Flow control method, computer program product, device and storage medium

Abstract

The invention provides a flow control method, a computer program product, equipment and a storage medium, the flow control method is applied to electronic equipment, and comprises the following steps: obtaining current attribute information of service flow and current state information of a service board card; based on the current attribute information and the current state information, utilizing a strategy prediction model to predict and obtain a first target strategy; in the process of controlling the service board card to process the service flow according to the first target strategy, new attribute information and new state information are continuously obtained, and the attribute information of the service flow and the state information of the service board card are dynamically changed; obtaining result associated information corresponding to the first target strategy to adjust the strategy prediction model; determining a second target strategy by using the adjusted strategy prediction model based on the new attribute information and the new state information under the condition that the new service flow and / or the new service board card meet the adjustment condition; and controlling the service board card to process the service flow according to the second target strategy.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a flow control method, a computer program product, a device and a storage medium. Background Art

[0002] With the rapid development of information technology, distributed computing systems have been widely used in various fields. In distributed rack equipment, multiple service boards are usually equipped to jointly process a large amount of business traffic, among which these business traffic has diversity and uncertainty, including but not limited to real-time data processing, file transfer, multimedia streaming, etc. Therefore, it is necessary to accurately allocate business traffic to different business boards.

[0003] At present, the solution for adjusting the business traffic flow is: when packet loss occurs, the business traffic flow is adjusted by the load size and idle state of the business card. However, this solution cannot effectively avoid packet loss, and the flow diversion is not accurate. Summary of the invention

[0004] In view of this, an object of the embodiments of the present application is to provide a flow control method, a computer program product, a device and a storage medium.

[0005] On the one hand, an embodiment of the present application provides a flow control method, which is applied to an electronic device, including:

[0006] Acquire current attribute information of service traffic and current status information of a service board, wherein the electronic device includes a plurality of service boards, and the service boards generate service traffic when processing services;

[0007] Based on the current attribute information and the current state information, a first target strategy is predicted using a strategy prediction model;

[0008] In the process of controlling the service board to process the service traffic according to the first target strategy, new attribute information and new status information are continuously obtained, and the attribute information of the service traffic and the status information of the service board are dynamically changed;

[0009] Obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model;

[0010] In the case where the new service flow and / or the new service board meet the adjustment conditions, based on the new attribute information and the new state information, the second target policy is determined by using the adjusted policy prediction model;

[0011] According to the second target strategy, the service board is controlled to process the service traffic.

[0012] In a possible implementation, the step of predicting the first target strategy using a strategy prediction model based on the current attribute information and the current state information includes:

[0013] Calculating the current attribute information and the current state information using a strategy prediction model to predict a plurality of first candidate strategies;

[0014] In the case of obtaining a plurality of the first candidate strategies, screening out a first target strategy from the plurality of the first candidate strategies; or

[0015] In the case where the current attribute information and the current state information trigger a screening condition, a first target policy is screened out from a plurality of first candidate policies.

[0016] In a possible implementation, the step of predicting the first target strategy using a strategy prediction model based on the current attribute information and the current state information includes:

[0017] Calculating the current attribute information and the current state information using a strategy prediction model to predict a change trend of the service flow;

[0018] Based on the change trend and the current state information, a target policy template is selected from a plurality of candidate policy templates, wherein the candidate policy templates are pre-generated and stored in the policy prediction model;

[0019] Based on the target policy template, the first target policy is generated.

[0020] In a possible implementation manner, obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model includes:

[0021] After controlling the service board to process the service traffic according to the first target strategy, obtaining a processing result of the service board processing the service traffic;

[0022] Displaying the current attribute information, the current state information, a plurality of the first candidate strategies, and the first target strategy;

[0023] Collecting operations performed by users to obtain user feedback information, wherein the feedback information at least includes user evaluation and user modification information;

[0024] The strategy prediction model is adjusted based on the current attribute information, the current state information, the first candidate strategy, the first target strategy, the processing result, and the feedback information.

[0025] In a possible implementation manner, when the new service flow and / or the new service board meets the adjustment condition, determining the second target policy by using the adjusted policy prediction model based on the new attribute information and the new state information includes:

[0026] When the new service flow and / or the new service board meet the adjustment conditions, the new attribute information and the new status information are calculated by using the adjusted policy prediction model to obtain a plurality of second candidate policies;

[0027] A second target strategy is selected from the plurality of second candidate strategies.

[0028] In a possible implementation manner, the adjustment conditions include at least one or more of the following:

[0029] The new attribute information and the new status information represent the load imbalance of the service board;

[0030] According to the first target strategy, the time duration for the service board to process the service traffic reaches a preset time duration;

[0031] Receive new business requirements;

[0032] The network status of the network environment in which the service board is located changes.

[0033] In a possible implementation, the determining the second target strategy based on the new attribute information and the new state information using the adjusted strategy prediction model includes:

[0034] Obtaining one or more of the priority level of the service traffic, the new service demand, and the network status of the network environment in which the service board is located;

[0035] Using the adjusted strategy prediction model, calculating the new attribute information, the new state information, and one or more of the priority level, the new service demand, and the network state to obtain a plurality of the second candidate strategies;

[0036] The second target strategy is selected from a plurality of second candidate strategies.

[0037] On the other hand, an embodiment of the present application provides a computer program product storing a computer program, which implements the steps of the above method when executed by a processor.

[0038] On the other hand, an embodiment of the present application proposes a computer device, which includes at least a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps of the above method when executing the computer program in the memory.

[0039] On the other hand, an embodiment of the present application proposes a storage medium, which carries one or more computer programs, and the one or more computer programs implement the steps of the above method when executed by a processor.

[0040] The embodiment of the present application predicts and obtains the first target strategy based on the current attribute information of the business traffic and the current status information of the business board. It can predict the strategy for controlling the business board to process the business traffic before packet loss occurs, effectively avoiding the problem of business being affected by packet loss; and after controlling the business board to process the business traffic according to the first target strategy, the result association information corresponding to the first target strategy is obtained to adjust the strategy prediction model, so as to determine the second target strategy using the adjusted strategy prediction model, thereby achieving the purpose of dynamically adjusting the strategy with high real-time performance; furthermore, the embodiment of the present application determines strategies such as the first target strategy and the second target strategy based on attribute information and status information, which can control the business board to process the business traffic more comprehensively and accurately, thereby ensuring the system stability of the electronic equipment.

[0041] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0043] Figure 1 A flow chart of a flow control method provided by the present application is shown;

[0044] Figure 2 A flow chart showing an adjustment strategy prediction model in a flow control method provided by the present application is shown;

[0045] Figure 3 A schematic structural diagram of a flow control device provided by the present application is shown;

[0046] Figure 4 A schematic diagram of the structure of a computer device provided by the present application is shown. DETAILED DESCRIPTION

[0047] Various aspects and features of the present application are described herein with reference to the accompanying drawings.

[0048] It should be understood that various modifications may be made to the embodiments of the present application. Therefore, the above description should not be considered as limiting, but only as an example of an embodiment. Other modifications within the scope and spirit of the present application will occur to those skilled in the art.

[0049] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0050] These and other characteristics of the present application will become apparent from the following description of a preferred form of embodiment given as a non-limiting example with reference to the accompanying drawings.

[0051] It should also be understood that, although the present application has been described with reference to some specific examples, those skilled in the art will be able to realize many other equivalent forms of the present application that have the features described in the claims and are therefore within the scope of protection defined thereby.

[0052] The above and other aspects, features and advantages of the present application will become more apparent in view of the following detailed description when taken in conjunction with the accompanying drawings.

[0053] Specific embodiments of the present application are described hereinafter with reference to the accompanying drawings; however, it should be understood that the embodiments applied for are merely examples of the present application, which may be implemented in a variety of ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that obscure the present application. Therefore, the specific structural and functional details applied for herein are not intended to be limiting, but merely serve as a basis and representative basis for the claims to teach those skilled in the art to use the present application in a variety of ways with substantially any suitable detailed structure.

[0054] This specification may use the phrases "in one embodiment," "in another embodiment," "in yet another embodiment," or "in other embodiments," all of which may refer to one or more of the same or different embodiments according to the present application.

[0055] Firstly, in order to facilitate the understanding of the present application, a flow control method for a multi-service board provided by the present application is first introduced in detail.

[0056] like Figure 1 As shown, it is a flow chart of the flow control method of the multi-service board provided in the embodiment of the present application, according to Figure 1 The method steps shown can not only effectively prevent services from being affected by packet loss, but also achieve the purpose of dynamically adjusting strategies, and can more comprehensively and accurately control the service board to process service traffic, thereby ensuring the system stability of the electronic device. Specific steps include S101-S106.

[0057] S101, obtaining current attribute information of service traffic and current status information of service boards, wherein the electronic device includes a plurality of service boards, and service traffic is generated when the service boards process services.

[0058] Optionally, the electronic device may include multiple service boards, such as a graphics card, a sound card, a network card, etc. The service board is used to process services. Specifically, when processing services, the service board receives and sends data to generate service traffic, and different types of services correspond to different service boards. The main control board or central processor of the electronic device can obtain the service traffic and status information generated on each service board in real time.

[0059] After obtaining each service flow, the service flow is analyzed to obtain the current attribute information of the service flow, and the current attribute information of the service flow includes one or more of the following: size, type, source and destination addresses, port number, time series characteristics of the service flow, and periodicity, etc. In addition, the current status information of the service board includes one or more of the following: central processing unit (CPU) utilization rate, total memory space, unused memory space, network bandwidth utilization rate, number of I / O operations, task queue length, variance of response time, etc.

[0060] In the embodiment of the present application, the CPU utilization rate is measured as a percentage, which is used to characterize the load size of the business board and evaluate the performance and efficiency of the business board. For example, the higher the CPU utilization rate, the greater the load of the business board, and its performance and efficiency may be lower. The network bandwidth utilization rate is also measured as a percentage, which is used to characterize the proportion of the bandwidth used by the business traffic on the business board to the total bandwidth. The number of I / O operations is used to characterize the number of transmissions of the business traffic on the business board. The task queue length is used to characterize the length of the business traffic with transmission on the business board. The variance of the response time is used to characterize the response performance of the business board.

[0061] Here, the current attribute information is obtained for the service traffic generated for the first time, and the current state information is obtained when the service board processes the service for the first time.

[0062] S102: Based on the current attribute information and the current state information, a first target strategy is predicted using a strategy prediction model.

[0063] After obtaining the current attribute information and the current state information, a first target strategy is predicted using a strategy prediction model based on the current attribute information and the current state information, wherein the strategy prediction model is pre-trained.

[0064] Optionally, when the first target strategy is predicted based on the current attribute information and the current state information, any one of the following two methods may be used:

[0065] The first

[0066] When the first target strategy is predicted, the strategy prediction model is used to first calculate the current attribute information and the current state information to predict a plurality of first candidate strategies.

[0067] Among them, the strategy prediction model includes artificial intelligence traffic prediction technology and multi-source data fusion technology. Artificial intelligence traffic prediction technology is used to predict the changing trend of business traffic, and multi-source data fusion technology is used to obtain the current attribute information of business traffic from different channels. Therefore, the strategy predicted by the strategy prediction model, such as the first candidate strategy, is more comprehensive and accurate. In addition, by predicting the changing trend of business traffic and then predicting the first candidate strategy, that is, before the packet loss occurs, the strategy for controlling the business board to process business traffic, that is, the first candidate strategy, is predicted, which effectively avoids the problem of business being affected by packet loss.

[0068] In addition, the strategy prediction model also includes an evaluation technology, which is used to calculate the current state information of the business board to obtain an evaluation result of the business board, and the evaluation result at least includes the load level of the business board, and the load level can characterize the current load situation of the business board and the ability to undertake new loads, etc. The embodiment of the present application uses multiple dimensions of state information, namely CPU usage, total memory space, unused memory space, network bandwidth usage, I / O operation times, task queue length, response time variance, etc., to evaluate the load level of the business board, thereby ensuring the accuracy of the evaluation result.

[0069] Furthermore, after obtaining the current attribute information and the current state information in real time, the current attribute information and the current state information can be calculated in real time, and then the first candidate strategy can be obtained in real time, that is, the first candidate strategy has high real-time performance and accuracy.

[0070] In the case of obtaining multiple first candidate strategies, a first target strategy is screened out from the multiple first candidate strategies. Optionally, each first candidate strategy carries an identifier, which is a score of the first candidate strategy, and based on the identifier carried by each first candidate strategy, a first target strategy is screened out from the multiple first candidate strategies, wherein the score of the first target strategy is the highest score among the multiple first candidate strategies.

[0071] In another example, after a plurality of first candidate strategies are predicted by the strategy prediction model, and when the current attribute information and the current state information trigger the screening condition, the first target strategy is screened out from the plurality of first candidate strategies. The screening condition can be set as the CPU usage rate reaches the maximum usage rate allowed by the business card, the task queue length is greater than the preset length, etc., which is not specifically limited in the embodiments of the present application.

[0072] Optionally, in the embodiment of the present application, three models may be set, such as the first model, the second model and the first model, and the first model may include artificial intelligence traffic prediction technology and multi-source data fusion technology, the second model may include evaluation technology, and the third model may include strategy formulation technology. That is, the first model is used to calculate the current attribute information to obtain the change trend of the business traffic; the second model is used to calculate the current state information to obtain the load level of the business board; the third model is used to calculate the change trend of the business traffic and the load level of the business board to obtain the first candidate strategy.

[0073] Compared with setting up three models, setting up a policy prediction model can reduce the model construction cost and training cost; compared with setting up a policy prediction model, when setting up three models, the number of parameters and the amount of calculation of each model are smaller than the policy prediction model, and thus setting up three models can reduce the requirements of hardware performance for calculations.

[0074] It is worth noting that when constructing a model such as a strategy prediction model, the embodiment of the present application may adopt a model structure such as a deep neural network (DNN) or a convolutional neural network (CNN). Optionally, the input layer may be configured to receive a feature vector of a variety of data such as current attribute information and current state information after preprocessing; the middle layer may be configured to have multiple hidden layers, and nonlinear transformations may be performed through the connection of neurons and activation functions to achieve the calculation process; the output layer may be configured to output strategies, etc.

[0075] In the embodiment of the present application, when training the constructed model, such as training the strategy prediction model, a large amount of historical data can be used for training, and the historical data includes attribute information, state information, and corresponding strategies. Optionally, the historical data is divided into a training set, a validation set, and a test set. The training set is used to update the parameters of the model, the validation set is used to adjust the hyperparameters of the model and evaluate the performance of the model, and the test set is used to finally evaluate the generalization ability of the model.

[0076] In addition, when training the policy prediction model, the loss function is used to optimize the policy prediction model, wherein the loss function can select mean square error (MSE) or cross entropy, etc. The error between the predicted policy and the actual policy is minimized by the loss function, and then the parameters of the model are continuously adjusted through the back propagation algorithm and the optimization algorithm (such as stochastic gradient descent) to gradually reduce the loss function, so as to complete the training of the policy prediction model, and obtain the final policy prediction model.

[0077] As another example, when screening out the first target strategy from multiple first candidate strategies, the selection model can also be used to screen out the first target strategy. The selection model can adopt a machine learning algorithm, such as a decision tree, a random forest, or a support vector machine, so that the selection model evaluates different first candidate strategies to screen out the first target strategy.

[0078] In the embodiments of the present application, the training steps and processes of each model are not elaborated in detail, and the training steps and processes of the model can refer to the common training steps and processes of the current models.

[0079] The second

[0080] When the first target strategy is predicted based on the current attribute information and the current state information, the strategy prediction model is used to first calculate the current attribute information and the current state information to predict the change trend of the business traffic. Afterwards, based on the change trend and the current state information, the target strategy template is selected from multiple candidate strategy templates, and the candidate strategy templates are pre-generated and stored in the strategy prediction model. Among them, the candidate strategy template is generated based on historical data, and the historical data includes the historical attribute information of the business traffic, the historical state information of the business board, and the corresponding historical strategy, etc., and there are multiple candidate strategy templates, and different businesses and different attribute information can correspond to different candidate strategy templates, etc.

[0081] For example, for high-priority services, setting the corresponding candidate policy template includes: allocating high-priority service traffic to service boards with lighter loads and stronger performance, and at the same time, setting a higher allocation ratio; for low-priority services, setting the corresponding candidate policy template includes allocating low-priority service traffic to service boards with medium loads and medium performance.

[0082] For example, for large-scale business traffic, setting the corresponding candidate policy template includes: allocating the traffic in proportion according to the load level of the business board and the business processing capability.

[0083] For example, for a low-load service card, setting the corresponding candidate policy template includes: giving priority to selecting an idle service card for allocation.

[0084] Of course, the number and specific contents of the candidate policy templates can be adjusted and set according to actual needs, and the embodiments of the present application do not make specific limitations on this.

[0085] After selecting a target policy template from multiple candidate policy templates, a first target policy is generated based on the target policy template. For example, the target policy template is adjusted and supplemented based on current attribute information and current state information to obtain the first target policy.

[0086] It should be noted that the strategy prediction model in the first method and the strategy prediction model in the second method can adopt one or more of the same technical means, but the strategy prediction model in the first method and the strategy prediction model in the second method have different execution steps in a different order, and some parameters are also different.

[0087] S103, in the process of controlling the service board to process the service traffic according to the first target strategy, new attribute information and new status information are continuously obtained, and the attribute information of the service traffic and the status information of the service board are dynamically changed.

[0088] After the first target strategy is determined, the service board is controlled to process the service traffic according to the first target strategy, that is, the service traffic is diverted and processed.

[0089] As an example, during the process of the service card processing services, the attribute information of the service flow and the status information of the service card change dynamically. Based on this, in the process of the first target strategy controlling the service card to process the service flow, new attribute information and new status information are continuously obtained to ensure the real-time and accuracy of the current attribute information of the service flow and the current status information of the service card.

[0090] S104, obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model.

[0091] During the process of controlling the business card to process business traffic according to the first target strategy or after completing business processing, result association information corresponding to the first target strategy is obtained, wherein the result association information at least includes the processing result of the business card processing business traffic and user feedback information.

[0092] After obtaining the result association information, the strategy prediction model is adjusted based on the result association information. Figure 2 The flowchart shown is used to adjust the strategy prediction model, which specifically includes S201-S204.

[0093] S201, after controlling the service board to process service traffic according to the first target strategy, obtain the processing result of the service board processing the service traffic.

[0094] S202, display current attribute information, current state information, multiple first candidate strategies and a first target strategy.

[0095] S203, collecting operations performed by the user to obtain user feedback information, where the feedback information at least includes user evaluation and user modification information.

[0096] S204, adjusting the strategy prediction model based on the current attribute information, the current state information, the first candidate strategy, the first target strategy, the processing result and the feedback information.

[0097] When the service board processes the service traffic, the packet loss rate, load level, etc. in the processing process are recorded and stored. Based on this, after the service board is controlled to process the service traffic according to the first target strategy, the processing result of the service board processing the service traffic can be obtained. The processing result at least includes the packet loss rate, load level, etc.

[0098] Optionally, in order to avoid affecting the first target policy controlling the business board to process business traffic, after controlling the business board to process business traffic according to the first target policy, the display device on the electronic device is controlled to display current attribute information, current status information, multiple first candidate policies and the first target policy.

[0099] Of course, while controlling the business board to process business traffic according to the first target strategy, it is also possible to control the display device on the electronic device to display current attribute information, current status information, multiple first candidate strategies and the first target strategy, so that the user can know the currently executed first target policy in a timely manner and determine whether the first target policy needs to be adjusted, thereby ensuring the accuracy of controlling the business board to process business traffic.

[0100] Afterwards, the operations performed by the user are collected to obtain the user's feedback information. Optionally, the operations performed by the user may be operations performed on the first target strategy, operations performed on the first candidate strategy, or operations performed on the first candidate strategy and the first target strategy. Accordingly, the feedback information includes at least user evaluation and user-modified information, the user evaluation is the user's evaluation of the first candidate strategy and / or the first target strategy, which may be expressed as a score, and the user-modified information is the modification made by the user or the modification suggestion proposed by the user on the first target strategy.

[0101] Furthermore, after obtaining the processing results of the business board processing the business traffic and the user's feedback information, the strategy prediction model can be adjusted based on the current attribute information, the current state information, the first candidate strategy, the first target strategy, the processing results and the feedback information. Optionally, based on the processing results and the feedback information, the response time, throughput and user satisfaction with the target strategy of the business board are extracted (the satisfaction can be determined based on the score and whether to modify it), and then the response time, throughput and satisfaction are used as reward signals for the strategy prediction model. For example, when the response time, throughput and satisfaction represent the improvement of the system performance of the electronic device (the overall performance of multiple business boards), the strategy prediction model is given a positive reward; when the response time, throughput and satisfaction represent the decrease of the system performance of the electronic device, the strategy prediction model is given a negative reward, so as to achieve the purpose of optimizing and adjusting the strategy prediction model.

[0102] Optionally, the strategy prediction model can be optimized and updated using an online learning algorithm, such as an online version of stochastic gradient descent or an incremental deep learning algorithm, etc. In the embodiment of the present application, the strategy prediction model can be automatically optimized and adjusted without human intervention, and has high adaptability.

[0103] S105, when the new service flow and / or the new service board meet the adjustment conditions, based on the new attribute information and the new state information, the second target policy is determined by using the adjusted policy prediction model.

[0104] Optionally, the adjustment conditions include at least one or more of the following:

[0105] The first

[0106] The new attribute information and the new status information represent the load imbalance of the service board. For example, after calculating the new status information, the load levels of service board A, service board B, and service board C are obtained, and the load levels of service board A and service board B are both high, and service board C is idle, and there is no business traffic, i.e., load, that needs to be processed. This scenario is the load imbalance of the service board, and the adjustment condition is triggered at this time, and the step of updating the first target strategy is executed.

[0107] The second

[0108] The duration of the service board processing service traffic according to the first target strategy reaches a preset duration. In a specific implementation, in order to save monitoring resources, a preset duration can be set in advance, and the first target strategy can be periodically updated with the preset duration as a period. Therefore, the timing starts while the service board is controlled to process service traffic according to the first target strategy, and when the timing duration reaches the preset duration, the step of updating the first target strategy is executed.

[0109] The third

[0110] Receive new business requirements. Optionally, the business board can monitor its own business processing status in real time, such as completed business, business being processed, and newly added business. When monitoring the existence of newly added business, that is, receiving new business requirements, the step of updating the first target strategy is executed.

[0111] The fourth

[0112] The network status of the network environment where the service board is located changes. Among them, a certain bandwidth is required when processing services. Based on this, the service board needs to monitor the network status of the network environment in real time. For example, changes in the location of network equipment or changes in network time will cause bandwidth reduction. When the bandwidth reduction is detected, the step of updating the first target strategy is executed.

[0113] As an example, after determining that the new service flow and / or the new service board meet the adjustment conditions, the second target policy is determined based on the new attribute information and the new state information using the adjusted policy prediction model. Optionally, the policy prediction model is used to calculate the new attribute information and the new state information to obtain multiple second candidate policies, and the second target policy is selected from the multiple second candidate policies.

[0114] Optionally, when the new business traffic and / or the new business board meet the adjustment conditions, when the adjusted policy prediction model is used to predict multiple second candidate strategies, the adjusted policy prediction model can also be used to calculate one or more of the new attribute information, new status information, priority level, new business requirements, and network status to obtain multiple second candidate strategies, so that the second candidate strategies are more accurate, and thus the second target strategy is more accurate.

[0115] When new business traffic and / or new business boards meet the adjustment conditions, the adjusted strategy prediction model is used to promptly determine the second target strategy to control the business boards to process business traffic according to the second target strategy. This allows for more rapid determination of the optimal strategy to control the business boards to process business traffic when business traffic and network environment fluctuate, thereby effectively ensuring the stability of the electronic equipment system.

[0116] Moreover, when determining the second candidate strategy, the adjusted strategy prediction model can combine processing results, feedback information, priority levels, new business requirements, and one or more of the network status, so that the adjusted strategy prediction model, that is, the traffic control method of the present application, can be applicable to different scenarios with high flexibility.

[0117] S106, controlling the service board to process service traffic according to the second target strategy.

[0118] After the first target policy is updated with the second target policy, the service board is controlled to process the service traffic according to the second target policy, so that the load of each service board is more balanced and the service traffic can be accurately distributed.

[0119] The embodiment of the present application predicts and obtains the first target strategy based on the current attribute information of the business traffic and the current status information of the business board. It can predict the strategy for controlling the business board to process the business traffic before packet loss occurs, effectively avoiding the problem of business being affected by packet loss; and after controlling the business board to process the business traffic according to the first target strategy, the result association information corresponding to the first target strategy is obtained to adjust the strategy prediction model, so as to determine the second target strategy using the adjusted strategy prediction model, thereby achieving the purpose of dynamically adjusting the strategy with high real-time performance; furthermore, the embodiment of the present application determines strategies such as the first target strategy and the second target strategy based on attribute information and status information, which can control the business board to process the business traffic more comprehensively and accurately, thereby ensuring the system stability of the electronic equipment.

[0120] It is worth noting that before using the strategy prediction model to predict multiple first candidate strategies, one or more of the current attribute information, current status information, processing results, feedback information, priority levels, new business requirements, and network status are preprocessed to ensure the accuracy of the predicted first candidate strategy or second candidate strategy.

[0121] Among them, preprocessing can include data cleaning and normalization. Data cleaning is used to remove outliers and noise data, such as obviously unreasonable business traffic values ​​or load indicators. Specifically, outliers can be identified and eliminated by setting thresholds or statistical methods; normalization is used to unify different types of data into the same dimension and range, so that the strategy prediction model can analyze and calculate different types of data. For example, the min-max normalization method can be used to map the data to the [0, 1] interval, etc. Of course, other preprocessing methods can also be used, and the embodiments of this application will not be listed one by one.

[0122] On the second aspect, an embodiment of the present application also provides a flow control device corresponding to a flow control method. Since the principle of solving the problem by the flow control device in the present application is similar to the flow control method of the multi-service board mentioned above in the present application, the implementation of the electronic device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0123] Figure 3 A schematic diagram of a flow control device provided in an embodiment of the present application is shown, which specifically includes:

[0124] An acquisition module 301 is configured to acquire current attribute information of service traffic and current status information of a service board, wherein the electronic device includes a plurality of service boards, and the service boards generate service traffic when processing services;

[0125] A prediction module 302 is configured to predict a first target strategy using a strategy prediction model based on current attribute information and current state information;

[0126] An acquisition module 303 is configured to continuously acquire new attribute information and new status information in the process of controlling the service board to process the service traffic according to the first target strategy, wherein the attribute information of the service traffic and the status information of the service board are dynamically changed;

[0127] An adjustment module 304 is configured to obtain result association information corresponding to the first target strategy to adjust the strategy prediction model;

[0128] A determination module 305, configured to determine a second target strategy using an adjusted strategy prediction model based on the new attribute information and the new state information when the new service flow and / or the new service board meet the adjustment conditions;

[0129] The control module 306 is configured to control the service board to process the service traffic according to the second target policy.

[0130] In yet another embodiment, the prediction module 302 is specifically configured as follows:

[0131] Calculating the current attribute information and the current state information using a strategy prediction model to predict a plurality of the first candidate strategies;

[0132] In the case of obtaining a plurality of the first candidate strategies, screening out a first target strategy from the plurality of the first candidate strategies; or

[0133] In the case where the current attribute information and the current state information trigger a screening condition, a first target policy is screened out from a plurality of first candidate policies.

[0134] In yet another embodiment, the prediction module 302 is further configured to:

[0135] The method of predicting a first target strategy based on the current attribute information and the current state information by using a strategy prediction model includes:

[0136] Calculating the current attribute information and the current state information using a strategy prediction model to predict a change trend of the service flow;

[0137] Based on the change trend and the current state information, a target policy template is selected from a plurality of candidate policy templates, wherein the candidate policy templates are pre-generated and stored in the policy prediction model;

[0138] Based on the target policy template, the first target policy is generated.

[0139] In yet another embodiment, the adjustment module 304 is specifically configured as follows:

[0140] After controlling the service board to process the service traffic according to the first target strategy, obtaining a processing result of the service board processing the service traffic;

[0141] Displaying the current attribute information, the current state information, a plurality of the first candidate strategies, and the first target strategy;

[0142] Collecting operations performed by users to obtain user feedback information, wherein the feedback information at least includes user evaluation and user modification information;

[0143] The strategy prediction model is adjusted based on the current attribute information, the current state information, the first candidate strategy, the first target strategy, the processing result, and the feedback information.

[0144] In yet another embodiment, the determination module 305 is specifically configured as follows:

[0145] When the new service flow and / or the new service board meet the adjustment conditions, the new attribute information and the new status information are calculated by using the adjusted policy prediction model to obtain a plurality of the second candidate policies;

[0146] A second target strategy is selected from the plurality of second candidate strategies.

[0147] In another embodiment, the adjustment conditions include at least one or more of the following:

[0148] The new attribute information and the new status information represent the load imbalance of the service board;

[0149] According to the first target strategy, the time duration for the service board to process the service traffic reaches a preset time duration;

[0150] Receive new business requirements;

[0151] The network status of the network environment in which the service board is located changes.

[0152] In yet another embodiment, the determining module 305 is further configured to:

[0153] Obtaining one or more of the priority level of the service traffic, the new service demand, and the network status of the network environment in which the service board is located;

[0154] Using the adjusted strategy prediction model, calculating the new attribute information, the new state information, and one or more of the priority level, the new service demand, and the network state to obtain a plurality of the second candidate strategies;

[0155] The second target strategy is selected from a plurality of second candidate strategies.

[0156] The embodiment of the present application predicts and obtains the first target strategy based on the current attribute information of the business traffic and the current status information of the business board. It can predict the strategy for controlling the business board to process the business traffic before packet loss occurs, effectively avoiding the problem of business being affected by packet loss; and after controlling the business board to process the business traffic according to the first target strategy, the result association information corresponding to the first target strategy is obtained to adjust the strategy prediction model, so as to determine the second target strategy using the adjusted strategy prediction model, thereby achieving the purpose of dynamically adjusting the strategy with high real-time performance; furthermore, the embodiment of the present application determines strategies such as the first target strategy and the second target strategy based on attribute information and status information, which can control the business board to process the business traffic more comprehensively and accurately, thereby ensuring the system stability of the electronic equipment.

[0157] In a third aspect, an embodiment of the present application further provides a computer program product, which is a computer-readable medium storing a computer program. When the computer program is executed by a processor, the method provided by any embodiment of the present invention is implemented, including the following steps S11 to S16:

[0158] S11, obtaining current attribute information of service traffic and current status information of a service board, wherein the electronic device includes a plurality of service boards, and service traffic is generated when the service boards process services;

[0159] S12, based on the current attribute information and the current state information, using the strategy prediction model to predict and obtain a first target strategy;

[0160] S13, in the process of controlling the service board to process the service traffic according to the first target strategy, continuously obtaining new attribute information and new status information, wherein the attribute information of the service traffic and the status information of the service board are dynamically changed;

[0161] S14, obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model;

[0162] S15, when the new service flow and / or the new service board meet the adjustment conditions, based on the new attribute information and the new state information, determine the second target strategy using the adjusted strategy prediction model;

[0163] S16: Control the service board to process the service traffic according to the second target strategy.

[0164] In a fourth aspect, the present application also provides a computer device, the structural diagram of which can be as follows: Figure 4 As shown, at least a memory 401 and a processor 402 are included. The memory 401 stores a computer program. The processor 402 implements the method provided by any embodiment of the present invention when executing the computer program on the memory 401. Exemplarily, the computer program steps of the computer device are as follows S21 to S26:

[0165] S21, obtaining current attribute information of the service traffic and current status information of the service board, wherein the electronic device includes a plurality of service boards, and the service boards generate service traffic when processing services;

[0166] S22, based on the current attribute information and the current state information, using the strategy prediction model to predict and obtain a first target strategy;

[0167] S23, in the process of controlling the service board to process the service traffic according to the first target strategy, continuously obtaining new attribute information and new status information, wherein the attribute information of the service traffic and the status information of the service board are dynamically changed;

[0168] S24, obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model;

[0169] S25, when the new service flow and / or the new service board meet the adjustment conditions, based on the new attribute information and the new state information, determine the second target strategy using the adjusted strategy prediction model;

[0170] S26: Control the service board to process the service traffic according to the second target policy.

[0171] In a fifth aspect, an embodiment of the present application further provides a storage medium, wherein the storage medium carries one or more computer programs, and when the one or more computer programs are executed by a processor, the method provided by any embodiment of the present invention is implemented, including the following steps S31 to S36:

[0172] S31, obtaining current attribute information of service traffic and current status information of a service board, wherein the electronic device includes a plurality of service boards, and the service boards generate service traffic when processing services;

[0173] S32, based on the current attribute information and the current state information, using the strategy prediction model to predict and obtain a first target strategy;

[0174] S33, in the process of controlling the service board to process the service traffic according to the first target strategy, continuously obtaining new attribute information and new status information, wherein the attribute information of the service traffic and the status information of the service board are dynamically changed;

[0175] S34, obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model;

[0176] S35, when the new service flow and / or the new service board meet the adjustment conditions, based on the new attribute information and the new state information, determine the second target strategy using the adjusted strategy prediction model;

[0177] S36: Control the service board to process the service traffic according to the second target policy.

[0178] The embodiment of the present application predicts and obtains the first target strategy based on the current attribute information of the business traffic and the current status information of the business board. It can predict the strategy for controlling the business board to process the business traffic before packet loss occurs, effectively avoiding the problem of business being affected by packet loss; and after controlling the business board to process the business traffic according to the first target strategy, the result association information corresponding to the first target strategy is obtained to adjust the strategy prediction model, so as to determine the second target strategy using the adjusted strategy prediction model, thereby achieving the purpose of dynamically adjusting the strategy with high real-time performance; furthermore, the embodiment of the present application determines strategies such as the first target strategy and the second target strategy based on attribute information and status information, which can control the business board to process the business traffic more comprehensively and accurately, thereby ensuring the system stability of the electronic equipment.

[0179] Optionally, in this embodiment, the above-mentioned computer program product and storage medium may include, but are not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk. Optionally, in this embodiment, the processor executes the method steps recorded in the above-mentioned embodiment according to the program code stored in the computer program product. Optionally, the specific examples in this embodiment can refer to the examples described in the above-mentioned embodiment and the optional implementation mode, and this embodiment will not be repeated here. Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented with a general computing device, which can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented with executable program codes of computing devices, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from here, or they can be made into individual integrated circuit modules respectively, or multiple modules or steps in them can be made into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0180] In addition, although exemplary embodiments have been described herein, the scope includes any and all embodiments based on the present application with equivalent elements, modifications, omissions, combinations (e.g., various embodiments intersecting schemes), adaptations or changes. The elements in the claims will be interpreted broadly based on the language adopted in the claims, and are not limited to the examples described in this specification or during the implementation of the application, and the examples will be interpreted as non-exclusive. Therefore, this specification and examples are intended to be considered as examples only, and the true scope and spirit are indicated by the following claims and the full scope of their equivalents.

[0181] The above description is intended to be illustrative rather than restrictive. For example, the above examples (or one or more of them) can be used in combination with each other. For example, those of ordinary skill in the art can use other embodiments when reading the above description. In addition, in the above-mentioned specific embodiments, various features can be grouped together to simplify the application. This should not be interpreted as an intention that a disclosed feature that does not require protection is necessary for any claim. On the contrary, the subject matter of the present application may be less than all the features of a specific disclosed embodiment. Thus, the following claims are incorporated into the specific embodiments as examples or embodiments, wherein each claim is independently used as a separate embodiment, and it is considered that these embodiments can be combined with each other in various combinations or arrangements. The scope of the present application should be determined with reference to the attached claims and the full scope of equivalent forms granted by these claims.

[0182] Multiple embodiments of the present application are described in detail above, but the present application is not limited to these specific embodiments. Those skilled in the art can make various variations and modifications to the embodiments based on the concept of the present application, and these variations and modifications should all fall within the scope of protection required by the present application.

Claims

1. A flow control method, characterized in that: Used in electronic equipment, including: Acquire current attribute information of service traffic and current status information of a service board, wherein the electronic device includes a plurality of service boards, and the service boards generate service traffic when processing services; Based on the current attribute information and the current state information, a first target strategy is predicted using a strategy prediction model; In the process of controlling the service board to process the service traffic according to the first target strategy, new attribute information and new status information are continuously obtained, and the attribute information of the service traffic and the status information of the service board are dynamically changed; Obtaining result association information corresponding to the first target strategy to adjust the strategy prediction model; In the case where the new service flow and / or the new service board meet the adjustment conditions, based on the new attribute information and the new state information, the second target policy is determined by using the adjusted policy prediction model; According to the second target strategy, the service board is controlled to process the service traffic.

2. The flow control method according to claim 1, characterized in that: The method of predicting a first target strategy based on the current attribute information and the current state information by using a strategy prediction model includes: Calculating the current attribute information and the current state information using a strategy prediction model to predict a plurality of first candidate strategies; In the case of obtaining a plurality of the first candidate strategies, screening out a first target strategy from the plurality of the first candidate strategies; or In the case where the current attribute information and the current state information trigger a screening condition, a first target policy is screened out from a plurality of first candidate policies.

3. The flow control method according to claim 1, characterized in that: The method of predicting a first target strategy based on the current attribute information and the current state information by using a strategy prediction model includes: Calculating the current attribute information and the current state information using a strategy prediction model to predict a change trend of the service flow; Based on the change trend and the current state information, a target policy template is selected from a plurality of candidate policy templates, wherein the candidate policy templates are pre-generated and stored in the policy prediction model; Based on the target policy template, the first target policy is generated.

4. The flow control method according to claim 2, characterized in that: The obtaining the result association information corresponding to the first target strategy to adjust the strategy prediction model includes: After controlling the service board to process the service traffic according to the first target strategy, obtaining a processing result of the service board processing the service traffic; Displaying the current attribute information, the current state information, a plurality of the first candidate strategies, and the first target strategy; Collecting operations performed by users to obtain user feedback information, wherein the feedback information at least includes user evaluation and user modification information; The strategy prediction model is adjusted based on the current attribute information, the current state information, the first candidate strategy, the first target strategy, the processing result, and the feedback information.

5. The flow control method according to claim 1, characterized in that: In the case where the new service flow and / or the new service board meet the adjustment condition, based on the new attribute information and the new state information, the second target policy is determined by using the adjusted policy prediction model, including: When the new service flow and / or the new service board meet the adjustment conditions, the new attribute information and the new status information are calculated by using the adjusted policy prediction model to obtain a plurality of second candidate policies; A second target strategy is selected from the plurality of second candidate strategies.

6. The flow control method according to claim 5, characterized in that: The adjustment conditions include at least one or more of the following: The new attribute information and the new status information represent the load imbalance of the service board; According to the first target strategy, the time duration for the service board to process the service traffic reaches a preset time duration; Receive new business requirements; The network status of the network environment in which the service board is located changes.

7. The flow control method according to claim 6, characterized in that: The determining the second target strategy based on the new attribute information and the new state information by using the adjusted strategy prediction model includes: Obtaining one or more of the priority level of the service traffic, the new service demand, and the network status of the network environment in which the service board is located; Using the adjusted strategy prediction model, calculating the new attribute information, the new state information, and one or more of the priority level, the new service demand, and the network state to obtain a plurality of the second candidate strategies; The second target strategy is selected from a plurality of second candidate strategies.

8. A computer program product storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

9. A computer device, comprising at least a memory and a processor, wherein a computer program is stored in the memory, wherein: The processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program on the memory.

10. A storage medium, characterized in that: The storage medium carries one or more computer programs, and when the one or more computer programs are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images