A dynamic resource configuration optimization method and system of an information system

Abstract

This invention relates to the field of resource allocation technology, specifically to a dynamic resource allocation optimization method and system for information systems. The method collects information system resource data in real time, analyzes the resource consumption coupling of historical requests among multi-dimensional resources, and judges the rationality of current resource requests by combining the dependency degree of current requests. By identifying differences in resource utilization, it analyzes the combined utilization rate of various resources in the system, and obtains a jitter index based on request rationality. It then dynamically optimizes resource allocation strategies to suppress system performance jitter caused by excessively frequent resource adjustments. This invention implements jitter suppression strategies by analyzing the dependencies between multi-dimensional resources and high utilization conditions, optimizing dynamic allocation, avoiding oversensitivity to a single dimension under high concurrency, improving the accuracy of resource scheduling, and ensuring overall system stability.

Description

Technical Field

[0001] This invention relates to the field of resource allocation technology, and specifically to a dynamic resource allocation optimization method and system for information systems. Background Technology

[0002] In information systems, computing resources, storage resources, and bandwidth resources are all limited and critical elements that need to be shared among multiple tasks. With the increasing concurrency and complexity of business operations, static allocation methods can no longer meet dynamic and changing service demands. The resource requirements of different tasks during operation may be sudden in time, unevenly distributed in space, and significantly different in priority, leading to resource waste or congestion problems with traditional scheduling methods. Therefore, building a real-time perception mechanism for multi-dimensional resource status, comprehensively understanding the utilization rate of various resources, and constructing a resource adaptability model have become key focuses of resource allocation in information systems. Through dynamic scheduling optimization algorithms, the relationship between resources and tasks can be intelligently matched to achieve elastic resource supply and dynamic adjustment mechanisms in multi-task and multi-scenario environments, comprehensively improving resource utilization efficiency, system service quality, and operational stability.

[0003] In high-concurrency scenarios, existing technologies often require information systems to dynamically adjust resources (such as CPU, memory, and bandwidth) based on load to maintain service availability and response performance when implementing dynamic resource allocation through optimization algorithms. However, systems are sensitive to short-term fluctuations, and frequent resource adjustments can cause performance "jitter," where the system frequently switches between different states, leading to decreased stability. For example, a business service might be frequently scheduled due to a sudden spike in CPU usage, triggering network connection rebuilding or load balancer adjustments, which in turn affects the stability and response time of other services. The interaction of multiple resources—CPU, memory, network, disk, and external dependencies—can amplify this jitter, potentially creating a cascading effect that ultimately leads to a chain reaction of performance degradation or even service interruption throughout the entire system. Summary of the Invention

[0004] To address the technical problem in existing technologies where dynamic resource allocation of information systems via optimization algorithms is susceptible to short-term fluctuations in individual resources, potentially leading to frequent resource scheduling triggers and performance jitter in high-concurrency scenarios, this invention aims to provide a dynamic resource allocation optimization method and system for information systems. The specific technical solution adopted is as follows:

[0005] This invention provides a method for dynamic resource allocation optimization in an information system, the method comprising:

[0006] Obtain the resource allocation, resource consumption, and resource request volume for each dimension in various business service requests within the information system;

[0007] In each current request for each business service, based on the similarity of the resource request volume fluctuations of each dimension resource to other dimension resources over a period of time, the demand dependency of each dimension resource on other resources is obtained;

[0008] Analyze the coupling between resource consumption of each dimension and other dimensions in the historical requests of each business service, and combine the demand dependency in the current request to obtain the rationality of the request for each dimension of resources in the current request of each business service;

[0009] In each business service historical request, based on the resource consumption and allocation between each of the two dimensions of resources, the resource utilization difference is analyzed, and combined with the actual resource utilization rate of the two dimensions of resources, the collaborative utilization rate of the two dimensions of resources is obtained; based on the collaborative utilization rate and coupling of each dimension of resources with other dimensions of resources, the joint utilization rate of each dimension of resources is obtained.

[0010] Based on the request rationality and joint utilization rate of each dimension of resources in the current request for each business service, analyze the jitter occurrence index of each dimension of resources; optimize dynamic resource allocation based on the jitter occurrence index.

[0011] Furthermore, the method for obtaining the demand dependency includes:

[0012] For any type of resource in any current request of any business service, the range of resource request volume for that resource in a preset unit of time is taken as the demand fluctuation range of that resource.

[0013] After calculating the difference between two consecutive resource requests for that dimension within a preset unit of time, the average of all resource request differences is taken as the magnitude of demand change for that dimension of resource.

[0014] By combining the range of demand fluctuations and the magnitude of demand changes for resources in this dimension, a demand fluctuation index for resources in this dimension can be obtained.

[0015] In the current request for this business service, negatively correlate the differences in demand fluctuation indicators between this dimension resource and each other dimension resource to obtain the demand dependency of this dimension resource on each other dimension resource.

[0016] Furthermore, the method for obtaining the coupling includes:

[0017] For any dimension resource in any historical request of any business service, obtain the resource consumption sequence of each dimension resource in the time sequence of the historical request of that business service; calculate the correlation between the resource consumption sequence of that dimension resource and each other dimension resource, and obtain the trend correlation between that dimension resource and each other dimension resource.

[0018] By combining the average demand dependency between this dimension resource and each other dimension resource in the historical requests of this business service, as well as the trend correlation, the coupling between this dimension resource and each other dimension resource can be obtained.

[0019] Furthermore, the method for obtaining the reasonableness of the request includes:

[0020] In each current request for each business service, the real-time demand rationality of each dimension resource and each other dimension resource is obtained by combining the demand dependency and coupling of each dimension resource with each other dimension resource.

[0021] The average of the real-time demand rationality of each dimension resource compared to all other dimension resources is used as the rationality of the request for each dimension resource in the current request of each business service.

[0022] Furthermore, the method for obtaining the resource utilization difference includes:

[0023] For any type of resource in any business service historical request, the ratio of the resource consumption to the allocation of each resource in each resource operation in the time sequence of the business service historical request is taken as the resource utilization rate of each resource in each operation.

[0024] After calculating the difference in resource utilization between this resource dimension and each other resource dimension for each instance, the average of all resource utilization differences is taken as the resource utilization difference degree between this resource dimension and each other resource dimension.

[0025] Furthermore, the method for obtaining the collaborative utilization rate includes:

[0026] The average resource utilization rate of each resource dimension in the historical request time series of this business service is taken as the actual utilization rate of each resource dimension; the value of negatively correlated mapping and normalized between the resource utilization difference of this resource dimension and each other resource dimension is taken as the utilization correlation coefficient.

[0027] The cumulative value of the correlation coefficient, the actual utilization rate of the resource in this dimension, and the actual utilization rate of each other dimension resource will be used as the collaborative utilization rate of the resource in this dimension and each other dimension resource.

[0028] Furthermore, the method for obtaining the joint utilization rate includes:

[0029] For any dimension resource in any historical request of any business service, the combined utilization rate of this dimension resource with each other dimension resource is used as a weight to calculate the weighted average of the coupling between this dimension resource and other dimension resources, and the combined utilization rate of this dimension resource is obtained.

[0030] Furthermore, the method for obtaining the jitter occurrence index includes:

[0031] In each current request for each business service, the product of the negative correlation mapping and normalization of the request rationality of each resource dimension and the joint utilization rate is used as the jitter indicator for each resource dimension.

[0032] Furthermore, the optimization of dynamic resource allocation based on jitter occurrence indicators includes:

[0033] The jitter index of all business service corresponding dimension resources is used as a penalty function and introduced into the fitness function of the genetic algorithm. The genetic algorithm is then used to obtain a resource allocation scheme for dynamic resource configuration.

[0034] The present invention also provides a dynamic resource allocation optimization system for an information system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the dynamic resource allocation optimization method for an information system as described above.

[0035] The present invention has the following beneficial effects:

[0036] This invention collects real-time information system resource data, analyzes the resource consumption coupling of historical requests between multi-dimensional resources, and combines this with the current request's dependency to determine the rationality of the current resource request. It quantifies the reality of multi-resource parallel pressure during resource allocation, and allocates resources at real pressure points more quickly, improving the rationality of resource allocation. Furthermore, by identifying differences in resource utilization, it analyzes the combined utilization of various resources in the system, identifying whether multiple resources are simultaneously under high load, avoiding over-response and frequent adjustments due to short-term fluctuations in a single resource. Combined with request rationality, it obtains a jitter index to dynamically optimize resource allocation strategies, suppressing system performance jitter caused by excessively frequent resource adjustments. This invention analyzes the dependencies between multi-dimensional resources and high utilization conditions to implement jitter suppression strategies, optimizes dynamic configuration, avoids oversensitivity to a single dimension under high concurrency, improves the accuracy of resource scheduling, and ensures overall system stability. Attached Figure Description

[0037] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of a dynamic resource allocation optimization method for an information system provided in one embodiment of the present invention;

[0039] Figure 2 This is a schematic diagram of the business service resource relationship of an information system provided in one embodiment of the present invention;

[0040] Figure 3 This is a schematic diagram illustrating the changes in resource request volume across two dimensions, as provided in one embodiment of the present invention.

[0041] Figure 4 This is a schematic diagram illustrating the fluctuation of resource request volume for a single-dimensional resource, provided as an embodiment of the present invention. Detailed Implementation

[0042] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a dynamic resource allocation optimization method and system for an information system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0044] The following description, in conjunction with the accompanying drawings, details a specific scheme for a dynamic resource allocation optimization method and system for an information system provided by the present invention.

[0045] Please see Figure 1 The diagram illustrates a flowchart of a dynamic resource allocation optimization method for an information system according to an embodiment of the present invention. The method includes the following steps:

[0046] S1: Obtain the resource allocation, resource consumption, and resource request volume for each dimension in various business service requests within the information system.

[0047] Business services are the smallest execution units in an information system that directly carry specific business functions. Examples include user interaction services, such as login or payment interface responses in an app; data processing services, such as order calculation or inventory updates during e-commerce promotions; and communication coordination services, such as message passing or API calls between microservices. Each business service requires various resources to operate, such as CPU, memory, and disk I / O, and different services have different resource requirements. For example, data processing services have high CPU and memory requirements, while communication services have high network bandwidth requirements.

[0048] For each business service, information system resource data is collected. Business service instances specifically include service nodes, container or virtual machine instances. Multi-dimensional resources include: CPU utilization, memory usage, disk I / O, network bandwidth, connection count, thread pool status, etc. Please refer to [link / reference]. Figure 2 This diagram illustrates a service resource relationship diagram of an information system provided by an embodiment of the present invention.

[0049] Therefore, in this embodiment of the invention, a lightweight monitoring agent (such as Telegraf) is used to capture operating system and service layer metrics in real time. A distributed log system is used to collect the correspondence between business requests and resource usage, i.e., to obtain the resource request volume for each dimension of business service requests. Combined with traffic metrics provided by the service mesh (Istio), the calls and resource consumption between microservices are monitored, i.e., to obtain the allocation and consumption of each dimension of resources required by the business service. The collected data is transmitted through a message queue, and outlier removal, missing value imputation, data smoothing, and normalization are performed to improve data quality. The data is then stored in a time-series database for subsequent direct analysis. It should be noted that lightweight data collection is a well-known technique in the art and will not be further elaborated here.

[0050] S2: In the current request for each business service, based on the similarity of the resource request volume fluctuations of each dimension resource and other dimension resources over the time period, obtain the demand dependency of each dimension resource on other dimension resources.

[0051] Each business service of an information system consumes certain computing, storage, or network resources. In high-concurrency scenarios, service requests change rapidly within a short period, leading to frequent fluctuations in resource allocation requirements. If the system continuously performs high-frequency resource adjustments, it may cause system instability, such as increased service response latency, frequent context switching, or increased virtual resource migration overhead, thereby affecting the overall system stability and user experience. Therefore, analyzing the resource fluctuations of current requests helps identify the risk of over-adjustment and determine whether resource adjustments are necessary.

[0052] A single business service often consumes multiple types of resources simultaneously, and these resources are often interdependent, especially in high-concurrency scenarios. A single business service may consume multiple resource dimensions concurrently. When one resource experiences a drastic fluctuation, it can trigger a chain reaction of changes in the demand for other resources; for example, a sudden increase in CPU usage may be accompanied by a simultaneous increase in memory and I / O. Please refer to [link / reference]. Figure 3 This diagram illustrates the changes in resource request volume for two dimensions of resources according to an embodiment of the present invention.

[0053] By analyzing the real-time fluctuations in the resource requirements of each business service, the synchronicity or correlation of different resource demands can be quantified. If the demand fluctuations of two resources are relatively similar, it indicates a strong real-time dependency in the use of that business service, i.e., a high degree of demand dependence.

[0054] Preferably, in this embodiment of the invention, the method for obtaining the demand dependency includes:

[0055] In high-concurrency scenarios, such as e-commerce promotions, the demand for computing resources may fluctuate drastically in the short term. Analyzing the changes in resource requests for a single dimension within a unit of time reveals that if the fluctuation in resource requests is excessively large and continues to oscillate in both directions within a short period, it indicates significant real-time fluctuations in resource demand. Please refer to [link / reference]. Figure 4 This diagram illustrates the fluctuation of resource request volume for a single-dimensional resource according to an embodiment of the present invention.

[0056] First, for any resource dimension in the current request of any business service, the range of resource request volume for that resource dimension within a preset unit of time is taken as the demand fluctuation range of that resource dimension. That is, the difference between the maximum and minimum values ​​of all resource request volumes for that resource dimension within a unit of time is used as the demand fluctuation range, reflecting the magnitude of fluctuation in request demand. In this embodiment of the invention, the preset unit of time is set to 1 minute; the specific value can be adjusted by the implementer according to the specific implementation situation.

[0057] Furthermore, after calculating the difference between two consecutive resource requests for that dimension within a preset unit of time, the average of all resource request differences is taken as the demand change range for that dimension of resource, reflecting the demand change range through the difference between adjacent requests.

[0058] Finally, by combining the demand fluctuation range and the demand change magnitude of the resource in this dimension, the demand fluctuation index of the resource in this dimension is obtained. In this embodiment of the invention, the demand fluctuation range and the demand change magnitude of the resource in this dimension are multiplied to obtain the demand fluctuation index of the resource in this dimension. The larger the fluctuation range and the higher the overall change magnitude, the faster the resource demand changes, which characterizes the current request for resource fluctuation.

[0059] By analyzing the similarity of request volume fluctuations for each business service's required dimensional resources, we can reflect the real-time demand dependency. In the current requests for that business service, we negatively correlate the differences in demand fluctuation metrics between that dimensional resource and each of the other dimensional resources to obtain the demand dependency degree between them. The smaller the difference, the higher the dependency between the resources.

[0060] It should be noted that negative correlation mapping is a technique well known to those skilled in the art, such as using inverse proportional values ​​or negative exponential functions with the natural constant as the base, etc., and will not be limited or elaborated here.

[0061] S3: Analyze the coupling between each dimension of resources and other dimension resources in the historical requests of each business service in terms of resource consumption, and combine this with the demand dependency in the current request to obtain the rationality of the request for each dimension of resources in the current request of each business service.

[0062] When resource demand in one dimension of a business service fluctuates drastically, it often leads to a simultaneous increase in demand for resources in other dimensions, resulting in increased overall resource pressure. Therefore, by analyzing the changing trends in resource consumption across business services and identifying the coupling between resources in different dimensions, it is possible to determine whether resources exhibit enhanced correlation.

[0063] By analyzing the changes in resource consumption across various dimensions during the actual historical operation of business services, it is possible to identify whether there are synchronous fluctuation trends among resources. If changes in the usage of one dimension of resources are often accompanied by changes in the usage of another dimension of resources, it indicates a strong coupling between the two.

[0064] Preferably, in this embodiment of the invention, the method for obtaining coupling includes:

[0065] For any dimension resource in any historical request of a business service, obtain the resource consumption sequence of each dimension resource in the time sequence of the historical request of that business service, representing the actual consumption change of a single dimension resource during operation. Then, calculate the correlation between the resource consumption sequences of this dimension resource and each other dimension resource to obtain the trend correlation between the dimension resource and each other dimension resource. In this embodiment of the invention, the similarity calculation between sequences uses the Pearson correlation coefficient. The larger the Pearson correlation coefficient, that is, the greater the trend correlation, the more consistent the changes in the usage and consumption of these two dimension resources during the operation of the business service, and the greater the coupling of the usage of these two dimension resources in the business service.

[0066] By combining the average demand dependency between this dimension resource and each other dimension resource in the historical requests of this business service, and the trend correlation, the coupling between this dimension resource and each other dimension resource is obtained. In this embodiment of the invention, the average demand dependency between this dimension resource and each other dimension resource in the historical requests of this business service is multiplied by the trend correlation to obtain the coupling between this dimension resource and each other dimension resource. The stronger the overall demand dependency in the actual historical operation, and the more similar the consumption trend, the higher the coupling between the two dimension resources, indicating that there is a strong dependency relationship between the dimension resources, that is, the use of this dimension resource is often accompanied by the common consumption of other resources.

[0067] During the operation of each business service, if the current requested dimensional resource is more coupled with other multi-dimensional resources and has a stronger real-time dependency, it indicates that the current request is more likely to be a multi-resource collaborative fluctuation. The more the business service is under real business pressure, the less likely resource adjustments will cause jitter, the more reasonable the request is, and the requested dimensional resources should be allocated and satisfied quickly.

[0068] Preferably, in this embodiment of the invention, the method for obtaining the reasonableness of a request includes:

[0069] In each current request for a business service, the real-time demand rationality of each dimension resource and each other dimension resource is obtained by combining the demand dependency and coupling of each dimension resource with each other dimension resource. In this embodiment of the invention, the product of the demand dependency and coupling of each dimension resource and each other dimension resource is used as the real-time demand rationality of each dimension resource and each other dimension resource. When the real-time demand dependency is higher and the historical coupling reflects the historical dependency more strongly, it indicates that the real-time mutual demand of the two dimension resources is stronger.

[0070] The average of the real-time demand rationality of each dimension resource with all other dimension resources is used as the request rationality of each dimension resource in the current request of each business service. Considering the overall dependency of each dimension resource with other dimension resources, the higher the dependency of a resource on the multi-resource synchronization requirement, the more reasonable the allocation of the request for that dimension resource required by the current business service is, and therefore the greater the request rationality.

[0071] S4: In each business service historical request, based on the resource consumption and allocation between each of the two dimensions of resources, analyze the difference in resource utilization, and combine the actual resource utilization rate of the two dimensions of resources to obtain the collaborative utilization rate of the two dimensions of resources; based on the collaborative utilization rate and coupling of each dimension of resources with other dimensions of resources, obtain the joint utilization rate of each dimension of resources.

[0072] If multiple resources are simultaneously at high utilization rates within the same time period, it indicates that the requested resources are more likely to be the actual needs of the business, rather than excessive response and frequent adjustments caused by short-term fluctuations in a single resource. When multiple resources are under high load and stress at the same time, timely coordination and allocation are necessary to prevent local bottlenecks from spreading into systemic congestion, thereby ensuring service continuity and performance stability.

[0073] In the operation of multi-dimensional resources, if a certain dimension resource is continuously consumed at a high level while another dimension resource is idle for a long time, it indicates that there is a difference in actual utilization between the two in the scheduling and they have failed to form an effective collaborative usage relationship. For example, if the CPU utilization rate is always 80% while the memory utilization rate is always 20%, then there may be errors in the joint state judgment by simply analyzing the trend situation. It is necessary to further comprehensively measure the overall collaborative usage efficiency of each dimension resource with other dimension resources in business services by analyzing the actual utilization and consumption situation during operation.

[0074] In this embodiment of the invention, the method for obtaining the resource utilization difference includes:

[0075] By analyzing the difference between the actual consumption and allocation of resources in two required dimensions during the historical operation of a business service, the degree of matching between resource allocation and actual use can be measured.

[0076] First, for any type of resource in any historical request of a business service, the ratio of the resource consumption to the allocation of each resource in each operation of the historical request of that business service is taken as the resource utilization rate of each resource in each operation, reflecting the actual resource utilization efficiency in the historical operation.

[0077] After further calculating the difference in resource utilization between this dimension resource and each other dimension resource in each instance, the average of all resource utilization differences is taken as the resource utilization difference degree between this dimension resource and each other dimension resource. When the consumption ratio of the two dimensions resources in business services is close, that is, the smaller the resource utilization difference degree, it indicates that the two dimensions resources have a high degree of real synergy in business service execution.

[0078] After determining whether the two types of resources are used in synergy, and combining the actual utilization of the two types of resources in historical operation, an analysis of the historical overall utilization level of each type of resource is conducted to avoid the impact of ineffective configurations with high synergy but low overall utilization on synergy efficiency.

[0079] Preferably, in this embodiment of the invention, the method for obtaining the collaborative utilization rate includes:

[0080] The average resource utilization rate of each resource dimension in the historical request time sequence of this service is taken as the actual utilization rate of each resource dimension. The actual utilization rate reflects the historical average resource consumption of each resource dimension and reflects the historical utilization level.

[0081] The value obtained by negatively correlated and normalized the resource utilization differences between this resource dimension and each other resource dimension is used as the utilization correlation coefficient. The smaller the resource utilization differences, the higher the collaborative efficiency. It should be noted that normalization is a technique well known to those skilled in the art. The choice of normalization can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0082] Ultimately, the cumulative value of the correlation coefficient, the actual utilization rate of the resource in this dimension, and the actual utilization rate of each other dimension will be used as the collaborative utilization rate of the resource in this dimension and each other dimension. The higher the historical synergy and historical utilization level, the higher the collaborative utilization efficiency between the two dimensions, and thus the higher the collaborative utilization rate.

[0083] By combining the coupling between a required dimension resource in each business service and other required resources in all dimensions, i.e. whether they appear in a high degree of synergy during task execution, and by adjusting the synergy utilization rate, we can more comprehensively and realistically represent the overall joint utilization of that dimension resource with other resources in the business service.

[0084] Preferably, in this embodiment of the invention, the method for obtaining the combined utilization rate includes:

[0085] For any resource dimension in any historical request for any business service, the combined utilization rate of that resource dimension with each of the other resource dimensions is used as a weight to calculate the weighted average of the coupling between that resource dimension and other resource dimensions. This yields the joint utilization rate of that resource dimension. By adjusting the coupling analysis through weighted averaging, the simultaneous high utilization level of multiple resources can be more accurately quantified. As an example, methods for obtaining the joint utilization rate include:

[0086] In the formula, Represented as the first The first type of business service required The combined utilization rate of resources across various dimensions This represents the total quantity of resources in other dimensions. Represented as the first The first type of business service required Type of resources and the first dimension Coupling of other dimensions of resources, Represented as the first The first type of business service required Type of resources and the first dimension The collaborative utilization rate of other dimensions of resources.

[0087] S5: Based on the request rationality and joint utilization rate of each dimension of resources in the current request of each business service, analyze the jitter occurrence index of each dimension of resources; optimize dynamic resource configuration based on the jitter occurrence index.

[0088] When the overall utilization of multiple resources is low, and only individual resources experience short-term fluctuations, it may indicate a memory leak or invalid resource usage. For example, if an application fails to release unnecessary objects, memory may swell, resulting in a high memory usage appearance without actual load pressure. In this case, excessive response may cause the system to frequently adjust resource configurations, leading to resource jitter, performance fluctuations, and even increased scheduling overhead.

[0089] Therefore, if the combined utilization rate of the required dimension resources of a certain business service with other multi-dimensional resources in the historical operation process is higher, but the rationality of the current request is lower, it indicates that the current request for this type of dimension resource is not suitable for frequent scheduling and allocation. Otherwise, it may cause system performance jitter, and it is necessary to reduce sensitivity to avoid triggering global rescheduling due to local anomalies.

[0090] In this embodiment of the invention, for each current request of a business service, the product of the negative correlation mapping and normalization of the request rationality of each resource dimension and the joint utilization rate is used as the jitter occurrence index for each resource dimension. A higher joint utilization rate indicates a higher historical level of collaborative strain on multiple resources, but a lower request rationality indicates an unreasonable current request, meaning the current request is more likely due to short-term fluctuations in a single resource, such as a sudden increase in CPU usage. Frequent resource adjustments can easily create a cascading effect, causing significant system jitter, thus resulting in a higher jitter occurrence index.

[0091] Genetic algorithms can optimize the global resource allocation of information systems, dynamically resolve multi-objective conflicts, and avoid local optima. They are used for dynamic resource configuration. In genetic algorithms, the primary objectives are Quality of Service (QoS), resource utilization, and overall system load balancing. A fitness function is constructed, and a jitter suppression penalty function is introduced to suppress large and frequent resource adjustments, limiting unnecessary dynamic configuration behavior. This helps enhance system stability and the robustness of resource allocation strategies, achieving an effective balance between responsiveness and stability in the dynamic resource optimization and adjustment of information systems.

[0092] In this embodiment of the invention, jitter is suppressed by using the jitter occurrence index of the resource allocation dimension required for each business service, thereby optimizing dynamic resource allocation. When the genetic algorithm derives the resource allocation scheme for each round of business services, the jitter occurrence index of the corresponding allocation dimension resources for all business services is calculated as a penalty function and introduced into the fitness function of the genetic algorithm, that is, the penalty term is subtracted, so that resource adjustments are not too frequent. That is, allocation schemes with a high probability of jitter occurrence score lower in fitness evaluation and are thus eliminated in the evolution process.

[0093] Optimizing dynamic resource allocation based on jitter metrics ensures that even if a particular resource allocation scheme offers higher performance gains, excessive system disturbances will reduce overall fitness and suppress its spread within the search space. This effectively avoids frequent system oscillations in dynamic scenarios, improving the smoothness, continuity, and stability of resource allocation strategies and business services.

[0094] In summary, this invention collects real-time information system resource data, analyzes the resource consumption coupling of historical requests between multi-dimensional resources, and combines this with the current request's dependency to determine the rationality of the current resource request. It quantifies the reality of multi-resource parallel pressure during resource allocation, ensuring that resource requests at real pressure points are allocated more quickly, thus improving the rationality of resource allocation. Furthermore, by identifying differences in resource utilization, it analyzes the combined utilization of various resources in the system, identifying whether multiple resources are simultaneously under high load, avoiding over-response and frequent adjustments due to short-term fluctuations in a single resource. Combined with request rationality, it obtains a jitter occurrence index to dynamically optimize resource allocation strategies, suppressing system performance jitter caused by excessively frequent resource adjustments. This invention analyzes the dependencies between multi-dimensional resources and high utilization scenarios to implement jitter suppression strategies, optimizes dynamic configuration, avoids oversensitivity to a single dimension under high concurrency, improves the accuracy of resource scheduling, and ensures overall system stability.

[0095] The present invention also provides a dynamic resource allocation optimization system for an information system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the dynamic resource allocation optimization method for an information system as described above.

[0096] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0097] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for dynamic resource allocation optimization in an information system, characterized in that, The method includes: Obtain the resource allocation, resource consumption, and resource request volume for each dimension in various business service requests within the information system; In each current request for each business service, based on the similarity of the resource request volume fluctuations of each dimension resource to other dimension resources over a period of time, the demand dependency of each dimension resource on other resources is obtained; Analyze the coupling between resource consumption of each dimension and other dimensions in the historical requests of each business service, and combine the demand dependency in the current request to obtain the rationality of the request for each dimension of resources in the current request of each business service; In each business service historical request, based on the resource consumption and allocation between each of the two dimensions of resources, the resource utilization difference is analyzed, and combined with the actual resource utilization rate of the two dimensions of resources, the collaborative utilization rate of the two dimensions of resources is obtained; based on the collaborative utilization rate and coupling of each dimension of resources with other dimensions of resources, the joint utilization rate of each dimension of resources is obtained. Based on the request rationality and joint utilization rate of each dimension of resources in the current request for each business service, analyze the jitter occurrence index of each dimension of resources; optimize dynamic resource allocation based on the jitter occurrence index; The method for obtaining the collaborative utilization rate includes: The average resource utilization rate of each dimension resource in the historical request time sequence of each business service is taken as the actual utilization rate of each dimension resource; the negative correlation mapping and normalization of the resource utilization difference between each dimension resource and each other dimension resource is taken as the utilization correlation coefficient. The cumulative value of the correlation coefficient, the actual utilization rate of each resource dimension, and the actual utilization rate of each other resource dimension will be used as the collaborative utilization rate of each resource dimension and each other resource dimension. The method for obtaining the combined utilization rate includes: For any dimension resource in any business service historical request, the combined utilization rate of this dimension resource with each other dimension resource is used as the weight to calculate the weighted average of the coupling between this dimension resource and other dimension resources, and the joint utilization rate of this dimension resource is obtained. The method for obtaining the jitter occurrence index includes: In each current request for each business service, the product of the negative correlation mapping and normalization of the request rationality of each resource dimension and the joint utilization rate is used as the jitter indicator for each resource dimension.

2. The dynamic resource allocation optimization method for an information system according to claim 1, characterized in that, The methods for obtaining the demand dependency include: For any type of resource in any current request of any business service, the range of resource request volume for that resource in a preset unit of time is taken as the demand fluctuation range of that resource. After calculating the difference between two consecutive resource requests for that dimension within a preset unit of time, the average of all resource request differences is taken as the magnitude of demand change for that dimension of resource. By combining the range of demand fluctuations and the magnitude of demand changes for resources in this dimension, a demand fluctuation index for resources in this dimension can be obtained. In the current request for this business service, negatively correlate the differences in demand fluctuation indicators between this dimension resource and each other dimension resource to obtain the demand dependency of this dimension resource on each other dimension resource.

3. The dynamic resource allocation optimization method for an information system according to claim 1, characterized in that, The method for obtaining the coupling includes: For any dimension resource in any historical request of any business service, obtain the resource consumption sequence of each dimension resource in the time sequence of the historical request of that business service; calculate the correlation between the resource consumption sequence of that dimension resource and each other dimension resource, and obtain the trend correlation between that dimension resource and each other dimension resource. By combining the average demand dependency between this dimension resource and each other dimension resource in the historical requests of this business service, as well as the trend correlation, the coupling between this dimension resource and each other dimension resource can be obtained.

4. The dynamic resource allocation optimization method for an information system according to claim 1, characterized in that, The methods for obtaining the reasonableness of the request include: In each current request for each business service, the real-time demand rationality of each dimension resource and each other dimension resource is obtained by combining the demand dependency and coupling of each dimension resource with each other dimension resource. The average of the real-time demand rationality of each dimension resource compared to all other dimension resources is used as the rationality of the request for each dimension resource in the current request of each business service.

5. The dynamic resource allocation optimization method for an information system according to claim 1, characterized in that, The methods for obtaining the resource utilization difference include: For any type of resource in any business service historical request, the ratio of the resource consumption to the allocation of each resource in each resource operation in the time sequence of the business service historical request is taken as the resource utilization rate of each resource in each operation. After calculating the difference in resource utilization between this resource dimension and each other resource dimension for each instance, the average of all resource utilization differences is taken as the resource utilization difference degree between this resource dimension and each other resource dimension.

6. The dynamic resource allocation optimization method for an information system according to claim 1, characterized in that, The optimization of dynamic resource allocation based on jitter occurrence indicators includes: The jitter index of all business service corresponding dimension resources is used as a penalty function and introduced into the fitness function of the genetic algorithm. The genetic algorithm is then used to obtain a resource allocation scheme for dynamic resource configuration.

7. A dynamic resource allocation optimization system for an information system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the dynamic resource allocation optimization method for an information system as described in any one of claims 1 to 6.

Images