Online service resource scheduling system based on cloud platform

Abstract

The invention discloses an online service resource scheduling system based on a cloud platform, particularly relates to the technical field of cloud platforms, and comprises a resource scheduling acquisition module, a resource information acquisition module, a resource information analysis module, a comprehensive analysis module, a resource proofreading module and a man-machine interaction module. The resource scheduling acquisition module is used for generating resource information meeting application requirements after sending the online service resource information, and the accuracy of the resource information is improved; resource information is collected through the resource information collection module, the accuracy of the resource information is improved, all coefficient values are analyzed through the comprehensive analysis module to obtain reasonable application resource scheduling values, the resource utilization efficiency is improved, resource scheduling of applications is optimized through the resource proofreading module, the cost is greatly reduced, and the application resource scheduling efficiency is improved. And the practicability of online service resource scheduling of the cloud platform of an enterprise is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud platforms. More specifically, the present invention relates to an online service resource scheduling system based on a cloud platform. Background Art

[0002] With the rapid expansion of digital services, the resource demand of cloud platforms has started to grow. Resource supply is tight and the cost is high due to unreasonable resource allocation and low scheduling efficiency. As various business systems become more dependent on cloud services, more stringent standards are put forward for the accuracy of resource scheduling. In particular, parameters such as the real-time nature, balance of resource allocation, and high efficiency of resource use must be further precisely controlled. At the same time, the phenomenon of resource idleness and waste is strictly restricted to ensure the optimization of business operations.

[0003] An existing online service resource scheduling system based on a cloud platform mainly includes: a resource data collection module, a resource analysis and processing module, and an evaluation module. The resource data collection module is deployed at key nodes of the cloud platform with the help of diverse network probes and performance monitoring tools to collect multi-dimensional data; the resource analysis and processing module uses advanced algorithm models and big data mining technologies to integrate and analyze massive resource data; the evaluation module generates and executes a scheduling evaluation based on the analysis results, combined with preset business rules and resource requirements.

[0004] However, in actual application scenarios, this system still has some drawbacks. For example, in the resource data collection module, the coverage of various collected data is not wide enough, only focusing on some common resource indicators; in the resource analysis and processing module, the analysis of the collected resource data is not fine enough. Facing complex and changing business load patterns, existing algorithm models are difficult to accurately obtain the deep associations between data, and there are often omissions of key information and large errors in analysis results due to improper data processing; this system lacks a comprehensive evaluation index for comprehensive analysis to accurately schedule the business resources of enterprises. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an online service resource scheduling system based on a cloud platform, through the technical field of cloud platforms, to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solution: An online service resource scheduling system based on a cloud platform, including: A resource scheduling acquisition module: After receiving the online service resource information sent by the target enterprise, it generates resource information that meets the application requirements and stores it in the database at the same time; Resource Information Collection Module: Collects resource information for application requirements to obtain cloud resource data and transmits it to the Resource Scheduling and Analysis Module simultaneously. Resource Information Analysis Module: Used to analyze the data of the Resource Information Collection Module and transmit the results to the Comprehensive Analysis Module; includes a computing resource analysis unit, a storage resource analysis unit, a network resource analysis unit, and a server performance analysis unit. Comprehensive Analysis Module: Used to establish a comprehensive analysis model, import the data transmitted by the Resource Information Analysis Module into the comprehensive analysis model, calculate the reasonable value of application resource scheduling, and transmit it to the Resource Verification Module; includes a comprehensive analysis unit. Resource Verification Module: Includes an application resource verification unit and a resource adjustment unit. Human-Machine Interaction Module: Imports the information of the Resource Verification Module into the database and sends it to relevant management personnel.

[0007] Preferably, the computing resource status data includes the leakage rate of server motherboard capacitors, the heat generation rate of GPU transistors, and the memory page swap rate; the storage resource status data includes the average disk seek time, the number of flash wear leveling times of solid-state drives, and the storage network bandwidth utilization rate; the network resource status data includes the network link packet loss rate, the network router cache overflow frequency, and the network link round-trip delay jitter value; the server performance data includes the server power conversion efficiency, the number of servers in use, and the number of server failures.

[0008] Preferably, the computing resource analysis unit is used to establish a computing resource analysis model, import the computing resource status data transmitted by the Resource Information Collection Module into the computing resource analysis model, and calculate the computing resource coefficient value. The calculation method of the computing resource analysis model is: , A 1 represents the computing resource coefficient value, wd i represents the leakage rate of server motherboard capacitors of the i-th application, wg i represents the heat generation rate of GPU transistors of the i-th application, wj i-1 represents the memory page swap rate of the (i - 1)-th application, wj i represents the memory page swap rate of the i-th application, and n represents the total number of applications.

[0009] Preferably, the storage resource analysis unit is used to establish a storage resource analysis model, import the storage resource status data transmitted by the Resource Information Collection Module into the storage resource analysis model, and calculate the storage resource coefficient value. The calculation method of the storage resource analysis model is: , A 2Represents the storage resource coefficient value, dg i Represents the average disk seek time of the i-th application, dt i Represents the number of times of solid-state drive flash wear leveling of the i-th application, dc i Represents the storage network bandwidth utilization rate of the i-th sub-application, and n represents the total number of applications.

[0010] Preferably, the network resource analysis unit is used to establish a network resource analysis model, import the network resource status data transmitted by the resource information collection module into the network resource analysis model, and calculate the network resource coefficient value. The calculation method of the network resource analysis model is: , A 3 Represents the network resource coefficient value, fd i Represents the network link packet loss rate of the i-th application, dc i Represents the network router cache overflow frequency of the i-th application, dr i Represents the network link round-trip delay jitter value of the i-th application, e represents the natural constant, and n represents the total number of applications.

[0011] Preferably, the server performance analysis unit is used to establish a server performance analysis model, import the computing resource status data transmitted by the resource information collection module into the server performance analysis model, and calculate the server performance coefficient value. The calculation method of the server performance analysis model is: , A 4 Represents the server performance coefficient value, kw i Represents the server power conversion efficiency of the i-th application, kt i Represents the number of servers used by the i-th application, kz i Represents the number of server failures of the i-th application, and n represents the number of applications.

[0012] Preferably, the comprehensive analysis unit is used to establish a comprehensive analysis model, import the computing resource coefficient value, computing resource coefficient value, storage resource coefficient value, and network resource coefficient value of the resource information analysis module into the comprehensive analysis model, and calculate the reasonable value of application resource scheduling. The specific calculation formula is: , Where P represents the reasonable value of application resource scheduling, A 1 Represents the computing resource coefficient value, A 2 Represents the storage resource coefficient value, A 3 Represents the network resource coefficient value, A 4 Represents the server performance coefficient value, where λ represents other influencing factors of application resource scheduling.

[0013] Preferably, the application resource verification unit is used to verify the reasonable value of application resource scheduling, and the resource range [P c , P l of the preset application resource scheduling; the resource scheduling allocation degree α, and the specific acquisition method: , When P < P c or P > P l , the application resources are abnormal and enter the resource adjustment unit; when Pc ≤ P ≤ Pl, the application resources are in a reasonable allocation state. The resource adjustment unit is used to adjust the abnormal resources required by the application, and at the same time generate an adjustment index denoted as η; compare the optimization index with the resource scheduling allocation degree, and the specific formula is as follows: , ε represents a positive number infinitely close to 0, indicating that the resources required by the application are reasonably scheduled.

[0014] Technical effects and advantages of the present invention: 1. Through the resource scheduling acquisition module of the present invention, after receiving the online service resource information sent by the target enterprise, the resource information that meets the application requirements is generated, which is conducive to the timely scheduling of the application; through the resource information collection module, the computing resource status data, storage resource status data, network resource status data, and server performance data are collected, which helps to improve the comprehensiveness of the data collected by the system; 2. Through the resource information analysis module of the present invention, various coefficient values are comprehensively analyzed, and the reasonable value of application resource scheduling is calculated, which greatly reduces resource waste and improves the resource scheduling utilization efficiency; through the application resource verification unit in the resource verification module for verification, the resource scheduling of the application is optimized, greatly reducing costs, and helping to improve the practicability and rationality of the online service resource scheduling of the enterprise's cloud platform. Brief Description of the Drawings

[0015] Figure 1 is a schematic diagram of the system structure of the present invention.

[0016] Figure 2 is a schematic diagram of the resource information analysis module of the present invention. Detailed Embodiments

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the 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 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.

[0018] As shown in the appendix Figure 1Shown is an online service resource scheduling system based on a cloud platform, including the following modules: a resource scheduling acquisition module, a resource information collection module, a resource information analysis module, a comprehensive analysis module, a resource verification module, a human-computer interaction module, and a database. The resource scheduling acquisition module is connected to the resource information collection module, the resource information collection module is connected to the resource information analysis module, the resource information analysis module is connected to the comprehensive analysis module, and the comprehensive analysis module is connected to the resource verification module.

[0019] Resource scheduling acquisition module: After receiving the online service resource information sent by the target enterprise, it generates resource information that meets the application requirements and stores it in the database at the same time.

[0020] In this embodiment, specifically, the applications are specifically divided into the 1st application, the 2nd application,..., the i-th application,..., the n-th application, and the resource information amounts required by each application are different.

[0021] Resource information collection module: Collects the resource information required by the application to obtain cloud resource data and transmits it to the resource scheduling analysis module at the same time.

[0022] Furthermore, the cloud resource data specifically includes computing resource status data, storage resource status data, network resource status data, and server performance data.

[0023] In this embodiment, specifically, the computing resource status data includes the leakage rate of the server motherboard capacitor, the heat generation rate of the GPU transistor, and the memory page swap rate; the storage resource status data includes the average disk seek time, the number of flash wear leveling times of the solid-state drive, and the storage network bandwidth utilization rate; the network resource status data includes the network link packet loss rate, the network router cache overflow frequency, and the network link round-trip delay jitter value; the server performance data includes the server power conversion efficiency, the number of servers in use, and the number of server failures.

[0024] Detect the leakage rate of server motherboard capacitors through a high-precision micro-current tester, install a high-precision temperature sensor on the GPU chip to detect the heat generation rate of GPU transistors, and obtain the memory page swap rate through the memory management module of the operating system; use a disk performance test tool to detect the average disk seek time, and build in a management chip to obtain the number of flash wear leveling times of the solid-state drive, configure and monitor the storage network bandwidth utilization rate on the switch of the storage network; detect the packet loss rate of the network link through a network tester, obtain the network router cache overflow frequency through the management interface of the router in combination with some automated script tools, and detect the round-trip delay jitter value of the network link through a professional network performance test tool; detect the server power conversion efficiency through a power analyzer, detect the number of servers in use by calling the API interface of the cloud platform, and regularly query these logs by writing scripts to count the number of faulty servers recorded therein.

[0025] Resource information analysis module: used to analyze the data of the resource information collection module and transmit the results to the comprehensive analysis module; includes a computing resource analysis unit, a storage resource analysis unit, a network resource analysis unit, and a server performance analysis unit.

[0026] In this embodiment, specifically, the computing resource analysis unit is used to establish a computing resource analysis model, import the computing resource status data transmitted by the resource information collection module into the computing resource analysis model, and calculate the computing resource coefficient value. The calculation method of the computing resource analysis model is: , A 1 represents the computing resource coefficient value, wd i represents the leakage rate of the server motherboard capacitor of the i-th application, wg i represents the heat generation rate of the GPU transistors of the i-th application, wj i-1 represents the memory page swap rate of the (i - 1)-th application, wj i represents the memory page swap rate of the i-th application, and n represents the total number of applications.

[0027] In this embodiment, specifically, the storage resource analysis unit is used to establish a storage resource analysis model, import the storage resource status data transmitted by the resource information collection module into the storage resource analysis model, and calculate the storage resource coefficient value. The calculation method of the storage resource analysis model is: , A 2 represents the storage resource coefficient value, dg i represents the average disk seek time of the i-th application, dt i represents the number of flash wear leveling times of the solid-state drive of the i-th application, dci represents the storage network bandwidth utilization rate of the i-th sub-application, and n represents the total number of applications.

[0028] In this embodiment, specifically, the network resource analysis unit is used to establish a network resource analysis model, import the network resource status data transmitted by the resource information collection module into the network resource analysis model, and calculate the network resource coefficient value. The calculation method of the network resource analysis model is as follows: , A 3 represents the network resource coefficient value, fd i represents the network link packet loss rate of the i-th application, dc i represents the network router cache overflow frequency of the i-th application, dr i represents the network link round-trip delay jitter value of the i-th application, e represents the natural constant, and n represents the total number of applications.

[0029] In this embodiment, specifically, the server performance analysis unit is used to establish a server performance analysis model, import the computing resource status data transmitted by the resource information collection module into the server performance analysis model, and calculate the server performance coefficient value. The calculation method of the server performance analysis model is as follows: , A 4 represents the server performance coefficient value, kw i represents the server power conversion efficiency of the i-th application, kt i represents the usage quantity of the server of the i-th application, kz i represents the number of failures of the server of the i-th application, and n represents the number of applications.

[0030] Comprehensive analysis module: It is used to establish a comprehensive analysis model, import the data transmitted by the resource information analysis module into the comprehensive analysis model, calculate the reasonable value of application resource scheduling, and transmit it to the resource verification module; it includes a comprehensive analysis unit.

[0031] In this embodiment, specifically, the comprehensive analysis unit is used to establish a comprehensive analysis model, import the computing resource coefficient value, computing resource coefficient value, storage resource coefficient value, and network resource coefficient value of the resource information analysis module into the comprehensive analysis model, and calculate the reasonable value of application resource scheduling. The specific calculation formula is as follows: , where P represents the reasonable value of application resource scheduling, A 1 represents the computing resource coefficient value, A 2 represents the storage resource coefficient value, A 3Represents the network resource coefficient value, A 4 Represents the server performance coefficient value, where λ represents other influencing factors for application resource scheduling.

[0032] Resource verification module: Includes an application resource verification unit and a resource adjustment unit.

[0033] In this embodiment, specifically, the application resource verification unit is used to verify the reasonable value of application resource scheduling, and the preset resource range [P c , P l of application resource scheduling; the resource scheduling allocation degree α, specific acquisition method: , When P < P c or P > P l , the application resources are abnormal and enter the resource adjustment unit; when Pc ≤ P ≤ Pl, the application resources are in a reasonable allocation state. The resource adjustment unit is used to adjust the abnormal resources required by the application, and at the same time generate an adjustment index denoted as η; compare the optimization index with the resource scheduling allocation degree, and the specific formula is as follows: , ε represents a positive number infinitely close to 0, indicating that the resources required by the application are reasonably scheduled.

[0034] Human-computer interaction module: Import the information of the resource verification module into the database and send it to relevant management personnel.

[0035] Through the resource scheduling acquisition module of the present invention, after receiving the online service resource information sent by the target enterprise, it generates resource information that meets the application requirements, which is beneficial to the timely scheduling of the application; through the resource information collection module, it collects computing resource status data, storage resource status data, network resource status data, and server performance data, which helps to improve the comprehensiveness of data collection in this system; through the resource information analysis module, it comprehensively analyzes various coefficient values and calculates the reasonable value of application resource scheduling, greatly reducing resource waste and improving the resource scheduling utilization efficiency; through the application resource verification unit in the resource verification module for verification, it optimizes the resource scheduling of the application, greatly reducing costs, and helps to improve the practicability and rationality of the online service resource scheduling of the enterprise's cloud platform.

[0036] Secondly, in the attached drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments of the present invention are involved. For other structures, reference can be made to the general design. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other; The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. An online service resource scheduling system based on a cloud platform, characterized in that: include: Resource scheduling and acquisition module: after receiving the online service resource information sent by the target enterprise, it generates resource information that meets the application requirements and stores it in the database; Resource information collection module: collects resource information required by the application, obtains cloud resource data, and transmits it to the resource scheduling and analysis module; Resource information analysis module: used to analyze the data of resource information collection module and transmit the results to comprehensive analysis module; It includes a computing resource analysis unit, a storage resource analysis unit, a network resource analysis unit and a server performance analysis unit; Comprehensive analysis module: used to establish a comprehensive analysis model, import the data transmitted by the resource information analysis module into the comprehensive analysis model, calculate the reasonable value of application resource scheduling, and transmit it to the resource proofreading module; Includes comprehensive analysis unit; Resource proofreading module: including application resource proofreading unit and resource adjustment unit; Human-computer interaction module: import the information of resource proofreading module into the database and send it to relevant managers.

2. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The computing resource status data includes the server motherboard capacitor leakage rate, GPU transistor heating rate and memory page swap rate; the storage resource status data includes the average disk seek time, the solid-state drive flash memory wear leveling times and the storage network bandwidth utilization; the network resource status data includes the network link packet loss rate, the network router cache overflow frequency and the network link round-trip delay jitter value; the server performance data includes the server power conversion efficiency, the number of servers in use and the number of server failures.

3. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The computing resource analysis unit is used to establish a computing resource analysis model, import the computing resource status data transmitted by the resource information acquisition module into the computing resource analysis model, and calculate the computing resource coefficient value. The calculation method of the computing resource analysis model is: ， A1 represents the computing resource coefficient value, wd i represents the capacitance leakage rate of the server motherboard in the ith application, wg i represents the GPU transistor heating rate of the i-th application, wj i-1 represents the memory page swap rate of the i-1th application, wj i represents the memory page swap rate of the i-th application, and n represents the total number of applications.

4. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The storage resource analysis unit is used to establish a storage resource analysis model, import the storage resource status data transmitted by the resource information acquisition module into the storage resource analysis model, and calculate the storage resource coefficient value. The calculation method of the storage resource analysis model is: ， A2 represents the storage resource coefficient value, dg i represents the average disk seek time of the ith application, dt i represents the number of SSD flash memory wear leveling times of the i-th application, dc i It represents the storage network bandwidth utilization of the i-th sub-application, and n represents the total number of applications.

5. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The network resource analysis unit is used to establish a network resource analysis model, import the network resource status data transmitted by the resource information acquisition module into the network resource analysis model, and calculate the network resource coefficient value. The calculation method of the network resource analysis model is: ， A3 represents the network resource coefficient value, fd i represents the network link packet loss rate of the i-th application, dc i represents the network router cache overflow frequency of the i-th application, dr i represents the round-trip delay jitter value of the network link of the ith application, e represents a natural constant, and n represents the total number of applications.

6. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The server performance analysis unit is used to establish a server performance analysis model, import the computing resource status data transmitted by the resource information acquisition module into the server performance analysis model, and calculate the server performance coefficient value. The calculation method of the server performance analysis model is: ， A4 represents the server performance coefficient value, kw i represents the server power conversion efficiency of the i-th application, kt i represents the number of servers used by the i-th application, kz i represents the number of server failures of the i-th application, and n represents the number of applications.

7. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The comprehensive analysis unit is used to establish a comprehensive analysis model, import the computing resource coefficient value, computing resource coefficient value, storage resource coefficient value and network resource coefficient value of the resource information analysis module into the comprehensive analysis model, and calculate the reasonable value of application resource scheduling. The specific calculation formula is: ， Where P represents the reasonable value of application resource scheduling, A1 represents the computing resource coefficient value, A2 represents the storage resource coefficient value, A3 represents the network resource coefficient value, A4 represents the server performance coefficient value, and λ represents other influencing factors of application resource scheduling.

8. The online service resource scheduling system based on a cloud platform according to claim 1, characterized in that: The application resource checking unit is used to check the reasonable value of application resource scheduling, the preset resource interval of application resource scheduling [P c , P l ]; Resource scheduling allocation degree α, specific acquisition method: ， When P <P c or P>P l When , the application resources are abnormal, and enter the resource adjustment unit; when Pc≤P≤Pl, the application resources are in a reasonable allocation state; the resource adjustment unit is used to adjust the resources required by the application, and at the same time generate an adjustment index recorded as η; the optimization index is compared with the resource scheduling allocation degree, and the specific formula is as follows: , ε represents a positive number infinitely close to 0, indicating that the resources required by the application are reasonably scheduled.

Images