Distributed system resource configuration method, device, equipment, medium and product

Abstract

The invention provides a distributed system resource allocation method which can be applied to the technical field of artificial intelligence. The method comprises the steps of performing structured processing on unstructured data in resource related data based on a preset optical character recognition algorithm to obtain an entity association relationship of resources, and performing special analysis on the entity association relationship through preset functional differentiation analysis agents to obtain special analysis data; carrying out dynamic weighted summation on the special analysis data by utilizing a decision-making agent to obtain fusion analysis data, and generating an execution instruction corresponding to the fusion analysis data in response to that the fusion analysis data passes compliance verification; and calling a regulation and modification program interface according to the execution instruction, and carrying out configuration regulation and modification on the distributed subsystem based on the regulation and modification program interface according to resource category allocation in the resource combination optimization scheme. The invention further provides a distributed system resource configuration device and equipment, a medium and a product.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, specifically to a method for optimizing resource allocation schemes, and more specifically to a method, apparatus, device, medium, and product for resource allocation in a distributed system. Background Technology

[0002] With the rapid development of the financial sector, the distributed systems upon which various financial enterprises or financial groups rely are developing rapidly. Resource integration of distributed systems is the core of efficient operation of various distributed subsystems. The principle of resource allocation for distributed subsystems is to balance resource allocation and resource demand within a compliant framework, thereby allocating resources to each distributed subsystem at the lowest cost and ensuring that the computing power needs of each distributed subsystem are met at the lowest cost.

[0003] The current mainstream resource allocation schemes are mainly divided into the following three categories: (1) Manual experience-driven: Based on macro system operation data, distributed system overall architecture construction reports, branch operation status and other information, the human expert judges the timing of resource combination rebalancing and manually adjusts the resource combination scheme based on personal professional experience. The core of this scheme relies on subjective judgment and industry experience, and the decision-making process lacks standardized process support; (2) Traditional quantitative model: Using fixed rule engines, such as duration matching, business detection strategies or single optimization algorithms to generate computing power combination suggestions. This type of scheme analyzes historical data through mathematical models and replaces some manual judgments through quantitative logic. However, there is data isolation between different mathematical models and there is a lack of effective collaboration mechanism; (3) Automated tool-assisted type: Introducing neural network models to execute resource allocation instructions. However, this type of scheme only applies automation technology to data extraction or end-point adjustment execution links, and does not achieve deep linkage with the decision-making link. The decision-making and execution processes are separated and cannot form a closed-loop response to changes in the branch business.

[0004] Therefore, the relevant technologies have the following technical problems: high dependence on manual labor, slow response to resource allocation, insufficient integration of multi-source heterogeneous data, low data utilization efficiency, disconnect between verification and decision-making processes, separation between decision-making and execution, and inability to form closed-loop management. Summary of the Invention

[0005] In view of the above problems, this application provides a method, apparatus, device, medium and product for resource configuration of distributed systems.

[0006] According to the first aspect of this application, a distributed system resource configuration method is provided, comprising: acquiring the latest resource-related data of each distributed subsystem in the current distributed system through a preset process automation mechanism; performing structured processing on unstructured data in the resource-related data based on a preset optical character recognition algorithm to obtain entity associations of resources, including data center equipment resources and computing resources configured in the data center equipment resources; performing specialized analysis on the entity associations through preset functionally differentiated analytical agents to obtain specialized analysis data, which indicates the resource configuration status of the distributed subsystem in different scenarios; dynamically weighting and summing the specialized analysis data using a preset decision-making agent to obtain fused analysis data; generating an execution instruction corresponding to the fused analysis data in response to the compliance verification of the fused analysis data, which indicates a resource combination optimization scheme corresponding to the fused analysis data; calling a modification program interface according to the execution instruction; configuring and modifying the distributed subsystem according to the resource category allocation in the resource combination optimization scheme based on the modification program interface; and performing real-time data monitoring on the resource categories after configuration modification to update resource-related data.

[0007] According to an embodiment of this application, unstructured data in resource-related data is structured based on a preset optical character recognition algorithm to obtain entity associations of resources. This includes: extracting indicators from resource-related data using the optical character recognition algorithm to obtain key indicators of the distributed subsystem. The key indicators include at least one of the following: computational resource attributes, resource association information, financial indicators, news data of the enterprise corresponding to the distributed subsystem, and market preferences of the enterprise corresponding to the distributed subsystem; importing the key indicators into a preset graph neural network so that the graph neural network generates a graph structure based on the key indicators; and extracting features from the graph structure to obtain entity associations of resources in the distributed subsystem.

[0008] According to embodiments of this application, the various functionally differentiated analytical agents include at least one of a computing power analysis agent, an enterprise demand agent, and a combinatorial optimization agent. The analytical agents perform specialized analyses on entity relationships to obtain specialized analysis data, including: using the computing power analysis agent to estimate computing power values ​​for entity relationships to obtain a computing power scoring matrix representing each distributed subsystem; using the enterprise demand agent to analyze enterprise demands for entity relationships to obtain demand dimension data for each distributed subsystem; using the combinatorial optimization agent to obtain stickiness data for each distributed subsystem based on entity relationships, and optimizing resource allocation based on the stickiness data, computing power scoring matrix, and demand dimension data to obtain a resource optimization plan, as well as combinatorial information data describing the resource optimization plan. The stickiness data indicates the correlation between each distributed subsystem; the computing power scoring matrix, demand dimension data, and combinatorial information data are used as specialized analysis data.

[0009] According to an embodiment of this application, a pre-defined decision-making agent is used to dynamically weight and sum the specialized analysis data to obtain fused analysis data, including: allocating weight ratios based on the branch business fluctuation index issued by the central node of the distributed system; and weighting and summing the computing power scoring matrix, demand dimension data, and combined information data based on the weight ratios to obtain fused analysis data.

[0010] According to an embodiment of this application, the weight ratio is allocated based on the branch company's business volatility index, including: allocating initial ratios to the computing power scoring matrix, demand dimension data, and combined information data, wherein the initial ratios include a first initial weight, a second initial weight, and a third initial weight; updating the initial ratios based on the branch company's business volatility index to obtain the weight ratios; wherein, if the branch company's business volatility index is higher than a preset volatility threshold, the proportion of the first initial weight is reduced to obtain a first dynamic weight, and the proportion of the second initial weight is increased to obtain a second dynamic weight; if the branch company's business volatility index is lower than the preset volatility threshold, the proportion of the first initial weight is increased to obtain a first dynamic weight, and the proportion of the second initial weight is decreased to obtain a second dynamic weight; the first dynamic weight, the second dynamic weight, and the third initial weight are used as the weight ratios; if the branch company's business volatility index is equal to the preset volatility threshold, the initial ratios are used as the weight ratios.

[0011] According to an embodiment of this application, in response to the compliance verification of the fusion analysis data passing, an execution instruction corresponding to the fusion analysis data is generated, including: in response to the compliance verification of the fusion analysis data passing, the resource optimization plan is used as a resource combination optimization scheme; the resources to be added and the resources to be reduced in the resource combination optimization scheme are obtained, and the process of adding the resources to be added and the process of reducing the resources to be reduced are digitized into a preset instruction template to obtain the execution instruction.

[0012] According to an embodiment of this application, the configuration of a distributed subsystem is modified based on the resource category allocation in the resource combination optimization scheme according to the modification program interface. This includes: parsing the resource category allocation in the resource combination optimization scheme to obtain configuration modification data containing the configuration target, configuration quantity, modification range, and modification type; verifying the identity and permissions of the configuration modification data based on a preset encrypted transmission mechanism; and, in response to the configuration modification data passing the identity and permission verification, performing an addition operation on the resource to be added and a reduction operation on the resource to be reduced.

[0013] According to the embodiments of this application, the analysis agent and the decision agent are integrated in a preset large model. Under the constraints of preset analysis prompts, the large model performs special analysis on the entity relationships to obtain special analysis data. Under the constraints of preset decision prompts, the large model performs dynamic weighted summation on the special analysis data to obtain fused analysis data. In response to the compliance verification of the fused analysis data, an execution instruction corresponding to the fused analysis data is generated.

[0014] The second aspect of this application provides a distributed system resource configuration device, comprising: a structured processing module, used to acquire the latest resource-related data of each distributed subsystem in the current distributed system through a preset process automation mechanism, and to perform structured processing on the unstructured data in the resource-related data based on a preset optical character recognition algorithm to obtain the entity association relationship of the resources, including data center equipment resources and the computing resources configured in the data center equipment resources; a special analysis module, used to perform special analysis on the entity association relationship through preset analysis agents with different functions to obtain special analysis data, which indicates the resource configuration status of the distributed subsystem in different scenarios; a data fusion module, used to perform dynamic weighted summation of the special analysis data using a preset decision-making agent to obtain fused analysis data, and to generate an execution instruction corresponding to the fused analysis data in response to the compliance verification of the fused analysis data, which indicates the resource combination optimization scheme corresponding to the fused analysis data; and a configuration execution module, used to call the adjustment program interface according to the execution instruction, and to perform configuration adjustment on the distributed subsystem according to the resource category allocation in the resource combination optimization scheme based on the adjustment program interface, and to perform real-time data monitoring on the resource categories after configuration adjustment to update the resource-related data.

[0015] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0016] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0017] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0018] The above one or more embodiments have the following beneficial effects: Based on process automation mechanisms and optical character recognition algorithms, the entity association relationships of resources are obtained. Then, through various functionally differentiated analytical agents, the entity association relationships are analyzed separately to obtain specialized analysis data. A decision-making agent then performs a weighted summation of the specialized analysis data to obtain fused analysis data, thereby determining whether the fused data is reasonable. If the fused analysis data passes compliance verification, the resource combination optimization scheme corresponding to the fused data is executed. Based on this, the problem of insufficient integration of multi-source heterogeneous data regarding resource demand and resource allocation status in existing distributed system resource allocation can be solved. This addresses the technical problems of decision-making bias in single-model systems, response lags caused by the disconnect between decision-making and execution, and low execution efficiency that are difficult to manage. It aims to improve the efficiency, accuracy, and rationality of resource allocation and combination management in distributed systems. By calling and modifying program interfaces according to execution instructions, and then configuring and modifying the resource category allocation in the resource combination optimization scheme based on these interfaces, the distributed subsystem is improved. Real-time data monitoring of the configured and modified resource categories is then performed to update resource-related data. This achieves a closed loop of data updates and resource allocation optimization based on the updated data, thereby improving the level of intelligent resource allocation and enhancing the scientific nature of resource allocation and the efficiency of distributed system resource management. Attached Figure Description

[0019] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0020] Figure 1 The illustration shows an application scenario diagram of a distributed system resource configuration method, apparatus, device, medium, and program product according to embodiments of this application;

[0021] Figure 2 A flowchart illustrating a distributed system resource allocation method according to an embodiment of this application is shown schematically.

[0022] Figure 3 The diagram illustrates the data flow output of various functionally differentiated analytical agents involved in the distributed system resource allocation method according to an embodiment of this application.

[0023] Figure 4 This illustration shows a schematic diagram of the data flow involved in the distributed system resource allocation method according to an embodiment of this application;

[0024] Figure 5 This schematically illustrates a structural block diagram of a distributed system resource allocation apparatus according to an embodiment of this application; and

[0025] Figure 6A block diagram schematically illustrates an electronic device suitable for implementing a distributed system resource allocation method according to an embodiment of this application. Detailed Implementation

[0026] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0027] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0028] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0029] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0030] In the technical solution of this application, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0031] This application provides a distributed system resource configuration method, apparatus, device, medium, and product. The distributed system resource configuration method includes: acquiring the latest resource-related data of each distributed subsystem in the current distributed system through a preset process automation mechanism; performing structured processing on unstructured data in the resource-related data based on a preset optical character recognition algorithm to obtain entity associations of resources, including data center equipment resources and computing resources configured within the data center equipment resources; performing specialized analysis on the entity associations through preset functionally differentiated analytical agents to obtain specialized analysis data, which indicates the resource configuration status of the distributed subsystem in different scenarios; dynamically weighting and summing the specialized analysis data using a preset decision-making agent to obtain fused analysis data; generating an execution instruction corresponding to the fused analysis data in response to the compliance verification of the fused analysis data, which indicates a resource combination optimization scheme corresponding to the fused analysis data; calling a modification program interface according to the execution instruction; configuring and modifying the distributed subsystem according to the resource category allocation in the resource combination optimization scheme based on the modification program interface; and monitoring the resource categories after configuration modification in real time to update the resource-related data.

[0032] Figure 1 The diagram illustrates an application scenario of a distributed system resource allocation method according to an embodiment of this application.

[0033] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105.

[0034] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0035] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0036] Network 104 may include various connection types, such as wired or wireless communication links or fiber optic cables. Network 104 serves as the medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. Network 104 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0037] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0038] It should be noted that the distributed system resource configuration method provided in this application embodiment can generally be executed by server 105. Correspondingly, the distributed system resource configuration device provided in this application embodiment can generally be located in server 105. The distributed system resource configuration method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the distributed system resource configuration device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0039] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0040] The following will be based on Figure 1 The described scene, through Figures 2-4 The distributed system resource configuration method according to the embodiments of this application will be described in detail.

[0041] Figure 2 A flowchart illustrating a distributed system resource configuration method according to an embodiment of this application is shown.

[0042] like Figure 2 As shown, the distributed system resource configuration method of this embodiment includes S210~S240, as follows:

[0043] Operation S210 acquires the latest resource-related data of each distributed subsystem in the current distributed system through a preset automated process mechanism. Based on a preset optical character recognition algorithm, it performs structured processing on the unstructured data in the resource-related data to obtain the entity associations of resources, including data center equipment resources and the computing resources configured in the data center equipment resources. Operation S220 performs specialized analysis on the entity associations through preset analytical agents with different functions to obtain specialized analysis data, which indicates the resource configuration status of the distributed subsystem in different scenarios. Operation S230 uses a preset decision-making agent to dynamically weight and sum the specialized analysis data to obtain fused analysis data. In response to the compliance verification of the fused analysis data, it generates an execution instruction corresponding to the fused analysis data, which indicates the resource combination optimization scheme corresponding to the fused analysis data. Operation S240 calls the adjustment program interface according to the execution instruction, and adjusts the configuration of the distributed subsystem according to the resource category allocation in the resource combination optimization scheme based on the adjustment program interface. It also performs real-time data monitoring on the resource categories after configuration adjustment to update the resource-related data.

[0044] As an example, the process automation mechanism is a pre-defined Robotic Process Automation (RPA) mechanism. RPA is a business process automation technology based on software robots and artificial intelligence. By simulating human operations on a computer, RPA automatically executes a large number of repetitive, rule-based tasks. It can process data across systems without changing existing IT systems, improving work efficiency and accuracy. The optical character recognition algorithm is OCR (Optical Character Recognition) technology. Based on this, it accurately acquires unstructured data from unstructured text / documents / web pages / images, facilitating the establishment of entity relationships among resources, such as the relationship between a distributed system, a server, a distributed subsystem, and business volume. In short, first, unstructured data, such as images from the operation report of a distributed subsystem, is extracted using RPA+OCR. Then, the unstructured data is structured to obtain structured data and entity relationships among resources. The relationship node graph / graph structure of these entity relationships is stored in a graph database, and then the entity relationships of the resource, such as the relationship between a distributed subsystem and business volume, are stored.

[0045] Then, through pre-defined analytical agents with different functions, specific analyses are conducted on the entity relationships to obtain specialized analysis data. For example, one agent can output a computing power scoring matrix, a second agent can perform enterprise demand analysis based on entity relationships to obtain demand dimension data for each distributed subsystem, and a third agent can calculate a combined rebalancing scheme based on the aforementioned computing power scoring matrix, demand dimension data, and other parameters to obtain a resource optimization plan. For example, reducing the memory of distributed subsystem A by 5% of the original memory. This resource optimization plan corresponds to combined information data, which is parameter data describing the various attributes / names / numbers involved in the resource optimization plan. The aforementioned computing power scoring matrix, demand dimension data, and combined information data can all be used as specialized analysis data.

[0046] Then, the decision-making intelligence agent dynamically weights and sums the specialized analysis data to obtain fused analysis data. Once the fused analysis data passes compliance verification, an execution instruction corresponding to the fused analysis data is generated. This instruction specifies the resource combination optimization scheme corresponding to the fused analysis data. The resources to be added and reduced in this resource combination optimization scheme are then configured and adjusted accordingly. Furthermore, after configuration and adjustment, real-time data monitoring of the resource categories in the resource combination optimization scheme continues to update resource-related data. This allows the system to again obtain the latest resource-related data involved in the current distributed system through a pre-defined automated process mechanism. Then, based on a pre-defined optical character recognition algorithm, the unstructured data in the resource-related data is structured to obtain the entity relationships of the resources. The updated resource combination optimization scheme is then executed again, followed by monitoring. This process is repeated continuously to achieve a closed loop in distributed system resource configuration, improving the execution efficiency of resource combination configuration, increasing the utilization rate of computing resources, and ensuring that each distributed subsystem has the required computing resources at the lowest cost, thus improving the scientific nature and accuracy of computing resource configuration.

[0047] Based on this, entity relationships of resources are obtained using process automation mechanisms and optical character recognition algorithms. Then, differentiating analytical agents perform specialized analyses on these relationships to obtain specialized analysis data. A decision-making agent then weights and sums this specialized analysis data to obtain fused analysis data. This fused data is then used to determine its rationality. If the fused analysis data passes compliance verification, the corresponding resource combination optimization scheme is executed. This approach addresses the shortcomings in existing distributed system resource allocation and configuration management, such as insufficient integration of multi-source heterogeneous data regarding resource requirements and allocation status, single-model decision bias, and the lack of integration between decision-making and resource allocation. This addresses the technical issues of delayed response, low execution efficiency, and difficulty in managing compliance risks caused by fragmented execution. It aims to improve the efficiency, accuracy, rationality, and compliance of resource allocation and combination management in distributed systems. The system invokes and modifies program interfaces based on execution instructions, and then adjusts the configuration of distributed subsystems according to resource categories in the resource combination optimization scheme. Real-time data monitoring of the configured resource categories is conducted to update resource-related data, thereby achieving a closed loop of data updates and resource allocation optimization based on the updated data. This improves the level of intelligent resource allocation, enhances the scientific nature of resource allocation and combination, and increases the efficiency of distributed system resource management.

[0048] In this embodiment, unstructured data in resource-related data is structured based on a preset optical character recognition algorithm to obtain entity relationships of resources. This includes: extracting indicators from resource-related data using the optical character recognition algorithm to obtain key indicators of the distributed subsystem. Key indicators include at least one of the following: computational resource attributes, resource association information, financial indicators, news data of the enterprise corresponding to the distributed subsystem, and market trends of the enterprise corresponding to the distributed subsystem; importing the key indicators into a preset graph neural network so that the graph neural network generates a graph structure based on the key indicators; and extracting features from the graph structure to obtain entity relationships of resources in the distributed subsystem.

[0049] As an example, the OCR algorithm first extracts key indicators from resource-related data, such as computing resource attributes, resource association information, financial indicators, and news data of the enterprise corresponding to the distributed subsystem. Computing resource attributes include, for example, the model of the central processing unit, memory capacity, and video memory; resource association information includes, for example, server S1 - distributed subsystem 2 - R&D business module; financial indicators include, for example, the annual maintenance cost of server S1 is 50,000 yuan; and news data of the enterprise corresponding to the distributed subsystem includes, for example, the content published by the enterprise and the market orientation of the branch company. Then, a graph neural network is used to establish the relationship between resources and distributed subsystems based on these key indicators, establishing relationships between entities such as computing power inventory / current computing power and computing power demand / resource demand in each distributed subsystem. This generates a graph structure containing the entity associations of resources, which is stored in a graph database and then associated with all graph structures in the database. Feature extraction is then performed on the associated graph structures to obtain the entity associations of the resource. For example, for a resource-system relationship, the association between "branch company B - distributed subsystem B" and "distributed subsystem B - distributed subsystem" can be obtained.

[0050] Based on this, this paper addresses the technical problems in the resource allocation process of distributed subsystems, such as the scattered processing of multi-source heterogeneous data and the lack of dynamic correlation analysis, which leads to low data utilization efficiency and incomplete decision-making basis. It achieves efficient extraction and structured storage of unstructured data, and establishes multi-dimensional entity associations, making data query and correlation analysis more convenient and faster in response. This provides comprehensive and accurate data support for subsequent collaborative decision-making by intelligent agents, avoiding decision-making biases caused by data fragmentation.

[0051] In this embodiment, the various functionally differentiated analytical agents include at least one of a computing power analysis agent, an enterprise demand agent, and a combinatorial optimization agent. The analytical agents perform specialized analyses on entity relationships to obtain specialized analysis data, including: using the computing power analysis agent to estimate computing power values ​​for entity relationships, obtaining a computing power scoring matrix representing each distributed subsystem; using the enterprise demand agent to analyze enterprise demands for entity relationships, obtaining demand dimension data for each distributed subsystem; using the combinatorial optimization agent to obtain stickiness data for each distributed subsystem based on entity relationships, and optimizing resource allocation based on the stickiness data, computing power scoring matrix, and demand dimension data to obtain a resource optimization plan, as well as combinatorial information data describing the resource optimization plan. The stickiness data indicates the correlation between each distributed subsystem. The computing power scoring matrix, demand dimension data, and combinatorial information data are used as specialized analysis data.

[0052] As an example, a computing power analysis agent is used to estimate computing power values ​​based on entity relationships, resulting in a computing power score matrix representing each distributed subsystem. For instance, based on the entity relationship of "server-distributed subsystem-computing load," a computing power estimation model is used, combining CPU utilization, memory usage, and task processing latency, to estimate the computing power of each distributed subsystem and generate a computing power score matrix. The feature values ​​corresponding to the computing power score matrix correspond to specific score values; for example, a maximum score of 100 points is used, and the distributed subsystem can be simply referred to as a subsystem. For instance, subsystem 1 has a score of 85, subsystem 2 has a score of 62, subsystem 3 has a score of 78, and subsystem 4 has a score of 90. Subsystem 2 suffers from insufficient computing power due to a surge in tasks. Furthermore, an enterprise demand agent is used to conduct sentiment analysis on entity relationships, obtaining demand dimension data for each distributed subsystem. For example, subsystem 2 needs an additional 30%. Graphics processing computing power; Subsystem 3 needs to retain 15% of its central processing unit resources; Subsystem 1 has no new requirements. Using a combinatorial optimization agent, sticky data of each distributed subsystem is obtained based on entity relationships. Resource allocation is optimized based on this sticky data, the computing power scoring matrix, and the demand dimension data to obtain a resource optimization plan and its combined information data. The sticky data characterizes the relationships between resources / computing resources / equipment resources of each category. For example, the data interaction frequency between subsystems is: Subsystem 1 interacts with Subsystem 2 20 times / hour, and Subsystem 2 interacts with Subsystem 3 5 times / hour. 20% of the central processing unit resources of Subsystem 1 and 10% of the graphics processing resources of Subsystem 3 are allocated to Subsystem 2, generating combined information data, including the type and quantity of allocated resources, the target subsystem, and the configuration information of the execution window, thus obtaining the resource optimization plan.

[0053] Based on this, the current traditional quantification model is single and lacks a collaborative mechanism, resulting in large decision-making biases and an inability to simultaneously consider current computing power and current demand dimensions. By having intelligent agents with differentiated functions work together, the current computing power and current demand dimensions can be captured synchronously. Combined with the combined optimization requirements, targeted rebalancing resource optimization plans are generated, avoiding the bias of a single model, making the decision more comprehensive and accurate, and reducing the tracking error of resource combination schemes.

[0054] In this embodiment, a pre-defined decision-making agent is used to dynamically weight and sum the specialized analysis data to obtain fused analysis data. This includes: allocating weight ratios based on the branch business fluctuation index issued by the central node of the distributed system; and weighting and summing the computing power scoring matrix, demand dimension data, and combined information data based on the weight ratios to obtain fused analysis data.

[0055] As an example, the decision-making agent calculates a comprehensive score based on weights, such as 40% for the computing power scoring matrix, 30% for demand dimension data, and 30% for combined information data. It then confirms whether the rebalancing scheme meets compliance requirements. If it does, it generates an execution instruction. The decision-making agent adopts a dynamic weighting mechanism, meaning that the weight allocation is dynamic. The weights are determined based on the allocated weight ratios, and then dynamically weighted summation is performed based on the dynamically changing weights to obtain fused analysis data. This allows for real-time data monitoring of resource categories, improving the flexibility and adaptability of decision-making and ensuring that resource allocation is tilted towards high-demand, high-fluctuation businesses.

[0056] In this embodiment, the weight ratio is allocated according to the branch business fluctuation index, including: allocating initial ratios to the computing power scoring matrix, demand dimension data, and combined information data respectively, the initial ratios including a first initial weight, a second initial weight, and a third initial weight; updating the initial ratios according to the branch business fluctuation index to obtain the weight ratios; wherein, when the branch business fluctuation index is higher than a preset fluctuation threshold, the proportion of the first initial weight is reduced to obtain a first dynamic weight, and the proportion of the second initial weight is increased to obtain a second dynamic weight; when the branch business fluctuation index is lower than the preset fluctuation threshold, the proportion of the first initial weight is increased to obtain a first dynamic weight, and the proportion of the second initial weight is decreased to obtain a second dynamic weight; the first dynamic weight, the second dynamic weight, and the third initial weight are used as the weight ratios; when the branch business fluctuation index is equal to the preset fluctuation threshold, the initial ratios are used as the weight ratios.

[0057] As an example, the threshold for the business fluctuation index of the distributed subsystem is preset to 30. The initial weights are as follows: the first initial weight for the computing power scoring matrix is ​​40%, the second initial weight for the demand dimension data is 30%, and the third initial weight for the combined information data is 30%. The fluctuation index of the distributed subsystem 1 obtained in real time is 35, which is higher than the threshold of 30. The decision-making agent triggers dynamic weight adjustment: the second initial weight of the enterprise demand agent result is increased to 45% as the second dynamic weight, the first initial weight of the computing power analysis agent result is decreased to 25% as the first dynamic weight, and the third initial weight of the combined optimization agent result remains unchanged at 30%.

[0058] At this point, the computing power analysis agent outputs the computing power score matrix of the distributed subsystem. The value of this score matrix is ​​70, and no warning is triggered. The decision-making agent weights and fuses the data according to the adjusted weights (25%+45%+30%) to obtain fused analysis data. If the fused analysis data is higher than the preset compliance threshold, it can be determined to be compliant. Alternatively, if the fused analysis data is lower than another preset compliance threshold, it can also be determined to be compliant. As long as the threshold and the specific judgment method are determined, the specific compliance judgment method is not restricted. After compliance verification, an RPA is generated. In a more specific example, the computing power analysis agent translates the compliance rules for resource allocation into mathematical constraints. These compliance rules include computing power scheduling constraints, cost budget constraints, business continuity constraints, and compliance verification constraints. These mathematical constraints are embedded in the generation process of the combined rebalancing scheme. For instance, the preset compliance rules are: peak computing power scheduling should not exceed 80% of the total cluster capacity, quarterly computing power cost should not exceed 35% of the business budget, and core business computing power redundancy should not be less than 15%. The combined optimization agent translates these compliance rules into mathematical constraints: for example, the computing power scheduling constraint is: peak computing power scheduling should not exceed 80% of the total cluster computing power capacity, and non-core business computing power scheduling priority is lower than core business, with its peak usage not exceeding 30% of the total computing power. For the cost budget constraint, the mathematical constraint is set as: the total quarterly computing power cost should not exceed 35% of the total quarterly business budget. The business continuity constraints include: core business computing power redundancy should not be less than 15%, and when the cluster computing power fluctuates within 10%, the core business computing power supply should not be interrupted, ensuring the scheme complies with all compliance rules.

[0059] Thus, the dynamic weight adjustment mechanism based on the distributed subsystem business volatility index enables decision-making to adapt to market volatility in real time, establishes standardized and quantifiable weight adjustment rules, prioritizes the prevention and control of resource allocation risks during periods of high market volatility, and focuses on precise analysis of computing power when the market is stable. This ensures that the weight allocation logic of the decision-making agent is consistent and traceable under different business volatility states, improves the flexibility and adaptability of decision-making, and further reduces portfolio risk exposure.

[0060] In this embodiment, in response to the compliance verification of the fusion analysis data passing, an execution instruction corresponding to the fusion analysis data is generated, including: in response to the compliance verification of the fusion analysis data passing, the resource optimization plan is used as a resource combination optimization scheme; the resources to be added and the resources to be reduced in the resource combination optimization scheme are obtained, and the process of adding the resources to be added and the process of reducing the resources to be reduced are digitized into a preset instruction template to obtain the execution instruction.

[0061] As an example, the aforementioned combinatorial optimization agent has already output a resource optimization plan. The relevant combinatorial information data of this resource optimization plan is part of the fusion analysis data. After the fusion analysis data passes the consistency check, an RPA execution instruction is generated. When generating the execution instruction, the resource optimization plan is first used as a resource combinatorial optimization scheme. Then, the resources to be added and the resources to be reduced in the resource combinatorial optimization scheme are determined. The process of adding the resources to be added and the process of reducing the resources to be reduced are digitized into a preset instruction template to obtain the execution instruction. Based on this execution instruction, the corresponding additions and reductions can be performed. Based on this, the additions and reductions in the execution instruction are executed efficiently and quickly, improving the implementation efficiency of the combinatorial configuration scheme and the scientific nature of market response.

[0062] In this embodiment, the distributed subsystem is configured and modified based on the resource category allocation in the resource combination optimization scheme according to the modification program interface. This includes: parsing the resource category allocation in the resource combination optimization scheme to obtain configuration modification data containing the configuration target, configuration quantity, modification range, and modification type; verifying the identity and permissions of the configuration modification data based on a preset encrypted transmission mechanism; and in response to the configuration modification data passing the identity and permission verification, performing an addition operation on the resource to be added and a reduction operation on the resource to be reduced.

[0063] As an example, the RPA engine that executes the command uses an encrypted transmission protocol to interface with the adjustment API (Application Programming Interface). Before executing a resource configuration delegation, authentication and permission verification are required. The RPA execution command includes information such as the configuration target, configuration quantity, adjustment range, and adjustment type. For instance, when the RPA engine interfaces with the adjustment API, an encrypted transmission protocol can be used to ensure data security during command transmission. The RPA execution command generated by the decision-making agent includes: the configuration target, such as CPU resources; the configuration quantity (e.g., adding 3 equivalent CPU resources to subsystem 1, reducing 3 equivalent CPU resources to subsystem 2); the adjustment range (resource cost sharing ratio 0%-5%); and the adjustment type, such as immediate adjustment. Before executing the adjustment, the RPA engine verifies authentication and role permissions, ensuring only system administrators and data center managers have execution permissions. After successful verification, the adjustment delegation is submitted. After submission, the execution status returned by the adjustment API is received in real time until the adjustment is completed. The transaction results for adding resources or reducing resources are then fed back to the system.

[0064] Based on this, by using encrypted transmission protocols, authentication, and permission verification, the transmission and execution security of RPA execution instructions are ensured, avoiding the risks of data leakage and unauthorized operations. At the same time, the core information of the execution instructions is clearly defined to ensure the accuracy of configuration modification delegation and further reduce the execution error rate.

[0065] In this embodiment, the analysis agent and the decision agent are integrated into a preset large model. Under the constraints of preset analysis prompts, the large model performs specialized analysis on the entity relationships to obtain specialized analysis data. Under the constraints of preset decision prompts, the large model dynamically weights and sums the specialized analysis data to obtain fused analysis data. In response to the compliance verification of the fused analysis data, an execution instruction corresponding to the fused analysis data is generated.

[0066] As an example, the "large model" refers to an artificial intelligence (AI) large model, or simply "large model." It is a type of AI model with a large number of parameters, constructed from artificial neural networks. Major categories of AI large models include: large language models, visual large models, multimodal large models, and basic science large models. For instance, under the constraint of preset analysis prompts, the enterprise demand agent uses natural language processing technology within the large language model to analyze industry news, social media sentiment, and business scenarios to obtain demand dimension data for each distributed subsystem. Based on the branch's business fluctuation index, the enterprise demand agent then... The rules dynamically allocate weight ratios to perform weighted summation on the specialized analysis data, retaining three significant figures. The computing power analysis agent works similarly, utilizing the language model within the larger model for analysis and prediction. Under the constraint of preset analysis prompts, this combinatorial optimization agent uses the decision-making model or multimodal model within the larger model to reorganize solutions. Under the constraint of decision prompts, the decision-making agent uses the decision model or multimodal model within the larger model to dynamically weight and sum the specialized analysis data, obtaining fused analysis data. Upon successful compliance verification of the fused analysis data, execution instructions corresponding to the fused analysis data are generated. Based on this, a closed-loop management system for distributed system resource allocation—"data acquisition - specialized analysis - collaborative decision-making - compliance execution - real-time monitoring"—is achieved. This improves the efficiency of unstructured data extraction, shortens the response time of graph database association analysis, reduces combinatorial tracking errors, and significantly improves the efficiency, accuracy, and compliance of combinatorial management.

[0067] As described above, the distributed system resource configuration method provided in this embodiment can solve the technical problems existing in the current resource combination management, such as insufficient integration of multi-source heterogeneous data, decision bias of single model, response lag caused by the separation of decision and execution, low execution efficiency, and difficulty in controlling compliance risks. It can improve the efficiency, accuracy and compliance of combination management. The method calls the adjustment program interface according to the execution instruction, and adjusts the configuration of the distributed subsystem in the resource combination optimization scheme based on the adjustment program interface. It also monitors the resource categories in the resource combination optimization scheme in real time to update the resource-related data, thereby realizing a closed loop of data update and resource configuration combination optimization based on the updated data. This improves the level of intelligent resource configuration and enhances the scientific nature and management efficiency of distributed system resource configuration.

[0068] Based on the above-described distributed system resource allocation method, this application also provides a distributed system resource allocation apparatus. The following will be combined with... Figure 5 The device is described in detail.

[0069] Figure 5 A schematic block diagram of a distributed system resource allocation apparatus according to an embodiment of this application is shown.

[0070] like Figure 5 As shown, the distributed system resource configuration device 500 of this embodiment includes a structured processing module 510, a special analysis module 520, a data fusion module 530, and a configuration execution module 540.

[0071] The structured processing module 510 executes operation S210, acquiring the latest resource-related data of each distributed subsystem in the current distributed system through a preset automated process mechanism. Based on a preset optical character recognition algorithm, it performs structured processing on the unstructured data in the resource-related data to obtain the entity relationships of the resources. The resources include data center equipment resources and the computing resources configured within those resources. The specialized analysis module 520 executes operation S220, performing specialized analysis on the entity relationships through preset, functionally differentiated analytical agents to obtain specialized analysis data. This specialized analysis data indicates the resource allocation of the distributed subsystems in different scenarios. The data fusion module 530 performs operation S230, dynamically weighting and summing the special analysis data using a preset decision-making intelligent agent to obtain fused analysis data. In response to the compliance verification of the fused analysis data, it generates an execution instruction corresponding to the fused analysis data, which indicates the resource combination optimization scheme corresponding to the fused analysis data. The configuration execution module 540 performs operation S240, calling the adjustment program interface according to the execution instruction. Based on the adjustment program interface, it performs configuration adjustment on the distributed subsystem according to the resource category allocation in the resource combination optimization scheme, and performs real-time data monitoring on the resource categories after configuration adjustment to update resource-related data.

[0072] In this embodiment, the structured processing module 510 performs structured processing on unstructured data in resource-related data based on a preset optical character recognition algorithm to obtain the entity association relationship of the resources. This includes: using the optical character recognition algorithm to extract indicators from the resource-related data to obtain key indicators of the distributed subsystem. The key indicators include at least one of the following: computing resource attributes, resource association information, financial indicators, news data of the enterprise corresponding to the distributed subsystem, and market preferences of the enterprise corresponding to the distributed subsystem; importing the key indicators into a preset graph neural network so that the graph neural network generates a graph structure based on the key indicators; and extracting features from the graph structure to obtain the entity association relationship of the resources in the distributed subsystem.

[0073] The specialized analysis module 520 includes at least one of the following functionally differentiated analytical agents: a computing power analysis agent, an enterprise demand agent, and a combinatorial optimization agent. Each analytical agent performs specialized analysis on entity relationships to obtain specialized analysis data. This includes: using the computing power analysis agent to estimate computing power values ​​based on entity relationships, obtaining a computing power scoring matrix representing each distributed subsystem; using the enterprise demand agent to analyze enterprise demands based on entity relationships, obtaining demand dimension data for each distributed subsystem; and using the combinatorial optimization agent to obtain stickiness data for each distributed subsystem based on entity relationships, and optimizing resource allocation based on the stickiness data, computing power scoring matrix, and demand dimension data to obtain a resource optimization plan, as well as combinatorial information data describing the resource optimization plan. The stickiness data indicates the correlation between each distributed subsystem. The computing power scoring matrix, demand dimension data, and combinatorial information data are used as specialized analysis data.

[0074] The data fusion module 530 uses a pre-set decision-making agent to dynamically weight and sum the specialized analysis data to obtain fused analysis data. This includes: allocating weight ratios based on the branch company business fluctuation index issued by the central node of the distributed system; weighting and summing the computing power scoring matrix, demand dimension data, and combined information data based on the weight ratios to obtain fused analysis data; and allocating weight ratios based on the branch company business fluctuation index, including: assigning initial ratios to the computing power scoring matrix, demand dimension data, and combined information data respectively, where the initial ratios include a first initial weight, a second initial weight, and a third initial weight; updating the initial ratios based on the branch company business fluctuation index to obtain the weight ratios; wherein, if the branch company business fluctuation index is higher than a preset fluctuation threshold, the proportion of the first initial weight is reduced to obtain a first dynamic weight, and the proportion of the second initial weight is increased. The proportion of the weight is used to obtain the second dynamic weight. When the branch's business volatility index is lower than the preset volatility threshold, the proportion of the first initial weight is increased to obtain the first dynamic weight, and the proportion of the second initial weight is decreased to obtain the second dynamic weight. The first dynamic weight, the second dynamic weight, and the third initial weight are used as the weight ratio. When the branch's business volatility index is equal to the preset volatility threshold, the initial ratio is used as the weight ratio. In response to the compliance verification of the fusion analysis data, execution instructions corresponding to the fusion analysis data are generated, including: in response to the compliance verification of the fusion analysis data, the resource optimization plan is used as the resource combination optimization scheme; the resources to be added and the resources to be reduced in the resource combination optimization scheme are obtained, and the process of adding the resources to be added and the process of reducing the resources to be reduced are digitized into the preset instruction template to obtain the execution instructions.

[0075] The configuration execution module 540 performs configuration adjustments on the distributed subsystem based on the resource category allocation in the resource combination optimization scheme according to the adjustment program interface. This includes: parsing the resource category allocation in the resource combination optimization scheme to obtain configuration adjustment data containing the configuration target, configuration quantity, adjustment range, and adjustment type; verifying the identity and permissions of the configuration adjustment data based on a preset encrypted transmission mechanism; and, in response to the configuration adjustment data passing the identity and permission verification, performing addition operations on new resources and reduction operations on resources to be reduced. The analysis agent and decision agent are integrated into a preset large model. Under the constraints of preset analysis prompts, the large model performs specialized analysis on entity relationships to obtain specialized analysis data. Under the constraints of preset decision prompts, the large model dynamically weights and sums the specialized analysis data to obtain fused analysis data. In response to the compliance verification of the fused analysis data passing, an execution instruction corresponding to the fused analysis data is generated.

[0076] Furthermore, according to embodiments of this application, any multiple modules of the structured processing module 510, the specialized analysis module 520, the data fusion module 530, and the configuration execution module 540 can be merged into one module, or any one of these modules can be split into multiple modules. Alternatively, at least some of the functions of one or more of these modules can be combined with at least some of the functions of other modules and implemented in one module. According to embodiments of this application, at least one of the structured processing module 510, the specialized analysis module 520, the data fusion module 530, and the configuration execution module 540 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the structured processing module 510, the special analysis module 520, the data fusion module 530, and the configuration execution module 540 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0077] It should be noted that the implementation methods, technical problems solved, functions achieved, and technical effects of each module in the device embodiment are the same as or similar to the implementation methods, technical problems solved, functions achieved, and technical effects of each corresponding step in the method embodiment, and will not be repeated here.

[0078] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a distributed system resource allocation method according to an embodiment of this application.

[0079] like Figure 6 As shown, an electronic device 600 according to an embodiment of this application includes a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage portion 608 into a random access memory (RAM) 603. The processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 601 may also include onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0080] RAM 603 stores various programs and data required for the operation of electronic device 600. Processor 601, ROM 602, and RAM 603 are interconnected via bus 604. Processor 601 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 602 and / or RAM 603. It should be noted that the programs may also be stored in one or more memories other than ROM 602 and RAM 603. Processor 601 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0081] According to embodiments of this application, the electronic device 600 may further include an input / output (I / O) interface 605, which is also connected to a bus 604. The electronic device 600 may also include one or more of the following components connected to the input / output (I / O) interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the input / output (I / O) interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 610 as needed so that computer programs read from it can be installed into the storage section 608 as needed.

[0082] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0083] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 602 and / or RAM 603 and / or one or more memories other than ROM 602 and RAM 603 described above.

[0084] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to enable the computer system to implement the distributed system resource allocation method provided in the embodiments of this application.

[0085] When the computer program is executed by the processor 601, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0086] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 609, and / or installed from the removable medium 611. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0087] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 609, and / or installed from the removable medium 611. When the computer program is executed by the processor 601, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0088] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0089] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0090] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A method for resource allocation in a distributed system, characterized in that, The method includes: The latest resource-related data of each distributed subsystem in the current distributed system is obtained through a preset process automation mechanism. The unstructured data in the resource-related data is structured based on a preset optical character recognition algorithm to obtain the entity association relationship of the resources. The resources include data center equipment resources and the computing resources configured in the data center equipment resources. The entity relationships are analyzed by pre-defined analytical agents with different functions to obtain specialized analysis data, which indicates the resource configuration status of the distributed subsystem in different scenarios. The special analysis data is dynamically weighted and summed using a pre-set decision-making intelligent agent to obtain fused analysis data. In response to the compliance verification of the fused analysis data, an execution instruction corresponding to the fused analysis data is generated. The execution instruction indicates the resource combination optimization scheme corresponding to the fused analysis data. The execution instruction calls the adjustment program interface, and based on the adjustment program interface, the distributed subsystem is configured and adjusted according to the resource category allocation in the resource combination optimization scheme. The resource categories after configuration adjustment are monitored in real time to update the resource-related data.

2. The method according to claim 1, characterized in that, Based on a preset optical character recognition algorithm, unstructured data in the resource-related data is processed into structured data to obtain entity associations of the resources, including: The optical character recognition algorithm is used to extract indicators from the resource-related data to obtain key indicators of the distributed subsystem. The key indicators include at least one of the following: computing resource attributes, resource association information, financial indicators, news data of the enterprise corresponding to the distributed subsystem, and market orientation of the enterprise corresponding to the distributed subsystem. The key indicators are imported into a preset graph neural network so that the graph neural network generates a graph structure based on the key indicators; Feature extraction is performed on the graph structure to obtain the entity association relationships of the resources in the distributed subsystem.

3. The method according to claim 2, characterized in that, The various functionally differentiated analytical agents include at least one of a computing power analysis agent, an enterprise demand agent, and a combinatorial optimization agent. The analytical agents perform specialized analyses on the entity relationships to obtain specialized analysis data, including: The computing power analysis agent is used to estimate the computing power value of the entity association relationship to obtain a computing power score matrix representing each distributed subsystem. The enterprise demand agent is used to analyze the enterprise demand of the entity association relationship to obtain demand dimension data of each distributed subsystem. The combinatorial optimization agent is used to obtain stickiness data of each distributed subsystem based on the entity association relationship. Resource allocation optimization is performed based on the stickiness data, the computing power score matrix and the demand dimension data to obtain a resource optimization plan and combinatorial information data to describe the resource optimization plan. The stickiness data indicates the association between the distributed subsystems. The computing power scoring matrix, demand dimension data, and combined information data are used as specialized analysis data.

4. The method according to claim 3, characterized in that, The specific analysis data is dynamically weighted and summed using a pre-defined decision-making intelligent agent to obtain fused analysis data, including: The weighting ratio is assigned based on the branch business fluctuation index issued by the central node of the distributed system. The computing power scoring matrix, demand dimension data, and combined information data are weighted and summed based on the weight ratio to obtain fusion analysis data.

5. The method according to claim 4, characterized in that, The weighting ratios are allocated based on the branch company's business volatility index, including: Initial ratios are assigned to the computing power scoring matrix, the demand dimension data, and the combined information data, respectively, and the initial ratios include a first initial weight, a second initial weight, and a third initial weight. The initial ratio is updated based on the branch company's business volatility index to obtain the weighted ratio; wherein... When the branch office's business volatility index is higher than a preset volatility threshold, the proportion of the first initial weight is reduced to obtain the first dynamic weight, and the proportion of the second initial weight is increased to obtain the second dynamic weight. When the branch office's business volatility index is lower than a preset volatility threshold, the proportion of the first initial weight is increased to obtain the first dynamic weight, and the proportion of the second initial weight is reduced to obtain the second dynamic weight. The first dynamic weight, the second dynamic weight, and the third initial weight are used as the weight ratio. When the branch office's business volatility index equals a preset volatility threshold, the initial ratio is used as the weighting ratio.

6. The method according to claim 3, characterized in that, In response to the compliance verification of the fused analysis data passing, an execution instruction corresponding to the fused analysis data is generated, including: In response to the successful compliance verification of the fused analysis data, the resource optimization plan will be adopted as the resource combination optimization scheme. Obtain the resources to be added and the resources to be reduced in the resource combination optimization scheme, and digitize the process of adding the resources to be added and the process of reducing the resources to be reduced into a preset instruction template to obtain the execution instruction.

7. The method according to claim 6, characterized in that, The configuration adjustment of the distributed subsystem based on the resource category allocation in the resource combination optimization scheme according to the adjustment program interface includes: Data parsing is performed on the resource category allocation in the resource combination optimization scheme to obtain configuration adjustment data including configuration target, configuration quantity, adjustment range, and adjustment type; Based on a preset encrypted transmission mechanism, the configuration modification data is authenticated and its permissions are verified. In response to the configuration modification data passing the authentication and permission verification, the resource to be added is added, and the resource to be reduced is reduced.

8. The method according to claim 1, characterized in that, The analytical agent and the decision-making agent are integrated into a pre-set large model. Under the constraints of pre-set analytical prompts, the large model performs specialized analysis on the entity relationships to obtain specialized analysis data. Under the constraints of pre-set decision prompts, the large model dynamically weights and sums the specialized analysis data to obtain fused analysis data. In response to the compliance verification of the fused analysis data, an execution instruction corresponding to the fused analysis data is generated.

9. A distributed system resource allocation device, characterized in that, The device includes: The structured processing module is used to obtain the latest resource-related data of each distributed subsystem in the current distributed system through a preset process automation mechanism, and to perform structured processing on the unstructured data in the resource-related data based on a preset optical character recognition algorithm to obtain the entity association relationship of the resources. The resources include data center equipment resources and the computing resources configured in the data center equipment resources. The specialized analysis module is used to perform specialized analysis on the entity relationships through preset analytical agents with different functions to obtain specialized analysis data, which indicates the resource configuration status of the distributed subsystem in different scenarios. The data fusion module is used to dynamically weight and sum the special analysis data using a preset decision intelligence agent to obtain fused analysis data. In response to the compliance verification of the fused analysis data, an execution instruction corresponding to the fused analysis data is generated. The execution instruction indicates the resource combination optimization scheme corresponding to the fused analysis data. The configuration execution module is used to call the adjustment program interface according to the execution instruction, and to adjust the configuration of the distributed subsystem according to the resource category allocation in the resource combination optimization scheme based on the adjustment program interface, and to perform real-time data monitoring on the resource categories after configuration adjustment in order to update the resource-related data.

10. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 8.

Images