Intelligent management method and system for service calls based on data exchange platform

Abstract

The present invention discloses a service call intelligent management method and system based on a data exchange platform, which relates to the technical field of service call, and comprises the following steps: obtaining a plurality of connection pool sizes to be screened; applying the plurality of connection pool sizes to be screened to intelligent management of service calls, and respectively obtaining corresponding scheduling data and concurrent exception data under high concurrent load; constructing a connection pool size evaluation model based on machine learning according to the scheduling data and concurrent exception data; constructing a scheduling evaluation curve with the connection pool size evaluation model and the corresponding connection pool size; performing data analysis on the scheduling evaluation curve to obtain the screened connection pool size, and applying it to intelligent management of service calls. The present application is used to solve the problem that the connection pool size setting is unbalanced, resulting in waste of system resources or increased request delay, thereby affecting the overall performance and stability of service calls.

Description

Technical Field

[0001] The present invention relates to the technical field of service invocation, and more specifically, to an intelligent management method and system for service invocation based on a data exchange platform. Background Art

[0002] The data exchange platform is a cross-system data sharing and integration technology that can transmit and integrate data between different systems, departments or organizations. Through standardized interfaces and protocol conversion, this platform ensures the security, real-time and compatibility of data transmission, and realizes seamless connection of data between multiple systems. Common technologies include API gateways, message queues, data format conversion and security authentication management, which provide an efficient and unified data exchange foundation for the system.

[0003] Service call refers to remotely accessing other services or modules through interfaces in a distributed system to complete specific business functions. Service calls can be implemented through technologies such as RESTful API and RPC, and are widely used in microservice architectures, enabling systems to perform flexible modular collaboration through standardized interfaces to improve the scalability and reusability of the system and the response efficiency of complex business processes.

[0004] Intelligent management uses AI and machine learning technologies in the service call process to automatically optimize the call path, resource allocation and fault handling. Through intelligent scheduling algorithms, dynamic routing and automated operation and maintenance technologies, the system can adjust the call strategy in real time to adapt to dynamic load requirements, ensure service efficiency and maximize resource utilization, effectively reduce human intervention and improve system stability.

[0005] For example, the invention patent with announcement number: CN117909322A announces a unified data management application system and method based on an intelligent computing platform, including a data source, a data lake, and a business area; the unified data management application system uses the data lake to provide services for data applications and data visualization; the data lake includes: a data lake public area, a data lake private area, a Daas platform area, and a data sharing and exchange platform area; wherein, the data lake public area is used to receive public data uploaded by the data source, and the data lake private area is used to receive private data uploaded by the data source; the Daas platform area provides data governance capabilities for the data lake, and the data sharing and exchange platform area provides data API sharing and exchange capabilities for the data lake; wherein, the business area uses the data in the data lake to empower data applications through API calls provided by the data sharing and exchange platform area; the business area includes: a data middle platform, the data stored in it is deposited in the data lake, and the data lake's capabilities are used for external sharing, exchange and API services.

[0006] For example, the method for constructing a unified information support platform for intelligent power distribution and consumption based on a full model disclosed in the invention patent announcement with the announcement number CN104239059A includes: step 1: communicating the control unit with the intelligent power distribution information interaction bus and the intelligent power consumption information interaction bus respectively; step 2: setting the intelligent power distribution information interaction bus in the safety zone I and the safety zone II, and exchanging information with the dispatching automation system and the distribution automation system through the bus adapter; setting the intelligent power consumption information interaction bus in the safety zone III and the safety zone IV, and exchanging information with the power consumption information acquisition unit and the distributed energy management unit through the model mapping method. Compared with the prior art, the method for constructing a unified information support platform for intelligent power distribution and consumption provided by the present invention enables the information support platform to have the functions of unified acquisition and centralized monitoring of full model information for substations, distribution lines, distribution substations, users, distributed energy and energy storage systems, and provide safe, stable, reliable and fast data exchange services for various departments.

[0007] The above disclosed technical solutions have at least the following technical problems: In the service call scenario, the connection pool is a technology for efficiently managing network connections, and plays a key role in distributed systems that need to frequently access external services. The connection pool reuses existing connections to avoid the resource overhead of creating and destroying connections for each request, thereby optimizing system performance.

[0008] When the connection pool is too large, the system will maintain a large number of idle connections, which will occupy system resources, such as memory and file handles. Especially in high concurrency situations, this resource occupation may cause resource competition and lead to insufficient overall system resources. At the same time, too many connections increase the management overhead of the connection pool, such as regular health checks, maintenance, and release of inactive connections, which will consume more computing resources and reduce system efficiency.

[0009] When the connection pool is too small, requests need to queue up to wait for connection resources in high concurrency situations, which increases the request latency and affects the service response speed. A connection pool that is too small may cause a decrease in throughput because the time it takes for requests to wait for connections increases and the system cannot fully utilize resources to handle concurrent requests, which in turn affects system performance and user experience.

[0010] The existing technology sets the size of the connection pool based on historical experience. Unbalanced connection pool size setting will lead to waste of system resources or increased request latency, thereby affecting the overall performance and stability of service calls.

[0011] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0012] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an intelligent service call management method and system based on a data exchange platform. By analyzing the size of the connection pool, it solves the problem of unbalanced connection pool size setting, which leads to waste of system resources or increased request delay, thereby affecting the overall performance and stability of the service call.

[0013] To achieve the above object, the present invention provides the following technical solutions:

[0014] The invention discloses a service call intelligent management method and system based on a data exchange platform, comprising the following steps: obtaining a plurality of different request frequencies under a high concurrent load of historical service call intelligent management, performing data analysis on the request frequencies, calculating an initial connection pool size, and obtaining a plurality of connection pool sizes to be screened by using an increasing and decreasing method to obtain the initial connection pool size; applying the plurality of connection pool sizes to be screened to the service call intelligent management, respectively obtaining corresponding scheduling data and concurrent exception data under a high concurrent load; constructing a connection pool size evaluation model based on machine learning according to the scheduling data and the concurrent exception data; constructing a scheduling evaluation curve with the connection pool size evaluation model and the corresponding connection pool size; performing data analysis on the scheduling evaluation curve to obtain the screened connection pool size, and applying it to the service call intelligent management.

[0015] In a preferred embodiment, several different request frequencies are obtained, and data analysis is performed on the request frequencies to calculate the initial connection pool size, specifically: the request frequencies under high concurrency loads at different times in historical data are obtained, and the corresponding average response time and number of requests per unit time are obtained respectively; the initial connection pool size is calculated based on the request frequency using the calculation formula of the initial connection pool size.

[0016] In a preferred embodiment, specifically: setting an adjustment range of the connection pool size, the adjustment range includes a minimum value and a maximum value; defining an adjustment step, and using the adjustment step as the adjustment value for each adjustment; gradually increasing the connection pool size starting from the initial connection pool size, through a loop, gradually increasing the current connection pool size by the step size until the set maximum value is reached, and recording the connection pool size after each adjustment as the connection pool size to be screened; when the maximum value is reached, starting to decrease, gradually reducing the connection pool size from the maximum value until the set minimum value is reached, and recording the connection pool size after each adjustment as the connection pool size to be screened.

[0017] In a preferred embodiment, the scheduling data includes a scheduling resource utilization impact coefficient and a scheduling efficiency impact coefficient; the specific method for obtaining the scheduling resource utilization impact coefficient is as follows: when performing intelligent management of service calls, obtaining scheduling request data in the operation path; performing denoising and calibration preprocessing on the obtained scheduling request data; establishing a scheduling resource utilization equation based on the processed scheduling request data, and constructing a utilization set; observing and updating the utilization equation through an observation equation to capture changes in scheduling resource utilization under different operation paths; and calculating the scheduling resource utilization impact coefficient of the operation path based on the updated utilization set.

[0018] In a preferred embodiment, the concurrent exception data includes a concurrent conflict exception coefficient; the specific method for obtaining the scheduling efficiency impact coefficient is as follows: obtain several scheduling efficiencies when the service calls intelligent management; the efficiency standard deviation and efficiency average of the scheduling efficiency at the same time interval when the service calls intelligent management; calculate the scheduling efficiency variation coefficient based on the efficiency standard deviation and the efficiency average; calculate the scheduling efficiency impact coefficient based on the scheduling efficiency variation coefficient based on the preset scheduling efficiency impact coefficient calculation formula.

[0019] In a preferred embodiment, the specific method for obtaining the concurrency conflict anomaly coefficient is as follows: based on the concurrency conflict anomaly type, a concurrency conflict anomaly set is obtained, wherein the concurrency conflict anomaly set includes the number of each type of concurrency conflict anomaly and its concurrency conflict anomaly proportion; a preset anomaly score is assigned to each concurrency conflict anomaly type in the concurrency conflict anomaly set; a regression model is constructed based on the concurrency conflict anomaly set and the corresponding anomaly score to generate a concurrency conflict anomaly coefficient.

[0020] In a preferred embodiment, a connection pool size assessment coefficient under different connection pool sizes is obtained from a connection pool size assessment model; a plurality of connection pool sizes and connection pool size assessment coefficients are connected to construct a scheduling assessment curve; data analysis is performed on the scheduling assessment curve to obtain the screened connection pool size, specifically: peak values ​​of a plurality of scheduling assessment curves are obtained, and the change rate of each peak value is calculated respectively, and the quadratic derivative is calculated as the change rate; and the connection pool size corresponding to the peak value of the scheduling assessment curve with the smallest quadratic derivative is used as the screened connection pool size, and it is applied to intelligent management of service calls.

[0021] The technical effects and advantages of the service call intelligent management method and system based on the data exchange platform of the present invention are as follows:

[0022] 1. The present invention realizes intelligent optimization of connection pool size through analysis of historical service call data. The method first obtains historical data under different request frequencies, calculates a reasonable initial connection pool size, and selects the optimal connection pool size through the increasing and decreasing method. Its main technical effects include improving the performance and stability of the system under high concurrent load, reducing request delays, and improving user experience. In addition, by monitoring the scheduling resource utilization influence coefficient and the scheduling efficiency influence coefficient, the method can accurately evaluate resource usage, optimize resource allocation, reduce maintenance costs, and enhance system adaptability and decision support capabilities. In short, the method provides a scientific and effective solution for service call management, ensuring the efficient operation of the system under high load conditions.

[0023] 2. The connection pool size evaluation model constructed by machine learning in the present invention effectively improves the intelligent level of scheduling management. The model comprehensively considers the scheduling resource utilization impact coefficient, the scheduling efficiency impact coefficient and the concurrency conflict exception coefficient, thereby generating a connection pool size evaluation coefficient and realizing a scientific evaluation of the connection pool. The main advantage of this method is that it can dynamically adjust the connection pool size to ensure the performance and stability of the system under high concurrency conditions. By constructing a scheduling evaluation curve and performing data analysis, the optimal configuration of the connection pool can be accurately identified, which significantly reduces the system request delay and improves the user experience. In addition, the analysis method based on the change rate provides a quantitative basis for the optimization of the connection pool, reduces the need for manual intervention, and enhances the system's adaptability to load changes and the level of automated management. In summary, this technology realizes the intelligent management of service calls, improves resource utilization efficiency, and reduces maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a structural diagram of the intelligent management method for service calls based on a data exchange platform of the present invention. DETAILED DESCRIPTION

[0025] 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.

[0026] Embodiment 1, Figure 1 The present invention provides a service call intelligent management method based on a data exchange platform, comprising the following steps:

[0027] S1, under the high concurrent load of historical service call intelligent management, obtains several different request frequencies, performs data analysis on the request frequencies, calculates the initial connection pool size, and obtains several connection pool sizes to be screened by increasing and decreasing the initial connection pool size.

[0028] The method obtains several different request frequencies, performs data analysis on the request frequencies, and calculates the initial connection pool size, specifically:

[0029] Obtain the request frequency under high concurrent load at different times in historical data, and obtain the corresponding average response time and number of requests per unit time respectively;

[0030] The initial connection pool size is calculated based on the request frequency using the calculation formula for the initial connection pool size.

[0031] The calculation formula for the initial connection pool size is as follows:

[0032]

[0033] In the formula, is the initial connection pool size, is the request frequency, is the average response time, The number of requests per unit time.

[0034] The initial connection pool size is obtained by increasing and decreasing the size of the connection pool to be screened, specifically:

[0035] Set the connection pool size adjustment range, including the minimum and maximum values;

[0036] Define the adjustment step size and use it as the adjustment value for each adjustment;

[0037] Starting from the initial connection pool size, the connection pool size is gradually increased. Through a loop, the current connection pool size is gradually increased by the step size until the set maximum value is reached, and the connection pool size after each adjustment is recorded as the connection pool size to be screened;

[0038] When the maximum value is reached, it starts to decrease, gradually reducing the connection pool size from the maximum value until the set minimum value is reached, and the connection pool size after each adjustment is recorded as the connection pool size to be screened.

[0039] It should be noted that under the high concurrency load of historical service call intelligent management, it is crucial to obtain several different request frequencies and perform data analysis on them, because the request frequency directly affects the performance and stability of the system. By analyzing different request frequencies, peak and trough periods can be identified to understand the performance of the system under different load conditions. This analysis process can help engineers calculate a reasonable initial connection pool size to ensure that the system can effectively handle requests in high-concurrency scenarios, reduce waiting time and resource conflicts, and thus improve the overall service quality and user experience. Through an in-depth understanding of request frequency, data support can be provided for subsequent connection pool configuration and adjustment, optimizing the utilization of system resources and preventing excessive resource waste or performance bottlenecks.

[0040] It should be noted that intelligent management of service calls under high concurrency load is crucial, because the system is under great pressure at this time, which can affect its performance and stability. In high-concurrency scenarios, multiple requests come in almost at the same time, and the system must respond quickly to meet user needs. If the connection pool is not configured properly, it will lead to request processing delays, system crashes, or resource waste. Therefore, by obtaining and analyzing request frequencies under high concurrency loads, the behavior of the system can be simulated and predicted more accurately, thereby optimizing the connection pool size and other resource configurations, ensuring efficient operation of the system under high load conditions, and improving user experience and satisfaction.

[0041] S2, applies several connection pool sizes to be screened to the intelligent management of service calls, and obtains the corresponding scheduling data and concurrent exception data under high concurrent load.

[0042] The scheduling data includes a scheduling resource utilization influence coefficient and a scheduling efficiency influence coefficient, and the concurrency exception data includes a concurrency conflict exception coefficient.

[0043] The scheduling resource utilization impact coefficient is an important indicator for evaluating system resource utilization. It is mainly used to measure the actual usage of resources such as CPU, memory, and network in service calls. By monitoring the usage data of these resources, the resource occupancy rate can be calculated and compared with the preset threshold to identify bottlenecks or redundancies in resource usage. For example, real-time monitoring tools can be used to collect the usage of each resource, and then the resource utilization rate can be obtained by calculating the ratio of actual usage to the maximum available amount. Under different connection pool sizes, this coefficient can reflect the efficiency of resource usage, help analyze the impact of connection pool configuration on system performance, and then adjust the connection pool size to ensure the best allocation of resources and avoid system performance degradation and response delays caused by improper resource utilization. By combining historical data and real-time monitoring results, this coefficient can also provide data support for optimization decisions, which helps to improve the overall performance and stability of service calls.

[0044] Analyzing the scheduling resource utilization impact coefficient has the following advantages for evaluating the scheduling effect under different connection pool sizes, selecting the optimal connection pool size, and solving the problem of unbalanced connection pool size settings, which leads to waste of system resources or increased request latency, thus affecting the overall performance and stability of service calls:

[0045] Accurately evaluate resource usage: By analyzing the impact coefficient of scheduling resource utilization, it can accurately reflect the actual usage of various system resources (such as CPU, memory, and network), help identify bottlenecks and waste in resource utilization, and thus provide targeted optimization solutions.

[0046] Optimize connection pool configuration: Based on resource utilization data under different connection pool sizes, the size of the connection pool can be scientifically adjusted to achieve the optimal state of resource allocation, thereby improving the system's response speed and processing capacity and avoiding performance degradation caused by improper configuration.

[0047] Improve system stability: Effective resource utilization and connection pool adjustment can reduce the risk of request delays and system crashes caused by excessive competition for resources, thereby enhancing the stability of the overall system and ensuring service continuity.

[0048] Reduce maintenance costs: By optimizing connection pool configuration and resource utilization, you can reduce unnecessary resource consumption and costs, improve the economy of the system, and reduce the complexity of operation and maintenance management.

[0049] Dynamically adapt to high-concurrency environments: Under high-concurrency load conditions, the real-time data support provided by the scheduling resource utilization impact coefficient can respond to changing request patterns in a timely manner, helping the system to dynamically adjust the connection pool size to ensure optimal performance under different loads.

[0050] Enhanced decision support capabilities: Combining historical data with real-time monitoring, the scheduling resource utilization impact coefficient provides system administrators with a more comprehensive perspective to support efficient decision-making processes, thereby promoting continuous improvement and optimization.

[0051] Through these advantages, the scheduling resource utilization impact coefficient plays a key role in the evaluation and optimization of the connection pool size, ensuring the efficiency and stability of service calls.

[0052] The specific method for obtaining the scheduling resource utilization influence coefficient is as follows:

[0053] When performing intelligent management of service calls, obtain the scheduling request data in the operation path;

[0054] Preprocess the acquired scheduling request data by denoising and calibration;

[0055] Based on the processed scheduling request data, a scheduling resource utilization equation is established and a utilization set is constructed;

[0056] The utilization equation is observed and updated through the observation equation to capture the changes in scheduling resource utilization under different operation paths;

[0057] According to the updated utilization set, the scheduling resource utilization impact coefficient of the operation path is calculated.

[0058] The utilization rate equation is specifically:

[0059]

[0060] The observation equation is specifically:

[0061]

[0062] The specific calculation formula of the scheduling resource utilization influence coefficient of the operation path is:

[0063]

[0064] In the formula, is the utilization set within the utilization set, k is the number of the utilization set, is the noise parameter, To control the parameters, is the updated utilization set, To measure noise, is the scheduling resource utilization impact coefficient of the operation path, and n is the total amount of scheduling request data.

[0065] The scheduling efficiency impact coefficient is a key indicator used to measure the efficiency of the intelligent scheduling system in service resource allocation and request processing, focusing on the processing speed and accuracy of requests. The coefficient can be calculated by monitoring the response time and success rate of the system under different connection pool sizes. During the analysis process, historical performance data and real-time monitoring information can be used to calculate multiple indicators such as the average response time, processing success rate, and resource utilization efficiency of each request to obtain the scheduling efficiency impact coefficient. In addition, a weighted algorithm can be introduced to assign different weights to different types of requests to more accurately reflect the overall performance of the scheduling system. By analyzing this coefficient, the scheduling effect under different connection pool configurations can be identified, and problems caused by unbalanced connection pool size settings, such as resource waste and request delays, can be discovered in a timely manner, thereby optimizing the connection pool size and improving the performance and stability of service calls.

[0066] Analyzing the scheduling efficiency impact coefficient has the following advantages for evaluating the scheduling effect under different connection pool sizes, selecting the optimal connection pool size, and solving the problem of unbalanced connection pool size settings, which leads to waste of system resources or increased request latency, thus affecting the overall performance and stability of service calls:

[0067] Accurately reflect scheduling performance: The scheduling efficiency impact coefficient can quantify the performance of the system in resource scheduling. By evaluating the request processing speed and success rate, we can clearly understand the scheduling efficiency under different connection pool configurations and help identify performance bottlenecks.

[0068] Optimize resource allocation: By analyzing the scheduling efficiency impact coefficient, the system can dynamically adjust the connection pool size according to actual load requirements, thereby optimizing resource allocation and ensuring that the system can still run efficiently under high load conditions.

[0069] Reduce request latency: Improving scheduling efficiency is directly related to the response speed of requests. The optimized connection pool size can reduce latency in request processing, improve user experience, and ensure timely response to customer needs.

[0070] Improve system stability: Reasonable connection pool configuration helps reduce the risk of service interruption due to resource competition and enhances system stability, especially in the face of high concurrent requests, ensuring service continuity.

[0071] Support real-time decision-making: The scheduling efficiency impact coefficient provides real-time feedback to help system administrators quickly identify unbalanced connection pool configurations and make adjustments to ensure that the system always maintains optimal performance under different load conditions.

[0072] Enhanced adaptability: This factor helps the system to flexibly adjust the size of the connection pool when facing changing request patterns, improve its adaptability to high-concurrency scenarios, and thus ensure the reliability and efficiency of the service.

[0073] Promote continuous improvement: By regularly analyzing the impact coefficient of scheduling efficiency, data support can be provided for the long-term optimization of the system, prompting the continuous improvement and development of service call management.

[0074] The above advantages make the scheduling efficiency impact coefficient play a key role in the evaluation and optimization of the connection pool size, effectively improving the performance of service calls and the stability of the system.

[0075] The specific method for obtaining the scheduling efficiency influence coefficient is as follows:

[0076] Obtain the scheduling efficiency of several services when intelligently managing service calls;

[0077] The efficiency standard deviation and efficiency average of the scheduling efficiency at the same time interval when the service call is managed intelligently;

[0078] Calculate the coefficient of variation of scheduling efficiency based on the efficiency standard deviation and the efficiency average;

[0079] The scheduling efficiency influence coefficient is calculated based on the preset scheduling efficiency influence coefficient calculation formula.

[0080] The specific calculation formula of the scheduling efficiency variation coefficient is as follows:

[0081]

[0082] The specific calculation formula of the scheduling efficiency influence coefficient is as follows:

[0083]

[0084] In the formula, is the coefficient of variation of scheduling efficiency, is the scheduling efficiency at the jth moment, is the total number of moments collected when the service calls intelligent management, and j is the moment number; is the scheduling efficiency impact coefficient.

[0085] The concurrency conflict anomaly coefficient is used to quantify the lock contention of the system in high-concurrency scenarios, which is specifically manifested as a comprehensive evaluation of key indicators such as lock waiting time, lock contention number, deadlock events, and concurrent transaction success rate. By monitoring the lock waiting time, the degree of delay in request acquisition can be analyzed. A long lock waiting time means that the system response is affected, thereby reducing the user experience. The statistics of the number of lock contention events reflect the frequency of lock contention events occurring in concurrent service calls. High-frequency lock contention usually indicates that there is a performance bottleneck in the system and the lock management strategy needs to be optimized. Monitoring the occurrence and frequency of deadlock events helps to identify and handle deadlock situations in a timely manner, avoid stagnation of service calls, and ensure the healthy operation of the system. In addition, analyzing the success rate of concurrent transactions can reveal the processing efficiency of concurrent requests. A low success rate may point to potential concurrency conflict problems, requiring in-depth analysis of lock usage. The comprehensive use of these indicators, through data analysis methods such as time series analysis and trend prediction, can provide a scientific basis for system tuning and optimization, thereby effectively improving the performance and stability of the system under high concurrent loads.

[0086] Analyzing the concurrency conflict anomaly coefficient has the following advantages for evaluating the scheduling effect under different connection pool sizes, selecting the optimal connection pool size, and solving the problem of unbalanced connection pool size settings, which leads to waste of system resources or increased request delays, thus affecting the overall performance and stability of service calls:

[0087] Quantifying the impact of lock contention: The concurrency conflict anomaly coefficient provides system administrators with a clear data basis by quantifying lock wait time, lock contention times, and deadlock events, making it easier to identify the specific impact of lock contention on performance under different connection pool configurations.

[0088] Optimize connection pool configuration: By analyzing the concurrency conflict anomaly coefficient, we can find the relationship between the connection pool size and lock contention, thereby optimizing the connection pool configuration to reduce lock contention and improve the effective use of resources.

[0089] Improve response speed: Reducing concurrency conflicts will directly reduce lock waiting time, thereby improving the system's response speed, ensuring that user requests are processed in a timely manner, and enhancing user experience.

[0090] Reduce system risks: Regularly monitoring the concurrency conflict anomaly coefficient helps to detect potential deadlock problems early, so that measures can be taken to prevent system stagnation and ensure service continuity and stability.

[0091] Support dynamic adjustment: Combined with the analysis of concurrent conflict anomaly coefficients, the system can dynamically adjust the size of the connection pool to adapt to changing request patterns and concurrent loads, thereby improving the flexibility of the system.

[0092] Promote data-driven decision-making: Through in-depth analysis of the concurrency conflict anomaly coefficient, it can provide data support for managers to help them make scientific connection pool size adjustment decisions based on actual conditions to avoid resource waste or request delays.

[0093] Improve system predictability: By understanding the patterns and trends of concurrent conflicts, the system can better predict its performance under high concurrent loads, develop more reasonable plans, and improve the overall scheduling effect.

[0094] Enhanced overall performance: Ultimately, the optimized connection pool settings can effectively reduce concurrency conflicts, improve the overall performance and stability of the system, and enable service calls to operate efficiently under different load conditions.

[0095] In summary, analyzing the concurrency conflict anomaly coefficient provides important support for the evaluation and optimization of the connection pool size, ensuring that the system achieves optimal performance and stability during service calls.

[0096] The specific method for obtaining the concurrent conflict anomaly coefficient is as follows:

[0097] Based on the concurrency conflict exception type, a concurrency conflict exception set is obtained, where the concurrency conflict exception set includes the number of each type of concurrency conflict exception and its proportion;

[0098] Assigning a preset anomaly score to each concurrency conflict anomaly type in the concurrency conflict anomaly set;

[0099] A regression model is constructed based on the concurrency conflict anomaly set and the corresponding anomaly scores to generate the concurrency conflict anomaly coefficient.

[0100] The specific calculation formula of the concurrent conflict anomaly coefficient is as follows:

[0101]

[0102] In the formula, is the concurrent conflict anomaly coefficient, … is the preset anomaly score corresponding to the concurrent conflict anomaly, … is the proportion of concurrent conflict exceptions in all concurrent conflict exceptions, is the number of types of concurrency conflict exceptions, n is the number of concurrency conflict exceptions, and e is a natural number.

[0103] This embodiment realizes intelligent optimization of the connection pool size by analyzing historical service call data. The method first obtains historical data under different request frequencies, calculates a reasonable initial connection pool size, and selects the optimal connection pool size by the increasing and decreasing method. Its main technical effects include improving the performance and stability of the system under high concurrent loads, reducing request delays, and improving user experience. In addition, by monitoring the scheduling resource utilization influence coefficient and the scheduling efficiency influence coefficient, the method can accurately evaluate resource usage, optimize resource allocation, reduce maintenance costs, and enhance system adaptability and decision support capabilities. In short, the method provides a scientific and effective solution for service call management, ensuring the efficient operation of the system under high load conditions.

[0104] Embodiment 2, S3, builds a connection pool size assessment model based on machine learning according to scheduling data and concurrent exception data.

[0105] The connection pool size evaluation model is constructed based on machine learning according to the scheduling data and concurrent exception data, specifically:

[0106] The obtained scheduling resource utilization impact coefficient, scheduling efficiency impact coefficient and concurrent conflict anomaly coefficient are used to establish a traffic shaping risk assessment model to generate a connection pool size assessment coefficient;

[0107] The specific calculation formula of the connection pool size evaluation coefficient is as follows:

[0108]

[0109] In the formula, is the connection pool size assessment factor, is the preset proportional coefficient of the scheduling resource utilization impact coefficient, is the preset proportional coefficient of the dispatching efficiency impact coefficient, is the preset proportional coefficient of the concurrent conflict anomaly coefficient, is the scheduling resource utilization impact coefficient, is the preset proportional coefficient of the dispatching efficiency impact coefficient, It is the preset proportional coefficient of the concurrent conflict exception coefficient.

[0110] S4, constructing a scheduling evaluation curve using the connection pool size evaluation model and the corresponding connection pool size.

[0111] The connection pool size evaluation model and the corresponding connection pool size are used to construct a scheduling evaluation curve, specifically:

[0112] Obtain connection pool size assessment coefficients under different connection pool sizes from the connection pool size assessment model;

[0113] A scheduling evaluation curve is constructed by connecting several connection pool sizes and connection pool size evaluation coefficients.

[0114] S5, performing data analysis on the scheduling evaluation curve to obtain the screened connection pool size, and applying it to the intelligent management of service calls.

[0115] The scheduling evaluation curve is analyzed to obtain the connection pool size after screening, which is specifically:

[0116] Obtain the peak values ​​of several scheduling evaluation curves, calculate the change rate of each peak value respectively, and calculate the second derivative as the change rate;

[0117] The connection pool size corresponding to the peak value of the scheduling evaluation curve of the minimum quadratic derivative is used as the screened connection pool size, and is applied to the intelligent management of service calls.

[0118] The connection pool size evaluation model constructed by machine learning in this embodiment effectively improves the intelligent level of scheduling management. The model comprehensively considers the scheduling resource utilization impact coefficient, the scheduling efficiency impact coefficient and the concurrency conflict exception coefficient, thereby generating a connection pool size evaluation coefficient and realizing a scientific evaluation of the connection pool. The main advantage of this method is that it can dynamically adjust the connection pool size to ensure the performance and stability of the system under high concurrency conditions. By constructing a scheduling evaluation curve and performing data analysis, the optimal configuration of the connection pool can be accurately identified, which significantly reduces system request delays and improves user experience. In addition, the analysis method based on the rate of change provides a quantitative basis for the optimization of the connection pool, reduces the need for manual intervention, and enhances the system's adaptability to load changes and the level of automated management. In summary, this technology realizes the intelligent management of service calls, improves resource utilization efficiency, and reduces maintenance costs.

[0119] The above formulas are all dimensionless and numerical calculations. The formula is a formula obtained by obtaining a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0120] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0121] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0122] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0123] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0124] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A service call intelligent management method based on a data exchange platform, characterized in that: The steps include: Under the high concurrent load of historical service call intelligent management, several different request frequencies are obtained, and data analysis is performed on the request frequencies to calculate the initial connection pool size. The initial connection pool size is incremented and decremented to obtain several connection pool sizes to be screened. Apply several connection pool sizes to be screened to intelligent management of service calls, and obtain the corresponding scheduling data and concurrent exception data under high concurrent loads; A connection pool size assessment model is built based on machine learning based on scheduling data and concurrent exception data. Specifically: The scheduling data includes a scheduling resource utilization influence coefficient and a scheduling efficiency influence coefficient; the concurrent exception data includes a concurrent conflict exception coefficient; The scheduling resource utilization impact coefficient is used to evaluate the system resource utilization; The scheduling efficiency impact coefficient is used to measure the efficiency of the intelligent scheduling system in service resource allocation and request processing; The concurrency conflict anomaly coefficient is used to quantify the lock contention of the system in high concurrency scenarios; The obtained scheduling resource utilization impact coefficient, scheduling efficiency impact coefficient and concurrent conflict anomaly coefficient are used to establish a traffic shaping risk assessment model to generate a connection pool size assessment coefficient; The specific calculation formula of the connection pool size evaluation coefficient is as follows: In the formula, is the connection pool size assessment factor, is the preset proportional coefficient of the scheduling resource utilization impact coefficient, is the preset proportional coefficient of the dispatching efficiency impact coefficient, is the preset proportional coefficient of the concurrent conflict anomaly coefficient, is the scheduling resource utilization impact coefficient, is the scheduling efficiency influence coefficient, is the concurrent conflict anomaly coefficient; Obtain connection pool size assessment coefficients under different connection pool sizes from the connection pool size assessment model; Connect several connection pool sizes and connection pool size evaluation coefficients to construct a scheduling evaluation curve; Obtain the peak values ​​of several scheduling evaluation curves, calculate the change rate of each peak value respectively, and calculate the second derivative as the change rate; The connection pool size corresponding to the peak value of the scheduling evaluation curve of the minimum quadratic derivative is used as the screened connection pool size, and is applied to the intelligent management of service calls.

2. The service call intelligent management method based on the data exchange platform according to claim 1 is characterized in that: The method obtains several different request frequencies, performs data analysis on the request frequencies, and calculates the initial connection pool size, specifically: Obtain the request frequency under high concurrent load at different times in historical data, and obtain the corresponding average response time and number of requests per unit time respectively; The initial connection pool size is calculated based on the request frequency using the calculation formula for the initial connection pool size.

3. The service call intelligent management method based on the data exchange platform according to claim 2 is characterized in that: The initial connection pool size is obtained by increasing and decreasing the size of the connection pool to be screened, specifically: Set the connection pool size adjustment range, including the minimum and maximum values; Define the adjustment step size and use it as the adjustment value for each adjustment; Starting from the initial connection pool size, the connection pool size is gradually increased. Through a loop, the current connection pool size is gradually increased by the step size until the set maximum value is reached, and the connection pool size after each adjustment is recorded as the connection pool size to be screened; When the maximum value is reached, it starts to decrease, gradually reducing the connection pool size from the maximum value until the set minimum value is reached, and the connection pool size after each adjustment is recorded as the connection pool size to be screened.

4. The service call intelligent management method based on the data exchange platform according to claim 3 is characterized in that: The specific method for obtaining the scheduling resource utilization influence coefficient is as follows: When performing intelligent management of service calls, obtain the scheduling request data in the operation path; Preprocess the acquired scheduling request data by denoising and calibration; Based on the processed scheduling request data, a scheduling resource utilization equation is established and a utilization set is constructed; The utilization equation is observed and updated through the observation equation to capture the changes in scheduling resource utilization under different operation paths; According to the updated utilization set, the scheduling resource utilization impact coefficient of the operation path is calculated.

5. The service call intelligent management method based on the data exchange platform according to claim 4 is characterized in that: The specific method for obtaining the scheduling efficiency influence coefficient is as follows: Obtain several scheduling efficiencies during intelligent management of service calls; Calculate the standard deviation and average efficiency of the scheduling efficiency at the same time interval when the service calls intelligent management; Calculate the coefficient of variation of scheduling efficiency based on the efficiency standard deviation and the efficiency average; The scheduling efficiency influence coefficient is calculated based on the preset scheduling efficiency influence coefficient calculation formula.

6. The intelligent management method for service calls based on a data exchange platform according to claim 5 is characterized in that: The specific method for obtaining the concurrent conflict anomaly coefficient is as follows: Based on the concurrency conflict exception type, a concurrency conflict exception set is obtained, where the concurrency conflict exception set includes the number of each type of concurrency conflict exception and its proportion; Assigning a preset anomaly score to each concurrency conflict anomaly type in the concurrency conflict anomaly set; A regression model is constructed based on the concurrency conflict anomaly set and the corresponding anomaly score to generate the concurrency conflict anomaly coefficient.

7. The intelligent management method for service calls based on a data exchange platform according to claim 6 is characterized in that: The calculation formula for the initial connection pool size is as follows: In the formula, is the initial connection pool size, is the request frequency, is the average response time, The number of requests per unit time.

8. The service call intelligent management method based on the data exchange platform according to claim 7 is characterized in that: The utilization rate equation is specifically: The observation equation is specifically: The specific calculation formula of the scheduling resource utilization influence coefficient of the operation path is: In the formula, is the utilization set within the utilization set, k is the number of the utilization set, is the noise parameter, To control the parameters, is the updated utilization set, To measure noise, is the scheduling resource utilization impact coefficient of the operation path, and n is the total amount of scheduling request data.

9. A system using the service call intelligent management method based on a data exchange platform as claimed in any one of claims 1 to 8, characterized in that It includes a connection pool size acquisition module, a data analysis module, a model building module, a curve building module and a screening module; The connection pool size acquisition module is used to obtain several different request frequencies under the high concurrent load of historical service call intelligent management, and perform data analysis on the request frequencies to calculate the initial connection pool size, and obtain several connection pool sizes to be screened by increasing and decreasing the initial connection pool size; The data analysis module is used to apply several connection pool sizes to be screened to the intelligent management of service calls, and obtain the corresponding scheduling data and concurrent exception data under high concurrent load; The model building module applies several connection pool sizes to be screened to the intelligent management of service calls, and obtains the corresponding scheduling data and concurrent exception data under high concurrent load respectively; the scheduling data includes the scheduling resource utilization influence coefficient and the scheduling efficiency influence coefficient; the concurrent exception data includes the concurrent conflict exception coefficient; the obtained scheduling resource utilization influence coefficient, scheduling efficiency influence coefficient and concurrent conflict exception coefficient are used to establish a traffic shaping hidden danger assessment model to generate a connection pool size assessment coefficient; A curve building module is used to obtain connection pool size evaluation coefficients under different connection pool sizes from a connection pool size evaluation model; connect several connection pool sizes and connection pool size evaluation coefficients to build a scheduling evaluation curve; The screening module is used to obtain the peak values ​​of several scheduling evaluation curves, calculate the change rate of each peak value respectively, and calculate the quadratic derivative as the change rate; and use the connection pool size corresponding to the peak value of the scheduling evaluation curve with the smallest quadratic derivative as the screened connection pool size, and apply it to the intelligent management of service calls.

Images