Distributed scheduling performance detection and evaluation method and system based on artificial intelligence

Abstract

The invention relates to the technical field of electronic digital data processing, in particular to a distributed scheduling performance detection and evaluation method and system based on artificial intelligence, and the method comprises the steps: S1, collecting the operation environment data of each distribution node, and obtaining the environment interference characteristic value of each distribution node through processing; s2, collecting system operation data of each distribution node, performing comprehensive analysis to obtain a system operation performance evaluation index of each distribution node, and performing task allocation according to the system operation performance evaluation index of each distribution node; and S3, collecting real-time operation data of the system, carrying out comprehensive analysis to obtain a system stability evaluation index, and carrying out system performance evaluation according to the system stability evaluation index. According to the artificial intelligence-based distributed scheduling performance detection and evaluation method and system provided by the invention, the stability of the system is comprehensively evaluated. System performance reduction or potential faults and node performance bottlenecks can be found in time, and the overall stability and processing capacity of the system can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and in particular to a distributed scheduling performance detection and evaluation method and system based on artificial intelligence. Background Art

[0002] With the popularization of technologies such as cloud computing, big data, and the Internet of Things, distributed systems have gradually become the mainstream computing model. These systems require efficient and reliable scheduling algorithms to manage resources and ensure system stability and performance. Artificial intelligence (AI), as a cutting-edge branch of computer science, has made significant progress in recent years. With the rapid development of technologies such as machine learning, data mining, and big data, AI has gradually been integrated into distributed scheduling systems.

[0003] For example, the invention patent with announcement number: CN111144701B is a method for classifying and evaluating ETL job scheduling resources in a distributed environment, including: step one, determining an indicator system for evaluating ETL server performance; step two, determining the ETL job type; step three, clustering and analyzing ETL server indicator data based on the indicator system to obtain a candidate set of ETL servers classified corresponding to the ETL job type; step four, establishing an indicator evaluation matrix and calculating the information entropy of the indicator and the weight of the information entropy; step five, sorting the candidate sets of ETL servers of each category; step six, determining the ETL job type according to the calculation of step two for the ETL job, and then selecting the ETL server with the highest ranking in the candidate set of ETL servers of the corresponding classification.

[0004] For example, the invention patent with publication number: CN112614013A is a multi-energy coupling scheduling system and method for a multi-station integrated distributed energy station. The system includes: a plan formulation module, which is used to formulate an energy demand plan based on user needs and constraints; set key parameters to impose conditional constraints on the load rate and operating time of the heat pump unit, and the operating power and duration of the air conditioner; an information collection module, which is used to collect measurement information on the operation of energy equipment and environmental sensing equipment; a data service module, which is used to store the information required for the formulation of energy demand plans and calculate the performance coefficient; an operation evaluation module, which is used to collect information based on the information collection module, and combine the historical data and historical scheduling plans in the database to perform comprehensive energy supply scheduling for the energy station; a result output module, which is used to output the scheduling plan and the evaluation results of the plan status.

[0005] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems: the performance of the distributed scheduling system is often affected by multiple factors, which may cause the existing performance detection and evaluation methods to produce different evaluation results at different times or in different scenarios. This increases the complexity and uncertainty of the evaluation results. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides an artificial intelligence-based distributed scheduling performance detection and evaluation method and system, which can effectively solve the problems involved in the above-mentioned background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: The first aspect of the present invention provides a distributed scheduling performance detection and evaluation method based on artificial intelligence, including: S1. Obtain distributed scheduling processing nodes, mark them as distributed nodes, collect operating environment data of each distributed node, and obtain the environmental interference characteristic value of each distributed node after processing.

[0008] S2. Collect the system operation data of each distributed node, and obtain the system operation performance evaluation index of each distributed node through comprehensive analysis based on the environmental interference characteristic value of each distributed node, and assign tasks based on the system operation performance evaluation index of each distributed node.

[0009] S3. Collect the real-time operation data of the system, and obtain the system stability evaluation index based on the system operation performance evaluation index of each distributed node through comprehensive analysis, and conduct system performance evaluation and feedback based on the system stability evaluation index.

[0010] As a further method, the operating environment data of each distributed node includes: real-time temperature, humidity, voltage, current and network delay of each distributed node within a preset detection period.

[0011] The distributed nodes are basic units that constitute a distributed scheduling system and can exchange data and work together through network connections and communication protocols.

[0012] As a further method, the environmental interference characteristic value of each distributed node is obtained through processing. The specific process is: extracting the reference standard temperature, reference standard humidity, allowable deviation temperature, allowable deviation humidity, rated voltage, rated current, allowable deviation voltage, allowable deviation current and critical network delay from the performance detection system database, normalizing the temperature deviation, humidity deviation, voltage deviation, current deviation and network delay, introducing weight factors and integrating them to obtain the environmental interference characteristic value of each distributed node. The environmental interference characteristic value of each distributed node is used to quantitatively evaluate the degree of interference of the environment in which each distributed node of the distributed scheduling system is located on the system performance of each distributed node, and provide a basis for evaluating the system operation performance of each distributed node.

[0013] As a further method, the comprehensive analysis obtains the system operation performance evaluation index of each distributed node. The specific analysis process is: the system operation data of each distributed node includes: the throughput, response time, CPU utilization, memory utilization and network bandwidth of each distributed node within a preset detection period.

[0014] The critical throughput, critical response time, critical CPU utilization, critical memory utilization and critical network bandwidth are extracted from the performance detection system database, and the throughput, response time, CPU utilization, memory utilization and network bandwidth of each distributed node are compared with the critical throughput, critical response time, critical CPU utilization, critical memory utilization and critical network bandwidth. The system operation performance evaluation index of each distributed node is obtained by accumulating the environmental interference characteristic values ​​of each distributed node and introducing a weight factor. The system operation performance evaluation index of each distributed node is used to quantitatively evaluate the operation performance of the distributed scheduling system at each distributed node, provide a basis for evaluating the overall operation performance of the distributed scheduling system, and can also help guide task allocation.

[0015] As a further method, the task allocation is performed according to the operation performance evaluation index of each distributed node system. The specific process is: sorting the operation performance evaluation index of each distributed node system in ascending order to obtain a performance evaluation index sequence, extracting the first quartile of the performance evaluation index sequence and marking it as the lower limit value of task scheduling, and extracting the third quartile of the performance evaluation index sequence and marking it as the upper limit value of task scheduling.

[0016] The system operation performance evaluation index of each distributed node is compared with the task scheduling lower limit and the task scheduling upper limit respectively. If the system operation performance evaluation index of a distributed node is greater than the task scheduling lower limit and the system operation performance evaluation index of the distributed node is less than the task scheduling upper limit, the distributed node is marked as a node with no demand scheduling.

[0017] If the system operation performance evaluation index of a distributed node is less than or equal to the task scheduling lower limit, the distributed node is marked as a demand reduction task node.

[0018] If the system operation performance evaluation index of a distributed node is greater than or equal to the upper limit of the task scheduling, the distributed node is marked as a demand-added task node.

[0019] Assign tasks to demand-increasing task nodes and demand-decreasing task nodes.

[0020] As a further method, the task allocation for demand-increased task nodes and demand-decreased task nodes is carried out, and the specific process is: marking the deviation value between the system operation performance evaluation index of the demand-increased task node and the task scheduling upper limit value as the expected demand-increased task parameter, and obtaining the increased task amount of the demand-increased task node from the performance detection system database according to the expected demand-increased task parameter.

[0021] The deviation value between the system operation performance evaluation index of the demand reduction task node and the task scheduling lower limit is marked as the expected demand reduction task parameter, and the reduced task amount of the demand reduction task node is obtained by matching from the performance detection system database according to the expected demand reduction task parameter.

[0022] The increased task amounts of all task nodes with increased demand and the reduced task amounts of all task nodes with reduced demand are counted for task allocation.

[0023] As a further method, the comprehensive analysis obtains system stability evaluation indicators, and the specific analysis process is: real-time operation data of the system, specifically including: task success rate, task execution time, task queue length and recovery time of the system within a preset detection period.

[0024] The critical task success rate, critical task execution time, critical task queue length and critical recovery time are extracted from the performance detection system database, and the ratio of the system's task success rate, task execution time, task queue length and recovery time to the critical task success rate, critical task execution time, critical task queue length and critical recovery time is combined with the system operation performance evaluation index of each distributed node, and a weight factor is set. A comprehensive analysis is performed to obtain a system stability evaluation index, which is used to quantitatively evaluate the stability of a distributed scheduling system and provide a basis for performance evaluation of a distributed scheduling system.

[0025] As a further method, the system performance evaluation is performed according to the system stability evaluation index. The specific evaluation process is: extracting the system stability evaluation index threshold from the performance detection system database, comparing the system stability evaluation index with the system stability evaluation index threshold, if the system stability evaluation index is higher than or equal to the system stability evaluation index threshold, then the system performance is evaluated as qualified, if the system stability evaluation index is lower than the system stability evaluation index threshold, then the system performance is evaluated as unqualified.

[0026] Conduct problem analysis on systems with substandard performance and provide feedback on the results.

[0027] As a further method, the specific numerical expression of the distribution node system operation performance evaluation index is: ; In the formula, represents the nth distributed node system operation performance evaluation index, e represents a natural constant, represents the throughput of the nth distribution node, Indicates the CPU usage of the nth distribution node, Indicates the memory usage of the nth distribution node, represents the network bandwidth of the nth distribution node, represents the response time of the nth distribution node, represents the critical throughput, represents the critical response time, Indicates the critical CPU usage, Indicates critical memory usage, represents the critical network bandwidth, represents the environmental interference characteristic value of the nth distribution node, Indicates the impact factor of the distributed node system operation performance corresponding to the preset throughput, Indicates the impact factor of the distributed node system operation performance corresponding to the preset CPU usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset memory usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset network bandwidth, Indicates the impact factor of the distributed node system operation performance corresponding to the preset response time, It represents the impact factor of the distributed node system operation performance corresponding to the preset environmental interference characteristic value.

[0028] The second aspect of the present invention provides an artificial intelligence-based distributed scheduling performance detection and evaluation system, including: an environmental interference detection module, a system node performance evaluation module, a system overall performance evaluation module and a performance detection system database.

[0029] The environmental interference detection module is used to obtain the distributed scheduling processing nodes, mark them as distributed nodes, collect the operating environment data of each distributed node, and obtain the environmental interference characteristic value of each distributed node after processing.

[0030] The system node performance evaluation module is used to collect the system operation data of each distributed node, and obtain the system operation performance evaluation index of each distributed node based on the environmental interference characteristic value of each distributed node through comprehensive analysis, and perform task allocation based on the system operation performance evaluation index of each distributed node.

[0031] The system overall performance evaluation module is used to collect the real-time operation data of the system, and obtain the system stability evaluation index based on the system operation performance evaluation index of each distributed node through comprehensive analysis, and then evaluate the system performance and provide feedback based on the system stability evaluation index.

[0032] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) The present invention provides a distributed scheduling performance detection and evaluation method and system based on artificial intelligence, which comprehensively analyzes the task success rate, task execution time, task queue length and recovery time, and system node performance, and more comprehensively evaluates the stability of the system. It can timely detect system performance degradation or potential failures, and take corresponding measures to optimize and improve them. At the same time, it can also timely detect node performance bottlenecks and expand or upgrade them, thereby improving the overall stability and processing capacity of the system; (2) The present invention comprehensively analyzes the temperature, humidity, voltage, current and network delay of the node environment to gain a more comprehensive understanding of the overall status of the node environment. Comprehensive analysis can timely discover potential fault hazards, which helps reduce the risk of data loss, damage or leakage, ensure data security, and thus enhance system reliability. At the same time, it can also reasonably allocate system resources according to changes in environmental parameters to help improve resource utilization efficiency; (3) By monitoring and analyzing the changing trends and interrelationships of the throughput, response time, CPU usage, memory usage and network bandwidth of each distributed node, the present invention can timely discover and optimize system performance bottlenecks, thereby improving the overall system performance and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The present invention is further described with the aid of the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other accompanying drawings based on the following accompanying drawings without creative work. Figure 1 It is a schematic diagram of the method flow of the present invention; Figure 2 It is a schematic diagram of system module connection of the present invention. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0035] Reference Figure 1 As shown, the first aspect of the present invention provides a distributed scheduling performance detection and evaluation method based on artificial intelligence, including: S1. Obtain distributed scheduling processing nodes, mark them as distributed nodes, collect operating environment data of each distributed node, and obtain environmental interference characteristic values ​​of each distributed node after processing.

[0036] It should be added that the distributed scheduling processing node in this embodiment refers to the device or node that the distributed scheduling system monitors in real time in different application scenarios. It has different meanings in different application scenarios. For example, in the cloud computing platform resource management application, the distributed scheduling processing node refers to the computing unit in the cloud computing platform, such as a virtual machine, container or physical server. These computing units are connected together through a network to form a large computing resource pool for executing the user's computing tasks. Each node can run applications independently and can dynamically allocate and release computing resources as needed.

[0037] In IoT device monitoring and maintenance applications, distributed scheduling processing nodes refer to various intelligent devices deployed at the edge of the network, such as sensors, actuators, or other hardware devices that can sense environmental changes and interact with the outside world. These devices usually send the collected data back to the central server for further processing, and also receive instructions from the cloud to adjust their own behavior.

[0038] In big data analysis and processing applications, distributed scheduling processing nodes refer to data processing nodes that participate in big data processing frameworks (such as Hadoop, Spark, etc.). These nodes are responsible for storing copies of some data sets and performing specific computing tasks, such as part of a MapReduce job. They are interconnected through high-speed networks and work together to complete complex queries and analysis on large-scale data sets.

[0039] In addition, the distributed dispatching system in this embodiment can also be applied to financial transaction systems and intelligent traffic management systems, etc.

[0040] Specifically, the operating environment data of each distributed node includes: the real-time temperature, humidity, voltage, current and network delay of each distributed node within a preset detection period.

[0041] The distributed nodes are basic units that constitute a distributed scheduling system and can exchange data and work together through network connections and communication protocols.

[0042] It should be explained that the network delay refers to the network delay of the network environment in which the node is located.

[0043] It should be explained that temperature sensors, humidity sensors, voltage sensors, and current sensors can be used to obtain the corresponding environmental data. Data center infrastructure management systems can also be used to obtain environmental data in real time. Data center infrastructure management systems usually integrate environmental monitoring functions. Network monitoring software such as PRTG, Nagios, and Zabbix can be used to obtain network latency data.

[0044] Furthermore, the environmental interference characteristic value of each distributed node is obtained after processing. The specific process is: extract the reference standard temperature, reference standard humidity, allowable deviation temperature, allowable deviation humidity, rated voltage, rated current, allowable deviation voltage, allowable deviation current and critical network delay from the performance detection system database, normalize the temperature deviation, humidity deviation, voltage deviation, current deviation and network delay, introduce weight factors and integrate them to obtain the environmental interference characteristic value of each distributed node. The environmental interference characteristic value of each distributed node is used to quantitatively evaluate the degree of interference of the environment in which each distributed node of the distributed scheduling system is located on the system performance of each distributed node, and provide a basis for evaluating the system operation performance of each distributed node.

[0045] In a specific embodiment, the numerical expression of the environmental interference characteristic value of each distribution node is: ; In the formula, represents the environmental interference characteristic value of the nth distribution node, n represents the number of each distribution node, , h represents the total number of distribution nodes, t represents the time variable, , Indicates the current time point. Indicates the detection start time point, represents the temperature of the nth distribution node at time t, represents the humidity of the nth distribution node at time t, represents the voltage of the nth distribution node at time t, represents the current of the nth distribution node at time t, represents the network delay of the nth distribution node at time t, Indicates the reference standard temperature, Indicates reference standard humidity, Indicates the allowable deviation temperature, Indicates the allowable deviation humidity, Indicates the rated voltage, Indicates the rated current, Indicates the allowable deviation voltage, Indicates the allowable deviation current, represents the critical network delay, Indicates the environmental interference impact factor of the distribution node corresponding to the preset temperature, Indicates the distribution node environmental interference impact factor corresponding to the preset humidity, Indicates the distributed node environmental interference impact factor corresponding to the preset voltage, Indicates the environmental interference impact factor of the distribution node corresponding to the preset current, Indicates the distributed node environment interference impact factor corresponding to the preset network delay.

[0046] It needs to be explained that when the absolute difference between temperature and humidity and the reference standard temperature and humidity is greater, the absolute difference between voltage and current and the rated voltage and current is greater, and the network delay is greater, the corresponding distributed node environmental interference characteristic value is greater, indicating that the abnormal environment around the distributed node interferes with the node operation to a greater extent.

[0047] It should be explained that in this embodiment Indicates the environmental interference impact factor of the distribution node corresponding to the preset temperature, Indicates the distribution node environmental interference impact factor corresponding to the preset humidity, Indicates the distributed node environmental interference impact factor corresponding to the preset voltage, Indicates the environmental interference impact factor of the distribution node corresponding to the preset current, The distributed node environment interference impact factor corresponding to the preset network delay represents the numerical value of the interference degree of the temperature, humidity, voltage, current and network delay unit value on the distributed node environment. When used, the distributed node environment interference impact factor corresponding to the temperature, the distributed node environment interference impact factor corresponding to the humidity, the distributed node environment interference impact factor corresponding to the voltage, the distributed node environment interference impact factor corresponding to the current and the distributed node environment interference impact factor corresponding to the network delay can be directly obtained from the performance detection system database. The corresponding relationship can be a pre-set mapping relationship. For example, the temperature, humidity, voltage, current and network delay at the node are respectively mapped with the distributed node environment interference impact factors corresponding to the temperature, humidity, voltage, current and network delay preset in the performance detection system database to form a mapping set. The real-time temperature, humidity, voltage, current and network delay are input into the mapping set to obtain the distributed node environment interference impact factor corresponding to the temperature, the distributed node environment interference impact factor corresponding to the humidity, the distributed node environment interference impact factor corresponding to the voltage, the distributed node environment interference impact factor corresponding to the current and the distributed node environment interference impact factor corresponding to the network delay. The mapping relationship can be one-to-one or a many-to-one relationship. The above-mentioned impact factors are all extracted from the performance detection system database, and the value range is between 0 and 1.

[0048] It needs to be explained that as the temperature rises, the resistance, capacitance and other parameters of electronic components may change, thus affecting the stability and performance of the circuit. Under extremely high or low temperature conditions, electronic components may even fail or be damaged. Excessive humidity may cause condensation on the surface of electronic components, causing short circuits or corrosion, and low humidity may cause static electricity accumulation, increasing the risk of damage to electronic components. Voltage fluctuations or instability may cause the performance of electronic equipment to deteriorate or fail. The selection and adjustment of node operating voltage will directly affect the performance and stability of electronic equipment. If the voltage is too high, it may cause electronic components to overheat or be damaged; if the voltage is too low, it may cause electronic equipment to not work properly or perform well. Excessive current, that is, much greater than the rated current, may cause electronic components to overheat, burn or be damaged; and unstable current may cause electronic equipment performance fluctuations or instability. In the case of large network delays, the real-time nature of data transmission will be affected, resulting in increased signal synchronization or timing problems.

[0049] It needs to be explained that in a high temperature environment, the resistance of electronic components may change, thereby affecting the voltage distribution of the circuit. In addition, temperature may also affect the efficiency of the power supply, thereby affecting the stability of the output voltage. Changes in temperature may also affect the magnitude and stability of the current. In a high temperature environment, the on-resistance of electronic components may decrease, resulting in an increase in current. At the same time, high temperatures may also accelerate the aging process of electronic components, thereby affecting their current handling capabilities. Excessive humidity may cause condensation or corrosion on the surface of electronic components, thereby affecting the voltage stability of the circuit. Unstable voltage or current may cause performance degradation or increased failure rate of network equipment, thereby increasing network latency.

[0050] It should be explained that by comprehensively analyzing the temperature, humidity, voltage, current and network delay of the node's environment, we can have a more comprehensive understanding of the overall condition of the node's environment. This helps to more accurately evaluate the environmental interference characteristic values. By real-time monitoring and analysis of changes in environmental parameters, potential fault hazards can also be discovered in a timely manner. This helps to take preventive measures in advance to avoid serious impacts on the system caused by failures. It also helps to reduce the risk of data loss, damage or leakage, ensure data security, and thus enhance system reliability. At the same time, according to changes in environmental parameters, system resources can be reasonably allocated, such as adjusting power supply, optimizing network bandwidth, etc. This helps to improve resource utilization efficiency.

[0051] S2. Collect the system operation data of each distributed node, and obtain the system operation performance evaluation index of each distributed node through comprehensive analysis based on the environmental interference characteristic value of each distributed node, and assign tasks based on the system operation performance evaluation index of each distributed node.

[0052] Specifically, a comprehensive analysis is performed to obtain the system operation performance evaluation index of each distributed node. The specific analysis process is as follows: the system operation data of each distributed node includes: the throughput, response time, CPU utilization, memory utilization and network bandwidth of each distributed node within a preset detection period.

[0053] It should be explained that the response time specifically refers to the processing time of a single task on each node. The network bandwidth refers to the amount of data that can be transmitted per unit time at each node, usually in bits per second. The network bandwidth determines the speed and efficiency of data transmission between nodes.

[0054] It should be explained that you can use performance monitoring tools such as Zabbix, Nagios, Prometheus, etc. to monitor the server's throughput, CPU usage, memory usage, response time, network bandwidth and other indicators in real time. You can also use network monitoring tools such as Wireshark, tcpdump, etc. to monitor the server's network traffic and bandwidth usage in real time.

[0055] It should be explained that when the number of tasks processed by the system increases, the response time may increase. Conversely, when the system response time is short, it usually means that the system can process tasks more efficiently, thereby improving throughput. When the CPU utilization rate is too high, the system may not be able to process new tasks in time, resulting in a decrease in throughput. Conversely, when the CPU utilization rate is moderate, the system can process tasks efficiently, thereby improving throughput. When the memory utilization rate is too high, the system may suffer from performance degradation or even crash due to insufficient memory. Conversely, when the memory utilization rate is moderate, the system can use memory resources more efficiently, thereby improving performance. Network bandwidth determines the speed and efficiency of data transmission between nodes. When the network bandwidth is insufficient, the data transmission speed will slow down, resulting in longer system response time and lower throughput. Conversely, when the network bandwidth is sufficient, the data transmission speed will increase, thereby improving system performance and throughput.

[0056] The critical throughput, critical response time, critical CPU utilization, critical memory utilization and critical network bandwidth are extracted from the performance detection system database, and the throughput, response time, CPU utilization, memory utilization and network bandwidth of each distributed node are compared with the critical throughput, critical response time, critical CPU utilization, critical memory utilization and critical network bandwidth. The system operation performance evaluation index of each distributed node is obtained by accumulating the environmental interference characteristic values ​​of each distributed node and introducing a weight factor. The system operation performance evaluation index of each distributed node is used to quantitatively evaluate the operation performance of the distributed scheduling system at each distributed node, provide a basis for evaluating the overall operation performance of the distributed scheduling system, and can also help guide task allocation.

[0057] In a specific embodiment, the numerical expression of the distribution node system operation performance evaluation index is: ; In the formula, represents the nth distributed node system operation performance evaluation index, e represents a natural constant, represents the throughput of the nth distribution node, Indicates the CPU usage of the nth distribution node, Indicates the memory usage of the nth distribution node, represents the network bandwidth of the nth distribution node, represents the response time of the nth distribution node, represents the critical throughput, represents the critical response time, Indicates the critical CPU usage, Indicates critical memory usage, represents the critical network bandwidth, represents the environmental interference characteristic value of the nth distribution node, Indicates the impact factor of the distributed node system operation performance corresponding to the preset throughput, Indicates the impact factor of the distributed node system operation performance corresponding to the preset CPU usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset memory usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset network bandwidth, Indicates the impact factor of the distributed node system operation performance corresponding to the preset response time, It represents the impact factor of the distributed node system operation performance corresponding to the preset environmental interference characteristic value.

[0058] It should be explained that when the node throughput and network bandwidth are larger, and the node response time, CPU utilization, memory utilization and environmental interference characteristic values ​​are smaller, the corresponding distributed node system operation performance evaluation index is larger, indicating that the distributed scheduling system has better operation performance in each distributed node.

[0059] It should be explained that in this embodiment Indicates the impact factor of the distributed node system operation performance corresponding to the preset throughput, Indicates the impact factor of the distributed node system operation performance corresponding to the preset CPU usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset memory usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset network bandwidth, Indicates the impact factor of the distributed node system operation performance corresponding to the preset response time, The distributed node system operation performance impact factor corresponding to the preset environmental interference characteristic value represents the numerical value of the interference degree of the unit value of the throughput, CPU utilization rate, memory utilization rate, network bandwidth, response time and environmental interference characteristic value on the distributed node system operation performance. When used, the distributed node system operation performance impact factor corresponding to the throughput, the distributed node system operation performance impact factor corresponding to the CPU utilization rate, the distributed node system operation performance impact factor corresponding to the memory utilization rate, the distributed node system operation performance impact factor corresponding to the network bandwidth, the distributed node system operation performance impact factor corresponding to the response time and the distributed node system operation performance impact factor corresponding to the environmental interference characteristic value can be directly obtained from the performance detection system database. The corresponding relationship can be a preset mapping relationship, for example, the throughput, CPU utilization rate, memory utilization rate, network bandwidth, response time and distributed node system operation performance impact factor corresponding to the node. The bandwidth, response time and environmental interference characteristic values ​​are respectively mapped with the distributed node system operation performance impact factors corresponding to the throughput, cpu usage, memory usage, network bandwidth, response time and environmental interference characteristic values ​​preset in the performance detection system database to form a mapping set, and the real-time throughput, cpu usage, memory usage, network bandwidth, response time and environmental interference characteristic values ​​are input into the mapping set to obtain the distributed node system operation performance impact factors corresponding to the throughput, the distributed node system operation performance impact factors corresponding to the cpu usage, the distributed node system operation performance impact factors corresponding to the memory usage, the distributed node system operation performance impact factors corresponding to the network bandwidth, the distributed node system operation performance impact factors corresponding to the response time and the distributed node system operation performance impact factors corresponding to the environmental interference characteristic values, wherein the mapping relationship can be one-to-one or many-to-one. The above impact factors are all extracted from the performance detection system database, and the value range is between 0 and 1.

[0060] It should be explained that in a distributed system, each node needs to process a certain number of tasks, so the throughput directly reflects the processing capacity of the node. When the throughput is high, it means that the node can process tasks efficiently and the overall performance of the system is also good. On the contrary, if the throughput is low, it may mean that the node has a performance bottleneck or insufficient resources and needs further optimization or expansion. In a distributed system, user requests may need to be transmitted and processed between multiple nodes, so the length of the response time directly affects the user experience. If the response time is short, it means that the system can respond to user requests quickly and provide smooth services. On the contrary, if the response time is long, it may cause the user to wait too long, affecting the user experience and the overall performance of the system. In a distributed system, if the CPU utilization rate is high, it means that the CPU resources on the node are fully utilized and the system's processing capacity is also strong. However, if the CPU utilization rate is too high, it may also cause the node to be overloaded or performance degradation. In a distributed system, memory resources are used to store and process the data and programs required for the task. If the memory utilization rate is high, it means that the memory resources on the node are fully utilized and the system's processing capacity is also strong. However, if the memory usage is too high, it may also cause memory leaks and memory overflows in the node, thus affecting the stability and performance of the system. In a distributed system, data transmission and communication between nodes are required frequently. If the network bandwidth is low, it may lead to slow data transmission speed, thus affecting the overall performance of the system. On the contrary, if the network bandwidth is high, it can support faster data transmission speed and improve the processing capacity and response speed of the system.

[0061] It needs to be explained that by monitoring and analyzing the changing trends and interrelationships of the throughput, response time, CPU utilization, memory utilization and network bandwidth of each distributed node, it is possible to timely discover and optimize system performance bottlenecks and improve the overall system performance and user experience.

[0062] Furthermore, tasks are allocated according to the system operation performance evaluation index of each distributed node. The specific process is: sort the system operation performance evaluation index of each distributed node in ascending order to obtain a performance evaluation index sequence, extract the first quartile of the performance evaluation index sequence and mark it as the lower limit value of task scheduling, and extract the third quartile of the performance evaluation index sequence and mark it as the upper limit value of task scheduling.

[0063] The system operation performance evaluation index of each distributed node is compared with the task scheduling lower limit and the task scheduling upper limit respectively. If the system operation performance evaluation index of a distributed node is greater than the task scheduling lower limit and the system operation performance evaluation index of the distributed node is less than the task scheduling upper limit, the distributed node is marked as a node with no demand scheduling.

[0064] If the system operation performance evaluation index of a distributed node is less than or equal to the task scheduling lower limit, the distributed node is marked as a demand reduction task node.

[0065] If the system operation performance evaluation index of a distributed node is greater than or equal to the upper limit of the task scheduling, the distributed node is marked as a demand-added task node.

[0066] Assign tasks to demand-increasing task nodes and demand-decreasing task nodes.

[0067] Furthermore, tasks are assigned to demand-increased task nodes and demand-decreased task nodes. The specific process is: the deviation value between the system operation performance evaluation index of the demand-increased task node and the task scheduling upper limit value is marked as the expected demand-increased task parameter, and the increased task volume of the demand-increased task node is obtained by matching the expected demand-increased task parameter from the performance detection system database.

[0068] The deviation value between the system operation performance evaluation index of the demand reduction task node and the task scheduling lower limit is marked as the expected demand reduction task parameter, and the reduced task amount of the demand reduction task node is obtained by matching from the performance detection system database according to the expected demand reduction task parameter.

[0069] The increased task amounts of all task nodes with increased demand and the reduced task amounts of all task nodes with reduced demand are counted for task allocation.

[0070] In a specific embodiment, the task allocation process is as follows: 1. Identify all demand-increasing task nodes and demand-reducing task nodes. 2. Determine the demand for task increase based on the difference between the performance evaluation index of the demand-increasing task node and the upper limit of task scheduling; then select tasks with corresponding demand from the task pool and assign them to the demand-increasing task node. 3. Determine the demand for task reduction based on the difference between the performance evaluation index of the demand-reducing task node and the lower limit of task scheduling; select tasks that can be migrated (i.e., the demand for task reduction) from the demand-reducing task node and reallocate them to the demand-increasing task node or put them into the task pool.

[0071] S3. Collect the real-time operation data of the system, and obtain the system stability evaluation index based on the system operation performance evaluation index of each distributed node through comprehensive analysis, and conduct system performance evaluation and feedback based on the system stability evaluation index.

[0072] Specifically, the real-time operation data of the system includes: the task success rate, task execution time, task queue length and recovery time of the system within a preset detection period.

[0073] It should be explained that the task success rate is a measure of the ratio of the number of tasks successfully executed by the system within a certain period of time to the total number of tasks. A high success rate indicates that the system is stable and can reliably execute tasks. The task execution time refers to the time required for a task to be executed from the beginning to the completion. A stable system should ensure that the task execution time is within a reasonable range to avoid being too long and causing users to wait too long. The task queue length refers to the number of tasks currently waiting to be scheduled. A stable system should ensure that the task queue length is not too long to avoid task accumulation and cause system instability. The recovery time refers to the time required to resume normal operation after a system failure. A stable system should have a shorter recovery time to ensure business continuity.

[0074] It needs to be explained that, with other conditions unchanged, an increase in task execution time may lead to a decrease in task success rate. This is because an increase in execution time may increase the risk of task failure, such as resource exhaustion, system errors, etc. An increase in the length of the task queue usually means that the system needs to process more tasks, which may lead to a decrease in the success rate of tasks, because system resources may become tight and the processing speed may decrease. A decrease in recovery time usually means that the system can recover from failures faster, which may increase the success rate of tasks. An increase in the length of the task queue may lead to an increase in task execution time, because the system needs to process more tasks and the waiting time for each task may increase.

[0075] Furthermore, a comprehensive analysis is performed to obtain a system stability evaluation index, and the specific analysis process is as follows: the real-time operation data of the system, specifically including: the task success rate, task execution time, task queue length and recovery time of the system within a preset detection period.

[0076] The critical task success rate, critical task execution time, critical task queue length and critical recovery time are extracted from the performance detection system database, and the ratio of the system's task success rate, task execution time, task queue length and recovery time to the critical task success rate, critical task execution time, critical task queue length and critical recovery time is combined with the system operation performance evaluation index of each distributed node, and a weight factor is set. A comprehensive analysis is performed to obtain a system stability evaluation index, which is used to quantitatively evaluate the stability of a distributed scheduling system and provide a basis for performance evaluation of a distributed scheduling system.

[0077] In a specific embodiment, the numerical expression of the system stability evaluation index is: ; In the formula, represents the system stability evaluation index, represents the nth distributed node system operation performance evaluation index, represents the task success rate of the system, Indicates the task execution time of the system, Indicates the length of the system's task queue. Indicates the system recovery time. represents the critical task success rate, represents the critical task execution time, Indicates the critical task queue length, represents the critical recovery time, Indicates the system stability impact factor corresponding to the preset distributed node system operation performance evaluation index, Indicates the system stability impact factor corresponding to the preset task success rate, Indicates the system stability impact factor corresponding to the preset task execution time, Indicates the system stability impact factor corresponding to the preset task queue length, Indicates the system stability impact factor corresponding to the preset recovery time.

[0078] It needs to be explained that when the system's task success rate and the distributed node system operation performance evaluation index are greater, and the system's task execution time, task queue length and recovery time are smaller, the corresponding system stability evaluation index will be greater, indicating that the distributed scheduling system has a higher degree of stability.

[0079] It should be explained that in this embodiment Indicates the system stability impact factor corresponding to the preset distributed node system operation performance evaluation index, Indicates the system stability impact factor corresponding to the preset task success rate, Indicates the system stability impact factor corresponding to the preset task execution time, Indicates the system stability impact factor corresponding to the preset task queue length, The system stability influencing factors corresponding to the preset recovery time respectively represent the numerical values ​​of the interference degree of the distributed node system operation performance evaluation index, task success rate, task execution time, task queue length and recovery time unit value on the system stability. When used, the system stability influencing factors corresponding to the distributed node system operation performance evaluation index, the system stability influencing factors corresponding to the task success rate, the system stability influencing factors corresponding to the task execution time, the system stability influencing factors corresponding to the task queue length and the system stability influencing factors corresponding to the recovery time can be directly obtained from the performance detection system database. The corresponding relationship can be a preset mapping relationship, for example, the distributed node system operation performance evaluation index, the device's task success rate, task execution time, The task queue length and recovery time are respectively mapped with the system stability influencing factors corresponding to the distributed node system operation performance evaluation index, task success rate, task execution time, task queue length and recovery time preset in the performance detection system database, and the real-time distributed node system operation performance evaluation index, task success rate, task execution time, task queue length and recovery time are input into the mapping set to obtain the system stability influencing factors corresponding to the distributed node system operation performance evaluation index, the system stability influencing factors corresponding to the task success rate, the system stability influencing factors corresponding to the task execution time, the system stability influencing factors corresponding to the task queue length and the system stability influencing factors corresponding to the recovery time, wherein the mapping relationship can be one-to-one or many-to-one. The above influencing factors are all extracted from the performance detection system database, and the value range is between 0 and 1.

[0080] It should be explained that by monitoring the changing trends of parameters such as task success rate, task execution time, task queue length and recovery time, system performance degradation or potential failures can be discovered in a timely manner, and corresponding measures can be taken to optimize and improve. At the same time, by paying attention to the changes in system node performance, node performance bottlenecks can be discovered in a timely manner and capacity can be expanded or upgraded to improve the overall stability and processing capacity of the system. Comprehensive analysis of task success rate, task execution time, task queue length and recovery time and system node performance can more comprehensively evaluate the stability of the system.

[0081] Furthermore, a system performance evaluation is performed based on the system stability evaluation index. The specific evaluation process is: extracting a system stability evaluation index threshold from a performance detection system database, comparing the system stability evaluation index with the system stability evaluation index threshold; if the system stability evaluation index is higher than or equal to the system stability evaluation index threshold, the system performance is evaluated as qualified; if the system stability evaluation index is lower than the system stability evaluation index threshold, the system performance is evaluated as unqualified.

[0082] Conduct problem analysis on systems with substandard performance and provide feedback on the results.

[0083] In a specific embodiment, when the system is evaluated as having unsatisfactory performance, immediate measures should be taken to identify the root cause of the problem and provide a corresponding solution. The specific process is as follows: U1. First check whether there are obvious hardware failures or software error reports, such as sensor failure, communication link interruption, etc. For such problems, they can be quickly resolved by restarting the device or replacing hardware components.

[0084] U2. Analyze the environmental interference characteristic values ​​to see if there are extreme environmental conditions (such as overheating, humidity) that cause performance degradation. If this is the case, you may need to adjust the deployment location or strengthen protective measures.

[0085] U3. Carefully check the consumption of key resources such as CPU usage and memory usage to determine whether there is a resource bottleneck. This may be caused by improper application configuration or because the task load on some nodes is too high.

[0086] U4. Evaluate the performance of network bandwidth and latency, especially for data flows transmitted across regions. High latency or low bandwidth may cause data processing to slow down, thus affecting overall performance.

[0087] U5. Review recent task scheduling records to confirm whether there is an unreasonable work distribution pattern, such as some nodes taking on too many tasks while other nodes are idle.

[0088] The specific process of result feedback is as follows: Q1. Organize all the above findings into a detailed analysis report, including a detailed description of the problem, potential causes and suggested solutions; Q2. Send alert information and analysis reports to the technical team responsible for maintenance in a timely manner through email, instant messaging tools or other means to ensure that they are aware of the problem as soon as possible; Q3.According to the recommendations in the analysis report, promptly perform necessary repair operations, such as optimizing resource allocation, adjusting task scheduling strategies, upgrading hardware facilities, etc.; Q4. After completing the rectification, continue to monitor the performance of the system to ensure that the problem has been effectively resolved and no new anomalies occur; Q5. Record this incident and the experience in handling it, and update the relevant operation and maintenance manuals and technical documents so that they can be used as a reference when encountering similar situations in the future.

[0089] Reference Figure 2 As shown, the second aspect of the present invention provides an artificial intelligence-based distributed scheduling performance detection and evaluation system, including: an environmental interference detection module, a system node performance evaluation module, a system overall performance evaluation module and a performance detection system database.

[0090] The environmental interference detection module is used to obtain the distributed scheduling processing nodes, mark them as distributed nodes, collect the operating environment data of each distributed node, and obtain the environmental interference characteristic value of each distributed node after processing.

[0091] The system node performance evaluation module is used to collect the system operation data of each distributed node, and obtain the system operation performance evaluation index of each distributed node based on the environmental interference characteristic value of each distributed node through comprehensive analysis, and perform task allocation based on the system operation performance evaluation index of each distributed node.

[0092] The system overall performance evaluation module is used to collect the real-time operation data of the system, and obtain the system stability evaluation index based on the system operation performance evaluation index of each distributed node through comprehensive analysis, and then evaluate the system performance and provide feedback based on the system stability evaluation index.

[0093] In a specific embodiment, the performance detection system database is used to store relevant data in the performance evaluation process of the distributed scheduling system, including critical task success rate, critical task execution time, critical task queue length and critical recovery time, system stability impact factor corresponding to the distributed node system operation performance evaluation index, system stability impact factor corresponding to the task success rate, system stability impact factor corresponding to the task execution time, system stability impact factor corresponding to the task queue length and system stability impact factor corresponding to the recovery time, system stability evaluation index threshold and the above data extracted from the performance detection system database. Performance-related data can be directly extracted from the database by using SQL or database-specific query language. System monitoring tools (such as Prometheus, Grafana, Zabbix, etc.) can also be used to collect and display system indicators. Log management systems (such as ELK Stack, Fluentd, Logstash, etc.) can also be used to collect and analyze log data. Specialized performance monitoring tools (such as Percona Monitoring and Management, New Relic, Dynatrace, etc.) can also be used to monitor database and application performance.

[0094] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A distributed scheduling performance detection and evaluation method based on artificial intelligence, characterized in that: include: S1. Obtain distributed scheduling processing nodes, mark them as distributed nodes, collect operating environment data of each distributed node, and obtain environmental interference characteristic values ​​of each distributed node after processing; S2. Collect system operation data of each distributed node, and obtain the system operation performance evaluation index of each distributed node based on the environmental interference characteristic value of each distributed node through comprehensive analysis, and assign tasks based on the system operation performance evaluation index of each distributed node; S3. Collect the real-time operation data of the system, and obtain the system stability evaluation index based on the system operation performance evaluation index of each distributed node through comprehensive analysis, and conduct system performance evaluation and feedback based on the system stability evaluation index.

2. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 1 is characterized in that: The operating environment data of each distributed node includes: real-time temperature, humidity, voltage, current and network delay of each distributed node within a preset detection period; The distributed nodes are basic units that constitute a distributed scheduling system and can exchange data and work together through network connections and communication protocols.

3. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 2 is characterized in that: The environmental interference characteristic value of each distributed node is obtained through processing, and the specific process is as follows: The reference standard temperature, reference standard humidity, allowable deviation temperature, allowable deviation humidity, rated voltage, rated current, allowable deviation voltage, allowable deviation current and critical network delay are extracted from the performance detection system database. The temperature deviation, humidity deviation, voltage deviation, current deviation and network delay are normalized, and weight factors are introduced and integrated to obtain the environmental interference characteristic values ​​of each distributed node. The environmental interference characteristic values ​​of each distributed node are used to quantitatively evaluate the degree of interference of the environment in which each distributed node of the distributed scheduling system is located on the system performance of each distributed node, providing a basis for evaluating the system operation performance of each distributed node.

4. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 3 is characterized in that: The comprehensive analysis obtains the system operation performance evaluation index of each distributed node, and the specific analysis process is as follows: The system operation data of each distributed node includes: the throughput, response time, CPU usage, memory usage and network bandwidth of each distributed node within a preset detection period; The critical throughput, critical response time, critical CPU utilization, critical memory utilization and critical network bandwidth are extracted from the performance detection system database, and the throughput, response time, CPU utilization, memory utilization and network bandwidth of each distributed node are compared with the critical throughput, critical response time, critical CPU utilization, critical memory utilization and critical network bandwidth. The system operation performance evaluation index of each distributed node is obtained by accumulating the environmental interference characteristic values ​​of each distributed node and introducing a weight factor. The system operation performance evaluation index of each distributed node is used to quantitatively evaluate the operation performance of the distributed scheduling system at each distributed node, provide a basis for evaluating the overall operation performance of the distributed scheduling system, and can also help guide task allocation.

5. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 4 is characterized in that: The task allocation is performed according to the system operation performance evaluation index of each distributed node, and the specific process is as follows: The system operation performance evaluation indexes of each distributed node are sorted in ascending order to obtain a performance evaluation index sequence, the first quartile of the performance evaluation index sequence is extracted and marked as the lower limit value of task scheduling, and the third quartile of the performance evaluation index sequence is extracted and marked as the upper limit value of task scheduling; The system operation performance evaluation index of each distributed node is compared with the lower limit value of task scheduling and the upper limit value of task scheduling respectively. If the system operation performance evaluation index of a distributed node is greater than the lower limit value of task scheduling and the system operation performance evaluation index of the distributed node is less than the upper limit value of task scheduling, the distributed node is marked as a node without demand scheduling; If the system operation performance evaluation index of a distributed node is less than or equal to the task scheduling lower limit, the distributed node is marked as a demand reduction task node; If the system operation performance evaluation index of a distributed node is greater than or equal to the upper limit of the task scheduling, the distributed node is marked as a node for adding tasks; Assign tasks to demand-increasing task nodes and demand-decreasing task nodes.

6. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 5 is characterized by: The specific process of allocating tasks to demand-increasing task nodes and demand-decreasing task nodes is as follows: The deviation value between the system operation performance evaluation index of the demand increase task node and the task scheduling upper limit value is marked as the estimated demand increase task parameter, and the increase task amount of the demand increase task node is obtained by matching from the performance detection system database according to the estimated demand increase task parameter; The deviation value between the system operation performance evaluation index of the demand reduction task node and the lower limit of the task scheduling is marked as the estimated demand reduction task parameter, and the reduction task amount of the demand reduction task node is obtained by matching from the performance detection system database according to the estimated demand reduction task parameter; The increased task amounts of all task nodes with increased demand and the reduced task amounts of all task nodes with reduced demand are counted for task allocation.

7. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 6 is characterized in that: The comprehensive analysis obtains the system stability evaluation index, and the specific analysis process is as follows: The real-time operation data of the system specifically includes: the task success rate, task execution time, task queue length and recovery time of the system within a preset detection period; The critical task success rate, critical task execution time, critical task queue length and critical recovery time are extracted from the performance detection system database, and the ratio of the system's task success rate, task execution time, task queue length and recovery time to the critical task success rate, critical task execution time, critical task queue length and critical recovery time is combined with the system operation performance evaluation index of each distributed node, and a weight factor is set. A comprehensive analysis is performed to obtain a system stability evaluation index, which is used to quantitatively evaluate the stability of a distributed scheduling system and provide a basis for performance evaluation of a distributed scheduling system.

8. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 7 is characterized by: The system performance evaluation and feedback are performed according to the system stability evaluation index. The specific evaluation process is as follows: Extracting a system stability evaluation index threshold from a performance detection system database, comparing the system stability evaluation index with the system stability evaluation index threshold, and if the system stability evaluation index is higher than or equal to the system stability evaluation index threshold, evaluating the system performance as qualified; if the system stability evaluation index is lower than the system stability evaluation index threshold, evaluating the system performance as unqualified; Conduct problem analysis on systems with substandard performance and provide feedback on the results.

9. The distributed scheduling performance detection and evaluation method based on artificial intelligence according to claim 4 is characterized by: The specific numerical expression of the system operation performance evaluation index of each distributed node is: ； In the formula, represents the nth distributed node system operation performance evaluation index, e represents a natural constant, represents the throughput of the nth distribution node, Indicates the CPU usage of the nth distribution node, Indicates the memory usage of the nth distribution node, represents the network bandwidth of the nth distribution node, represents the response time of the nth distribution node, represents the critical throughput, represents the critical response time, Indicates the critical CPU usage, Indicates critical memory usage, represents the critical network bandwidth, represents the environmental interference characteristic value of the nth distribution node, Indicates the impact factor of the distributed node system operation performance corresponding to the preset throughput, Indicates the impact factor of the distributed node system operation performance corresponding to the preset CPU usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset memory usage rate. Indicates the impact factor of the distributed node system operation performance corresponding to the preset network bandwidth, Indicates the impact factor of the distributed node system operation performance corresponding to the preset response time, It represents the impact factor of the distributed node system operation performance corresponding to the preset environmental interference characteristic value.

10. A system using the distributed scheduling performance detection and evaluation method based on artificial intelligence as described in any one of claims 1 to 9, characterized in that: include: Environmental interference detection module, system node performance evaluation module, system overall performance evaluation module and performance detection system database; The environmental interference detection module is used to obtain the distributed scheduling processing nodes, mark them as distributed nodes, collect the operating environment data of each distributed node, and obtain the environmental interference characteristic value of each distributed node after processing; The system node performance evaluation module is used to collect the system operation data of each distributed node, and obtain the system operation performance evaluation index of each distributed node based on the environmental interference characteristic value of each distributed node through comprehensive analysis, and perform task allocation based on the system operation performance evaluation index of each distributed node; The system overall performance evaluation module is used to collect the real-time operation data of the system, and obtain the system stability evaluation index based on the system operation performance evaluation index of each distributed node through comprehensive analysis, and then evaluate the system performance and provide feedback based on the system stability evaluation index.

Images