An artificial intelligence-based automatic server configuration management method

Abstract

The present invention relates to the technical field of server automatic configuration management. The present invention relates to a server automatic configuration management method based on artificial intelligence. It includes the following steps: S1. Collect the number of server resources available for configuration optimization, and at the same time determine the service users of the server; S2. Obtain the historical operation data of the service users, extract the load operation characteristics from the historical operation data, and at the same time obtain the number of server resource requirements corresponding to each operation characteristic; By calculating the weights of the server resources, the present invention preferentially adjusts those server resources with lower risks and smaller memory usage, which ensures that the adjustment decision is based on the actual load and risk situation rather than simply the resource quantity, effectively allocates and uses the server resources, avoids resource idleness or overcrowding, and improves the overall resource utilization rate.

Description

Technical Field

[0001] The present invention relates to the technical field of server automatic configuration management, and more specifically, to a method for automatically configuring and managing a server based on artificial intelligence. Background Art

[0002] The cloud center manages a large number of heterogeneous servers. During the construction and operation and maintenance management of the cloud platform, the environment preparation time is long, the configuration is complex, and the later operation and maintenance configuration models are not unified, which has become a bottleneck restricting the development of the cloud center.

[0003] Currently, the mainstream solutions for managing the network configurations of a large number of servers usually use automated operation and maintenance tools such as saltstack, ansible, and puppet to complete automated adjustments. However, in a dynamic environment, such as a frequently changing or automatically expanding cloud environment, the automated configuration is inconsistent with the actual system state. This disconnection may lead to unexpected behaviors or configuration errors because the automated operation and maintenance tools cannot timely understand the actual state changes of the system. As a result, when adjusting server resources, it is easy to have idle resources or overcrowding, leading to server resource interruptions, increasing the instability of the system load and the possibility of failures. Therefore, a method for automatically configuring and managing a server based on artificial intelligence is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for automatically configuring and managing a server based on artificial intelligence to solve the problems raised in the above background art.

[0005] To achieve the above purpose, a method for automatically configuring and managing a server based on artificial intelligence is provided, including the following steps:

[0006] S1. Collect the number of server resources available for configuration optimization, and at the same time determine the service users of the server.

[0007] S2. Obtain the historical operation data of the service users, extract the load operation characteristics from the historical operation data, and at the same time obtain the number of server resource requirements corresponding to each operation characteristic.

[0008] S3. Establish an artificial intelligence model, use the artificial intelligence model to monitor the real-time operation state of the service users, and at the same time input the operation characteristics and the number of characteristic resource requirements into the artificial intelligence model, and grant the server resource quantity configuration management authority.

[0009] S4. Set buffer configuration regulations in the artificial intelligence model, predict the load based on the historical operation data of the service users, compare the predicted server resources required for the load with the server resources required for the real-time operation characteristics, and then the artificial intelligence model selects the data that occupies more server resources as the management task.

[0010] Set the buffer configuration according to the predicted load to prevent resource shortage, and determine how to allocate resources to cope with load fluctuations;

[0011] S5. Set dynamic adjustment regulations in the artificial intelligence model. When configuring and adjusting server resources, obtain the running status of load data in the server resources, use the running status of load data and the memory as weights, select the server with the highest weight value for adjustable analysis, and screen out the adjustable servers for adjustment.

[0012] As a further improvement of this technical solution, the steps of S1 are as follows:

[0013] S1.1. Collect the number of server resources in the management system and obtain the service users of each server resource;

[0014] S1.2. Extract the configuration change diary of each server resource, screen the nature of the server resources according to the configuration change diary, and classify the server resources into those available for configuration optimization and fixed configurations, so as to obtain the number of server resources available for configuration optimization and the number of server resources with fixed configurations.

[0015] As a further improvement of this technical solution, the steps of S2 are as follows:

[0016] S2.1. Collect the historical operation data of service users, screen the historical operation data according to the load status, and then extract the load operation characteristics by combining each load status with the historical operation data to obtain the operation characteristics representing different load statuses of service users;

[0017] S2.2. Then match the operation characteristics with the number of server resources to obtain the required number of server resources for each operation characteristic.

[0018] As a further improvement of this technical solution, the formula of S2 is as follows:

[0019] ;

[0020] Among them, is the required number of CPU resources, is the number of CPU cores required by the operation characteristics, is the number of CPU cores of each server;

[0021] ;

[0022] Among them, is the required number of memory resources, is the memory capacity required by the operation characteristics, is the memory capacity of each server;

[0023] ;

[0024] Among them, is the required amount of storage resources, is the storage capacity required for the running characteristics, is the storage capacity of each server;

[0025] ;

[0026] Among them, is the required amount of network resources, is the network bandwidth required for the running characteristics, is the network bandwidth of each server.

[0027] As a further improvement of this technical solution, the steps of S3 are as follows:

[0028] S3.1. Access the server management system, establish an artificial intelligence model using artificial intelligence algorithms within the server management system, and then use the artificial intelligence model to monitor the load status of the server and the real-time running status of service users;

[0029] S3.2. Give the artificial intelligence model the server configuration management permission within the server management system, and then input the required quantity of server resources corresponding to each running characteristic obtained in S2.2 into the artificial intelligence model. Monitor each service user through the artificial intelligence model, and when a service user exhibits the corresponding running characteristic, configure the required quantity of server resources for the service user.

[0030] As a further improvement of this technical solution, it is determined that the overall server is in a high-load state by setting that the overall server usage exceeds 80% within the server management system;

[0031] When the server management system does not show an overall high-load state, the control permission priority of S3 is lower than that of S4;

[0032] When the server management system shows an overall high-load state, the control permission priority of S3 is higher than that of S4;

[0033] When S3 and S4 perform server resource configuration, they will both trigger dynamic adjustment regulations.

[0034] As a further improvement of this technical solution, the formula of S4 is as follows:

[0035] ;

[0036] Among them, is the predicted load at the current time point, is a constant term representing the average level of the load data, is an autoregressive parameter that controls the impact of historical load values on the current load, is a moving average parameter that controls the impact of past errors on the current load, is the white noise or error term at the current time point, representing the random error between the actual load and the predicted load;

[0037] ;

[0038] Among them, is the difference between the predicted load and the real-time load, is the predicted resource demand, is the real-time resource demand;

[0039] ;

[0040] Among them, B is the buffer size, and are adjustment parameters, is the predicted resource demand, is the load difference;

[0041] ;

[0042] Among them, is the priority score of task i, is the resource occupancy of task i, is the maximum resource occupancy value among all tasks.

[0043] As a further improvement of this technical solution, the steps of S5 are as follows:

[0044] S5.1. Set dynamic adjustment regulations in the artificial intelligence model. The dynamic adjustment regulations are used to first obtain the number of server resources that the service user needs to increase when configuring and adjusting the server resources;

[0045] S5.2. Obtain the server resources corresponding to each service user, and obtain the running status of the load data in the server resources. Then, use the running status of the load data and the memory usage as weights. The higher the adjustment risk coefficient indicated by the running status, the lower the weight; the smaller the proportion of the memory usage, the higher the weight proportion. Then, select the server with the highest weight value for adjustable analysis. When the server resources can still bear the running data generated by the service user after the adjustment of the server resources, it is determined that the server resources are adjustable servers. On the contrary, when the server resources cannot bear the running data generated by the service user after the adjustment of the server resources, it is determined that the server resources are non-adjustable servers.

[0046] As a further improvement of this technical solution, the formula of S5 is as follows:

[0047] ;

[0048] Where, L is the current load level, R is the adjustment risk coefficient of the running status, M is the proportion of the memory usage of the current server, and W is the weight;

[0049] C = A - L;

[0050] Where, A is the calculated load capacity after adjustment, and C is the current server load capacity. If , the server is adjustable; otherwise, it is non-adjustable.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] 1. In this server automatic configuration management method based on artificial intelligence, by calculating the weights of server resources, the server resources with lower risks and smaller memory usages are preferentially adjusted, which ensures that the adjustment decision is based on the actual load and risk situation, rather than simply the number of resources, effectively allocates and uses server resources, avoids resource idleness or overcrowding, improves the overall resource utilization rate. At the same time, the weight calculation takes into account the adjustment risk coefficient of the running status, reduces the risks that may occur during the adjustment process, and reduces the instability of the system load and the possibility of failures by preferentially considering the servers with lower risks for adjustment.

[0053] 2. In the server automatic configuration management method based on artificial intelligence, by dynamically adjusting resources, it is ensured that the server can carry the needs of service users, avoiding the situation where some servers are overloaded while others are idle, thus balancing the system load, improving the server's ability to process requests, improving the response time and service quality, enhancing the user experience. At the same time, the automated resource adjustment reduces manual intervention, simplifies the operation and maintenance work process, improves the operation and maintenance efficiency. Based on the analysis of load data and operating status, the resource adjustment decision is made more scientific and reasonable, reducing human errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is the overall flowchart of the present invention;

[0055] Figure 2 It is the flowchart of collecting the number of server resources in the collection management system of the present invention;

[0056] Figure 3 It is the flowchart of obtaining the required number of server resources corresponding to each operation characteristic of the present invention;

[0057] Figure 4 It is the flowchart of allocating server resources with the required number of configuration requirements for service users of the present invention;

[0058] Figure 5 It is the flowchart of obtaining the number of server resources that service users need to increase in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0060] Please refer to Figures 1 - 5 As shown, the purpose of this embodiment is to provide a server automatic configuration management method based on artificial intelligence, including the following steps:

[0061] S1. Collect the number of server resources available for configuration optimization, and at the same time determine the service users of the server;

[0062] The steps of S1 are as follows:

[0063] S1.1. Collect the number of server resources in the management system and obtain the service users of each server resource;

[0064] Use the system monitoring tool Prometheus to collect the quantity and usage of server resources;

[0065] Analyze server logs (such as application logs and system logs) to identify resource usage and service users.

[0066] S1.2. Extract the configuration change diary of each server resource, filter the nature of the server resources according to the configuration change diary, and classify the server resources into those available for configuration optimization and fixed configurations, so as to obtain the quantity of server resources available for configuration optimization and the quantity of server resources with fixed configurations. The specific working steps are as follows:

[0067] Collect the configuration change diary: Extract the configuration change diary from the server management system or log system, ensuring that the diary contains information such as timestamp, change content, and change type;

[0068] Parse the configuration change diary: Use text parsing technology to extract the configuration information of the resources from the diary, usually including parameters such as CPU, memory, storage, and network;

[0069] Determine the nature of resource configuration: Analyze the configuration change history of each server to determine the nature of the resources, by using change frequency analysis (if a certain configuration changes frequently, it indicates that the configuration is still being optimized) and change content analysis (check whether there is an obvious optimization trend or fixed pattern);

[0070] Classify server resources: Classify server resources into "available for configuration optimization (resource configurations change frequently, or the configuration changes reflect a certain optimization trend)" and "fixed configurations (configuration changes are rare, or the change records show that the configuration is stable)".

[0071] Statistics and reporting: Calculate the quantity of server resources in each category and generate a report to more accurately classify and count server resources. The formula is as follows:

[0072] ;

[0073] Among them, the number of changes is the number of changes of a certain configuration within the observation period. The observation period can be a time period such as one month or one year, and calculate the configuration change frequency of each server;

[0074] Stability = standard deviation (configuration parameters);

[0075] To calculate the stability of the configuration, statistical measures such as the standard deviation can be used to measure the degree of fluctuation of the configuration parameters. The smaller the standard deviation, the more stable the configuration;

[0076] S2. Obtain the historical operation data of the service user, extract the load operation characteristics from the historical operation data, and at the same time obtain the quantity of server resource requirements corresponding to each operation characteristic;

[0077] The steps of S2 are as follows:

[0078] S2.1. Collect the historical operation data of the service user, screen the historical operation data according to the load status, and then extract the load operation characteristics by combining each load status with the historical operation data to obtain the operation characteristics representing different load statuses of the service user. The specific working steps are as follows:

[0079] Collect historical data: Collect the historical operation data of the user service, which may include CPU usage, memory usage, network traffic, I / O operations, etc.;

[0080] Data types: Structured data (such as database logs), unstructured data (such as log files), time series data, etc.;

[0081] Define the load status: Classify according to different load statuses of the system (such as low load, medium load, high load);

[0082] Load status marking: Mark each piece of historical data with the corresponding load status;

[0083] Feature selection: Extract important features from the data of each load status;

[0084] Feature construction: Generate a feature vector representing the load status.

[0085] S2.2. Then, perform demand matching by combining the operation characteristics with the quantity of server resources to obtain the quantity of server resource requirements corresponding to the requirements of each operation characteristic. The specific working steps are as follows:

[0086] Determine the operation characteristics: It is necessary to define and identify the operation characteristics of the system or application, which include computing requirements, memory requirements, storage requirements, network requirements, GPU requirements, I / O operations, etc.;

[0087] Determine the server resources: Understand the resource configuration of each server, which usually includes CPU, memory, storage, network, GPU quantity, type, etc.;

[0088] Establish a resource requirement model: Establish a resource requirement model for each operation characteristic. Usually, the following formula can be used:

[0089] ;

[0090] Where, is the required quantity of CPU resources, is the number of CPU cores required by the running characteristics, is the number of CPU cores per server;

[0091] ;

[0092] Among them, is the required amount of memory resources, is the memory capacity required by the running characteristics, is the memory capacity per server;

[0093] ;

[0094] Among them, is the required amount of storage resources, is the storage capacity required by the running characteristics, is the storage capacity per server;

[0095] ;

[0096] Among them, is the required amount of network resources, is the network bandwidth required by the running characteristics, is the network bandwidth per server.

[0097] Calculation of demand quantity: For each running characteristic, calculate the required server resource quantity according to the above formula. Assuming there are multiple characteristics and resource configurations, these calculations can be summarized into a comprehensive demand table.

[0098] S3. Establish an artificial intelligence model, use the artificial intelligence model to monitor the real-time running status of service users, and at the same time input the running characteristics and the quantity of characteristic resource requirements into the artificial intelligence model, and give the server resource quantity configuration management permission;

[0099] The steps of S3 are as follows:

[0100] S3.1. Connect to the server management system, use artificial intelligence algorithms to establish an artificial intelligence model within the server management system, and then use the artificial intelligence model to monitor the load status of the server and the real-time running status of service users;

[0101] S3.2. Give the artificial intelligence model the server configuration management permission within the server management system, and then input the server resource requirement quantity corresponding to each running characteristic obtained in S2.2 into the artificial intelligence model. Monitor each service user through the artificial intelligence model. When a service user exhibits the corresponding running characteristic, configure the server resources with the required quantity for the service user. The specific working steps are as follows:

[0102] Access Server Management System: Collect the operation data of the server, such as CPU usage rate, memory usage rate, disk I / O, network traffic, etc., and integrate the data into a central database or data warehouse for subsequent processing and analysis;

[0103] Build an artificial intelligence model: Select a suitable AI model, such as a classification model, and extract the features related to the server load. The formula is as follows:

[0104] ;

[0105] Among them, is the number of server resources, is the running feature, is the regression coefficient;

[0106] Real-time monitoring and feedback: Use a stream processing system (such as Apache Kafka, Apache Flink) to process real-time data, input the real-time data into the AI model, and obtain the resource requirements of the current load;

[0107] Automatically configure server resources: Automatically adjust the server configuration according to the model output, such as increasing or decreasing virtual machine instances, adjusting CPU and memory allocation, etc., and feedback the adjustment results to the AI model for further optimization. The formula is as follows:

[0108] ;

[0109] Among them, is the total cost, is the cost of each server resource, is the number of each server resource to determine the optimal configuration of resources.

[0110] By setting that the overall server usage exceeds 80% within the server management system, it is determined that the overall server is in a high-load state;

[0111] When the server management system does not show an overall high-load state, the control permission priority of S3 is lower than that of S4;

[0112] When the server management system shows an overall high-load state, the control permission priority of S3 is higher than that of S4;

[0113] When S3 and S4 are configuring server resources, they will both trigger dynamic adjustment regulations.

[0114] S4. Set buffer configuration regulations in the artificial intelligence model, predict the load based on the historical operation data of service users, compare the predicted load with the server resources required by the real-time operation characteristics, and then the artificial intelligence model selects the data that occupies more server resources as the management task. The specific working steps are as follows:

[0115] Load prediction: Use a time series prediction model to predict the future load. The formula is as follows:

[0116] ;

[0117] Among them, is the predicted load at the current time point, is a constant term, representing the average level of load data, is the autoregressive parameter, controlling the influence of historical load values on the current load, is the moving average parameter, controlling the influence of past errors on the current load, is the white noise or error term at the current time point, representing the random error between the actual load and the predicted load;

[0118] Resource demand comparison: Collect real-time server resource usage data, compare the load predicted by the AI model with the real-time load. The formula is as follows:

[0119] ;

[0120] Among them, is the difference between the predicted load and the real-time load, is the predicted resource demand, is the real-time resource demand;

[0121] Buffer configuration regulations: Set buffer configuration according to the predicted load to prevent resource shortage, and determine how to allocate resources to cope with load fluctuations. The formula is as follows:

[0122] ;

[0123] Among them, B is the buffer size, and are adjustment parameters, is the predicted resource demand, is the load difference;

[0124] Management task selection: Evaluate the resource occupancy of each service user, and select the task with the most resource occupancy for priority management adjustment. The formula is as follows:

[0125] ;

[0126] Among them, is the priority score of task i, is the resource occupancy of task i, is the maximum resource occupancy value among all tasks.

[0127] S5. Set dynamic adjustment regulations in the artificial intelligence model. When configuring and adjusting server resources, obtain the running status of load data in the server resources, use the running status of the load data and the memory as weights, select the server with the highest weight value for adjustable analysis, and screen out the adjustable servers for adjustment.

[0128] The steps of S5 are as follows:

[0129] S5.1. Set dynamic adjustment regulations in the artificial intelligence model. The dynamic adjustment regulations are used to first obtain the number of server resources that the service user needs to increase when configuring and adjusting server resources;

[0130] S5.2. Obtain the server resources corresponding to each service user, and obtain the running status of the load data in the server resources. Then, use the running status of the load data and the memory as weights. The higher the adjustment risk coefficient shown by the running status, the lower the weight; the smaller the memory occupancy ratio, the higher the weight ratio. Then, select the server with the highest weight value for adjustable analysis. When the server resources can still bear the running data generated by the service user after the adjustment of the server resources, it is determined that the server resources are adjustable servers. On the contrary, when the server resources cannot bear the running data generated by the service user after the adjustment of the server resources, it is determined that the server resources are non-adjustable servers. The specific working steps are as follows:

[0131] Obtain user requirements and the existing resource status: Determine the number of server resources that need to be increased, and obtain the server resources corresponding to each service user, including its load data and running status;

[0132] Calculate resource weights: Calculate the weights according to the load data and running status of the server resources. The higher the adjustment risk coefficient shown by the running status, the lower the weight; the smaller the memory occupancy ratio, the higher the weight;

[0133] Select and analyze: Select the server with the highest weight value for adjustable analysis, and judge whether the server resources after adjustment can bear the running data generated by the user;

[0134] Decision: If the server resources after adjustment can bear the user's requirements, it is determined to be adjustable; otherwise, it is determined to be non-adjustable. The formula is as follows:

[0135] ;

[0136] Wherein, L is the current load level, R is the adjustment risk coefficient of the running state (the value range is from 0 to 1, and the higher the value, the higher the risk), M is the proportion of the running memory occupied by the current server (the value range is from 0 to 1, and the higher the value, the greater the running memory occupancy), and W is the weight;

[0137] The influence of the load data L on the weight calculation can be adjusted by comprehensively considering the relationship between the load data L and the running state risk coefficient R and the running memory occupancy M. The higher the running state risk coefficient R, the lower the weight, and the greater the running memory occupancy M, the lower the weight will also be;

[0138] After calculating the weights for all servers, select the server with the highest weight value for analysis. Assume that the adjusted server resources can increase ΔR (ΔR is the newly added resource amount), and calculate the adjusted load capacity A. The current server load capacity C can be expressed as:

[0139] C = A - L;

[0140] If , then the server is adjustable; otherwise, it is non-adjustable.

[0141] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. An automatic configuration management method for a server based on artificial intelligence, characterized in that: It includes the following steps: S1. Collect the quantity of server resources available for configuration optimization and determine the service users of the server simultaneously; S2. Obtain the historical operation data of the service users, extract the load operation characteristics from the historical operation data, and obtain the quantity of server resources required for each operation characteristic simultaneously; The steps of S2 are as follows: S2.

1. Collect the historical operation data of the service users, screen the historical operation data according to the load status, then extract the load operation characteristics by combining each load status with the historical operation data to obtain the operation characteristics representing different load statuses of the service users; S2.

2. Then perform demand matching by combining the operation characteristics with the quantity of server resources to obtain the quantity of server resources required for each operation characteristic; S3. Establish an artificial intelligence model, use the artificial intelligence model to monitor the real-time operation status of the service users, input the operation characteristics and the quantity of characteristic resources required into the artificial intelligence model, and grant the server resource quantity configuration management permission; The steps of S3 are as follows: S3.

1. Connect to the server management system, establish an artificial intelligence model using artificial intelligence algorithms within the server management system, and then use the artificial intelligence model to monitor the load status of the server and the real-time operation status of the service users; S3.

2. Grant the artificial intelligence model the server configuration management permission within the server management system, then input the quantity of server resources required for each operation characteristic obtained in S2.2 into the artificial intelligence model, monitor each service user through the artificial intelligence model, and configure the quantity of server resources required for the service user when the service user exhibits the corresponding operation characteristic; S4. Set buffer configuration regulations in the artificial intelligence model, perform load prediction based on the historical operation data of the service users, compare the server resources required for the predicted load with the server resources required for the real-time operation characteristics, and then the artificial intelligence model selects the data that occupies more server resources as the management task; Set buffer configuration according to the predicted load to prevent resource shortage and determine how to allocate resources to cope with load fluctuations; Determine that the overall server is in a high-load state by setting that the overall server usage in the server management system exceeds 80%; When the server management system does not show an overall high-load state, the control permission priority of S3 is lower than that of S4; When the server management system shows an overall high-load state, the control permission priority of S3 is higher than that of S4; When S3 and S4 perform server resource configuration, they will both trigger dynamic adjustment regulations; S5. Set dynamic adjustment regulations in the artificial intelligence model. When configuring and adjusting server resources, obtain the operation status of the load data in the server resources, use the operation status and memory of the load data as weights, select the server with the highest weight value for adjustable analysis, and screen out the adjustable servers for adjustment; The steps of S5 are as follows: S5.

1. Set dynamic adjustment regulations in the artificial intelligence model. The dynamic adjustment regulations are used to first obtain the quantity of server resources that service users need to increase when configuring and adjusting server resources. S5.

2. Obtain the server resources corresponding to each service user, and obtain the running status of the load data in the server resources. Then, use the running status of the load data and the memory usage as weights. The higher the adjustment risk coefficient indicated by the running status, the lower the weight; the smaller the proportion of memory usage, the higher the weight proportion. Then, select the server with the highest weight value for adjustable analysis. When the server resources can still bear the running data generated by the service user after the adjustment of the server resources, it is determined that the server resources are adjustable servers. Conversely, when the server resources cannot bear the running data generated by the service user after the adjustment of the server resources, it is determined that the server resources are non-adjustable servers.

2. The method for automatically configuring and managing a server based on artificial intelligence according to claim 1, wherein: The steps of S1 are as follows: S1.

1. Collect the quantity of server resources in the management system and obtain the service users of each server resource. S1.

2. Extract the configuration change diaries of each server resource, screen the nature of the server resources according to the configuration change diaries, and classify the server resources into those available for configuration optimization and fixed configurations, so as to obtain the quantity of server resources available for configuration optimization and the quantity of server resources with fixed configurations.

3. The method for automatically configuring and managing a server based on artificial intelligence according to claim 1, wherein: The formula of S2 is as follows: ； wherein, is the required amount of CPU resources, is the number of CPU cores required to run the feature, is the number of CPU cores per server; ； wherein, is the required amount of memory resources, is the memory capacity required for running features, is the memory capacity of each server; ； Among them, is the required amount of storage resources, is the storage capacity required to run the feature, is the storage capacity of each server; ； Among them, is the number of required network resources, is the network bandwidth required to run the feature, is the network bandwidth of each server.

4. The method for automatically configuring and managing a server based on artificial intelligence according to claim 1, wherein: The formula of S4 is as follows: ； Among them, is the predicted load at the current time point, is a constant term representing the average level of the load data, is an autoregressive parameter that controls the influence of historical load values on the current load, is a moving average parameter that controls the influence of past errors on the current load, is the white noise or error term at the current time point, representing the random error between the actual load and the predicted load; ； Among them, is the difference between the predicted load and the real-time load, is the predicted resource requirement, is the real-time resource requirement; ； where B is the buffer size, and are adjustment parameters, is the predicted resource requirement, is the load difference; ； wherein, is the priority score of task i, is the resource occupancy of task i, is the maximum resource occupancy value among all tasks.

5. The method for automatically configuring and managing a server based on artificial intelligence according to claim 1, wherein: The formula of S5 is as follows: ； Among them, R is the adjustment risk coefficient of the running status, M is the memory usage proportion of the current server, and W is the weight. C = A - L; Among them, A is the load capacity after calculation adjustment, L is the current load level, and C is the current server load capacity. If , the server is adjustable; otherwise, it is non-adjustable.

Images