can be a series of computer program instruction segments capable of accomplishing a specific function, which are used to describe the execution process of the computer program in the terminal device.

[0069] The processor 301 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor 301 can also be any conventional processor. The processor 301 is the control center of the terminal device, and connects various parts of the terminal device through various interfaces and lines.

[0070] The memory 302 mainly includes a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function, etc., and the data storage area can store related data, etc. In addition, the memory 302 can be a high-speed random access memory, and can also be a non-volatile memory, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., or the memory 302 can also be other volatile solid-state storage devices.

[0071] It should be noted that the above terminal device can include, but is not limited to, a processor, a memory, and those skilled in the art can understand that Figure 3 The structural diagram of the terminal device is only an example of the above terminal device, and does not constitute a limitation on the above terminal device, and can include more or fewer components than the diagram, or combine certain components, or different components.

[0072] The embodiment of the present application also provides a computer readable storage medium, which comprises a stored computer program, wherein the computer program controls a device where the computer readable storage medium is located to execute the cloud database health assessment method according to any one of the above embodiments when the computer program is running.

[0073] The embodiment of the present application also provides a computer program product, which comprises a computer program or computer instructions, and the computer program or the computer instructions realize the cloud database health assessment method according to any one of the above embodiments when executed by a processor.

[0074] Compared with the prior art, the present application can more accurately monitor and evaluate the health status of the cloud database through the quantitative evaluation method, thereby ensuring the stability and reliability of the database. This helps to reduce business interruption or delay caused by database performance problems, and improves customer satisfaction with cloud services. Secondly, based on the evaluation results, cloud service providers can more accurately predict and plan resource requirements, and realize the optimal allocation of resources. This not only improves resource utilization and reduces operating costs, but also better meets customer needs and improves service quality. Therefore, the present application can realize quantitative and qualitative analysis of the health status of the cloud database by constructing a quantitative evaluation method based on cloud database performance indicators, cloud server status and customer business cycle characteristics, thereby optimizing resource allocation and improving fault prevention capability.

[0075] In summary, the embodiment of the present application provides a cloud database health assessment method, device, equipment, storage medium and computer program product, which can combine various performance indicators of the cloud database, including CPU usage, memory usage, disk I / O, database connection number, etc., and use a unique exponential form to represent the health status of the cloud database, more directly representing the health status of the cloud database. The present application also introduces a time series prediction model, and dynamically adjusts the health status index according to the prediction result, realizes forward-looking detection and preventive maintenance, in addition, the present application realizes adaptive threshold adjustment through reinforcement learning and Bayesian optimization, dynamically adjusts the threshold of each performance indicator according to the current environment and historical data, improves the accuracy and flexibility of abnormal detection, it should be noted that the evaluation method provided by the present application also includes consideration of the periodic characteristics of customer business, and realizes dynamic change of the cloud database health status evaluation, improves the accuracy and availability of the evaluation.

[0076] It should be noted that the system embodiments described above are only illustrative, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e. can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment according to actual needs. In addition, the connection relationship between the modules in the system embodiment provided by the present application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0077] The above is the preferred embodiment of the present application. It should be noted that those skilled in the art can make several improvements and refinements without departing from the principles of the present application, and these improvements and refinements are also considered within the scope of protection of the present application.

Claims

1. A cloud database health assessment method, characterized in that, include: Obtain the first real-time parameters of the cloud server, and generate the health index of the cloud server based on the first real-time parameters; Obtain the second real-time parameter of the cloud database, and generate the performance index of the cloud database based on the second real-time parameter; The performance index is adjusted based on historical business data to obtain a corrected performance index. The health status of the cloud database is determined based on the health index and the corrected performance index.

2. The cloud database health assessment method as described in claim 1, characterized in that, The step of obtaining the first real-time parameters of the cloud server and generating the health index of the cloud server based on the first real-time parameters includes: Obtain the first real-time parameter of the cloud server; wherein the first real-time parameter includes at least one of the cloud server's CPU utilization, memory utilization, and storage space utilization; The health index of the cloud server is generated based on the comparison result between the first real-time parameter and the first threshold set. Wherein, when any of the first real-time parameters exceeds the first threshold set, a first discrimination value is output; or, When all the first real-time parameters do not exceed the first threshold set, output the second discrimination value.

3. The cloud database health assessment method as described in claim 1, characterized in that, The step of obtaining the second real-time parameter of the cloud database and generating the performance index of the cloud database based on the second real-time parameter includes: Obtain the second real-time parameters of the cloud database; wherein the second real-time parameters include at least one of the following: CPU utilization, memory utilization, IOPS utilization, number of connections, number of slow queries, and number of running threads of the cloud database; Based on the second real-time parameters and the preset adjustment strategy, a second threshold set is generated; The second real-time parameter is normalized based on the second threshold set to generate the performance index of the cloud server.

4. The cloud database health assessment method as described in claim 1, characterized in that, The process of adjusting the performance index based on historical business data to obtain a revised performance index includes: A periodic correction factor is generated based on the aforementioned historical business data; The performance index is dynamically corrected based on the correction factor to obtain the corrected performance index.

5. The cloud database health assessment method as described in claim 4, characterized in that, The generation of periodic correction factors based on the historical business data includes: Based on the aforementioned historical business data, analyze the periodic fluctuation characteristics of the business load; Based on the aforementioned periodic fluctuation characteristics, a correction function synchronized with the business load cycle is constructed.

6. The cloud database health assessment method as described in claim 1, characterized in that, Determining the health status of the cloud database based on the health index and the corrected performance index includes: By combining the health index and the modified performance index, a quantitative value of health status is generated. Based on the numerical range of the quantified health status values, a health status level system is mapped. The health status level system includes at least two states: inactive state and healthy state.

7. A cloud database health assessment device, characterized in that, include: The health calculation module is configured to obtain the first real-time parameters of the cloud server and generate a health index of the cloud server based on the first real-time parameters. The performance calculation module is configured to obtain the second real-time parameters of the cloud database and generate the performance index of the cloud database based on the second real-time parameters. The dynamic correction module is configured to adjust the performance index based on historical business data to obtain a corrected performance index; The status confirmation module is configured to determine the health status of the cloud database based on the health index and the corrected performance index.

8. A terminal device, characterized in that, The system includes a processor and a memory, wherein the memory stores a computer program and the computer program is configured to be executed by the processor, wherein the processor, when executing the computer program, implements the cloud database health assessment method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the cloud database health assessment method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program or computer instructions, which, when executed by a processor, implement the cloud database health assessment method as described in any one of claims 1 to 6.

Images