Interface flow control method, device, electronic device and storage medium

Abstract

The present application discloses an interface flow control method, device, electronic device, and storage medium, belonging to the field of electronic information technology. The interface flow control method includes obtaining request data, where the request data is data to be sent to the interface; inputting the request data into a threshold prediction model to obtain a threshold value for the number of requests sent to the interface; and performing flow control on the interface based on the threshold value for the number of requests. By adjusting the data input threshold using a model prediction method, this method achieves smooth threshold flow control, fully utilizes the resources of the request receiving host, and ensures maximum business processing while ensuring stable system operation.

Description

Technical Field

[0001] The present application relates to the field of electronic information technology, and in particular to an interface flow control method, device, electronic device and storage medium. Background Art

[0002] In the scenario of interaction between multiple business application systems, system A needs to send requests to system B through the interface. Since the load capacity of system B is limited, in order to prevent unexpected requests from putting too much pressure on the system and dragging down system B, the interface requests need to be limited. Limiting means that under limited resources, the service capacity of each interface per unit time is limited. If the number of accesses to the interface is not controlled, it will cause excessive pressure on the interface. In order to improve the high availability of the interface and business application system, it is necessary to control the number of interface requests per unit time.

[0003] In the existing technology, fixed thresholds are often used for flow control, which cannot fully utilize the resources of the request receiving host. At a certain moment, the host's resources are relatively idle and can process requests greater than the threshold, but resources are wasted due to flow control. Similarly, at a certain moment, the host's resources are relatively tight and can only process requests less than the threshold. This increases the risk of dragging down the host, thereby affecting business processing. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to provide an interface traffic control method, device, electronic device and storage medium to at least solve the problem that the existing interface traffic exceeds the system load and causes the host to crash.

[0005] The technical solution of this application is as follows:

[0006] According to the first aspect of an embodiment of the present application, an interface traffic control method is provided, which may include: obtaining request data, where the request data is data to be sent to the interface; inputting the request data into a threshold prediction model to obtain a threshold value for the number of requests sent to the interface; and performing traffic control on the interface based on the threshold value for the number of requests.

[0007] According to the second aspect of an embodiment of the present application, an interface flow control device is provided, which may include: a data acquisition module for obtaining request data, where the request data is data to be sent to the interface; a prediction module for inputting the request data into a threshold prediction model to obtain a threshold value for the number of requests sent to the interface; and a flow control module for performing flow control on the interface based on the threshold value for the number of requests.

[0008] According to a third aspect of an embodiment of the present application, an electronic device is provided, which may include:

[0009] processor;

[0010] a memory for storing processor-executable instructions;

[0011] The processor is configured to execute instructions to implement the information processing method as shown in any one of the embodiments of the first aspect.

[0012] According to the fourth aspect of the embodiments of the present application, a storage medium is provided. When the instructions in the storage medium are executed by the processor of an information processing device or server, the information processing device or server can implement the information processing method shown in any one of the embodiments of the first aspect.

[0013] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects:

[0014] The embodiment of the present application obtains request data, which is the data to be sent to the interface; inputs the request data into a threshold prediction model to obtain a threshold for the number of requests sent to the interface; and performs flow control on the interface based on the threshold for the number of requests. By adjusting the data input threshold using model prediction, this method achieves smooth threshold flow control, fully utilizing the resources of the request receiving host, and ensuring maximum service processing while ensuring stable system operation.

[0015] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application, and do not constitute an improper limitation on the present application.

[0017] Figure 1 is a flow chart illustrating a method for controlling interface flow according to an exemplary embodiment;

[0018] Figure 2 is a schematic diagram of an interface traffic intelligent control structure according to an exemplary embodiment;

[0019] Figure 3 is a schematic diagram of a model training process according to an exemplary embodiment;

[0020] Figure 4 is a comparison chart of the accuracy of different improved models according to an exemplary embodiment;

[0021] Figure 5 is a schematic diagram showing an execution process of a flow control module according to an exemplary embodiment;

[0022] Figure 6 is a prediction curve diagram shown according to an exemplary embodiment;

[0023] Figure 7 is a schematic structural diagram of an electronic device according to an exemplary embodiment;

[0024] Figure 8 The figure is a schematic diagram showing the hardware structure of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0025] In order to enable ordinary people in the art to better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0026] It should be noted that the terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0027] Figure 1 This is a flow chart of an embodiment of the interface flow control method provided by this application. Figure 1 As shown, the interface flow control method may include:

[0028] Step 100: Obtain request data, which is data to be sent to the interface;

[0029] Step 200: Input the request data into the threshold prediction model to obtain the threshold of the number of requests sent to the interface;

[0030] Step 300: Based on the request quantity threshold, flow control is performed on the interface.

[0031] The above embodiment obtains request data, which is the data to be sent to the interface; inputs the request data into a threshold prediction model to determine the threshold number of requests sent to the interface; and then performs flow control on the interface based on the threshold number of requests. Using model prediction to adjust the data input threshold, this approach provides smooth threshold flow control, fully utilizing the resources of the request receiving host, and ensuring maximum service processing while maintaining stable system operation.

[0032] In an embodiment of the present application, the request data may include basic data such as CPU, memory, I / O, number of threads, response time, etc.

[0033] In the embodiment of the present application, the threshold prediction model is constructed by the following method:

[0034] Step 210: Acquire target data;

[0035] Step 210: pre-process the target data to obtain comprehensive indicator data with labels;

[0036] Step 210: Use the labeled comprehensive indicator data to train and test the machine learning model to obtain a trained threshold prediction model.

[0037] In an embodiment of the present application, the target data includes: CPU usage, memory usage, I / O usage, number of threads, and response time.

[0038] In an embodiment of the present application, preprocessing the target data may include labeling the data and performing dimensionality reduction processing on the data set using PCA dimensionality reduction technology.

[0039] In the embodiment of the present application, the target data is preprocessed to obtain comprehensive indicator data with labels, including:

[0040] The target data is labeled to obtain labeled data;

[0041] The labeled data is processed for dimensionality reduction to obtain labeled comprehensive indicator data.

[0042] In the embodiment of the present application, the labeled data is subjected to dimensionality reduction processing to obtain labeled comprehensive indicator data, including:

[0043] Use principal component analysis technology to reduce the dimension of labeled data and obtain labeled comprehensive indicator data

[0044] In an embodiment of the present application, the machine learning model includes: an attention mechanism function;

[0045] The attention mechanism function is used to generate the attention allocation coefficient for the labeled comprehensive indicator data.

[0046] In an embodiment of the present application, the method further includes at least one of the following:

[0047] Use the update threshold of the forward propagation algorithm to predict the internal state of the model;

[0048] Use the back-propagation algorithm to train and update the threshold prediction model weights and biases;

[0049] The Adam optimization algorithm is used to minimize the loss function.

[0050] In an embodiment of the present application, step 300: Controlling interface traffic based on a request number threshold may specifically involve collecting real-time data using Logstash technology, transmitting it via Kafka to a threshold prediction model. The prediction model then loads a model weight file, predicts the incoming data, and outputs a prediction result. Traffic control is then performed based on the output prediction result, combined with the current number of interface requests.

[0051] Based on the same inventive concept, an embodiment of the present application further provides an interface flow control device, comprising:

[0052] The data acquisition module is used to obtain the request data, which is the data to be sent to the interface;

[0053] A prediction module is used to input the request data into the threshold prediction model to obtain the threshold number of requests sent to the interface;

[0054] The flow control module is used to control the flow of the interface based on the request quantity threshold.

[0055] The above embodiment obtains request data, which is the data to be sent to the interface; inputs the request data into a threshold prediction model to determine the threshold number of requests sent to the interface; and then performs flow control on the interface based on the threshold number of requests. Using model prediction to adjust the data input threshold, this approach provides smooth threshold flow control, fully utilizing the resources of the request receiving host, and ensuring maximum service processing while maintaining stable system operation.

[0056] In some embodiments of the present disclosure, the prediction module may include: a PCA dimensionality reduction module, a model training module, and a threshold prediction module;

[0057] Among them, the PCA dimension reduction module is used to reduce the dimension of the data set;

[0058] The model training module is used to pass the data after dimensionality reduction processing into the model for training, and generate a model weight file to be loaded into the threshold prediction module; real-time data is collected through Logstash technology and transmitted to the threshold prediction module via Kafka;

[0059] The threshold prediction module predicts the incoming data by loading the model weight file and outputs the result. The structure of the interface flow control device is as follows: Figure 2 shown.

[0060] The model training module mainly implements data collection and preprocessing, combines the improved three-layer LSTM neural network model, introduces the attention mechanism to assign different weights to parameter elements, trains historical data, generates a multi-factor prediction model, establishes a nonlinear relationship between multi-factor variables and dynamic thresholds, and enhances the fit. It includes the following steps, such as Figure 3 shown.

[0061] Step 1: Data Collection and Preprocessing. Collect data from the actual production system for the past three months on a minute-by-minute basis, including CPU usage, memory usage, and I / O usage of core production hosts. Operators will label the data and use the PCA algorithm to perform dimensionality reduction on the labeled data.

[0062] Step 2: Model Training. The reduced data is divided into a training set and a test set in an 8:2 ratio. The training set data serves as the input vector for training in a three-layer LSTM model. The attention mechanism, through the Attention function, generates attention allocation coefficients for parameters such as response time, I / O usage, memory usage, and CPU usage. The attention weight coefficients used are (0.6, 0.2, 0.1, and 0.1), respectively, reflecting the influence of each parameter on the predicted information. Through continuous training and tuning of model parameters, the test set data provides a basis for model prediction performance. The input vector dimension is (9, 4). The LSTM model effectively memorizes previous time series and overcomes the problems of limited parameters and low accuracy caused by the simple structure of the native LSTM model.

[0063] Step 3: Model Update. The LSTM network model updates its internal state using the forward propagation algorithm, uses the backpropagation algorithm to train and update its weights and biases, and uses the Adam optimization algorithm to minimize the loss function, so that the network output fits the real data as closely as possible.

[0064] The forward propagation algorithm updates the output gate value of the LSTM network model. The formula is as follows.

[0065] h t =σ(W0[h t-1 ,x t ]+b0)*tanh(C t )

[0066] Where W0 is the hyperparameter weight matrix, h t-1 is the output value of the LSTM network model at the t-1 sequence moment, x t is the training vector at the time of sequence t, b0 is the hyperparameter bias matrix, C t h t-1 Another output value of the node, σ is the activation function.

[0067] LSTM uses the backpropagation algorithm to train and update weights and biases. The loss function formula is as follows.

[0068]

[0069]

[0070] y t is the true value, is the predicted value, and L is the loss function value.

[0071] Gradient descent uses the Adam optimization algorithm to minimize the loss function. This algorithm not only stores the exponentially decaying average of the squared gradient, but also maintains the exponentially decaying average of the previous gradient. The formula is as follows.

[0072]

[0073]

[0074]

[0075] where θ t is the training parameter under the tth iteration, m t is the average value of the gradient at the first moment, v t is the non-central difference of the gradient at the second moment, β1, β2, ε, and η are all hyperparameters, η is the learning rate, which controls the update rate of the weight, and this paper takes the value of 0.0001, β1 is the exponential decay rate of the first-order moment estimate, and this paper takes the value of 0.9, β2 is the exponential decay rate of the second-order moment estimate, and this paper takes the value of 0.999, and ε is 10 -8 The above parameter values ​​are optimized by continuously tuning the model parameters based on historical data. In actual training, the Adam method converges quickly, achieves more effective learning results, and corrects issues common with other optimization methods (such as vanishing learning rate, slow convergence, and large loss function fluctuations).

[0076] This method conducted comparative experiments on the accuracy of native LSTM, three-layer LSTM, three-layer LSTM+attention mechanism, and four-layer LSTM models. Figure 4 As shown in the figure, the unmodified native LSTM model has lower accuracy than the other models; the three-layer LSTM model has lower accuracy than the three-layer LSTM + attention mechanism model; and the four-layer LSTM model has lower accuracy than the three-layer LSTM model. This shows that as the number of LSTM layers increases, its accuracy decreases. After comparing various improved models, the three-layer LSTM + attention mechanism model has the highest accuracy.

[0077] The flow control module mainly realizes the system interface flow prediction and control functions. It includes the following steps, such as Figure 5 shown.

[0078] Step 1: De-persist the model. The trained model is saved in the model library. When making predictions, the model's weights, biases, and other values ​​are loaded into memory for use in predictions.

[0079] Step 2: Use the Logstash data collection technology to collect real-time data from each host and transmit it to Kafka on a minute-by-minute basis. Then use the data preprocessing function to preprocess the data.

[0080] Step 3: The values ​​of the above two steps are passed as input to the indicator prediction module. The prediction module generates a prediction curve based on the information received. The area covered by the prediction curve is the flow control interval, such as Figure 6 shown.

[0081] in, is the predicted value, Y t is the flow control value. If the actual flow value y at time t t ≤Y t , then all are processed, if y t ≥Y t , then according to the maximum value Y t Process, process y at a later time (e.g. time t+1) t -Y t .

[0082] In terms of model structure, the above embodiment uses a three-layer LSTM model to train multi-factor variables, effectively memorizing the previous time series and solving the problems of few parameters and low accuracy caused by the single structure of the native LSTM model. It effectively learns the relationship between multiple factors such as the host CPU, memory, I / O, and solves the drawback of the linear model's insufficient fit to complex mathematical structures, and deeply explores the relationship between the dependent variable and the threshold. At the same time, an attention mechanism is introduced to assign weights to parameter elements. The attention mechanism will assign different amounts of attention to multiple parameter data, forming an attention weight coefficient for each parameter, reflecting the degree of influence of each parameter on the predicted value, improving the efficiency of model training and operation, obtaining more overall features of important time, and solving the problem of incomplete representation of the time features of the input data in the final output.

[0083] By continuously optimizing model parameters based on historical data, we ultimately used a (9,4)-dimensional vector, achieving a transition from single-factor variables to multi-factor variables and from linear to nonlinear fitting. The adjusted Adam optimizer corrected the learning rate vanishing, slow convergence, and large loss function fluctuations common in other optimization functions, resulting in faster convergence and more effective learning. Smooth threshold flow control was implemented to fully utilize the resources of the request receiving host, ensuring maximum service processing while maintaining stable system operation.

[0084] Optional, such as Figure 7 As shown, an embodiment of the present application also provides an electronic device 700, including a processor 701, a memory 702, and a program or instruction stored in the memory 702 and executable on the processor 701. When the program or instruction is executed by the processor 701, each process of the above-mentioned interface flow control method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0085] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0086] Figure 8 A schematic diagram of the hardware structure of an electronic device implementing an embodiment of the present application.

[0087] The electronic device 800 includes but is not limited to components such as a radio frequency unit 801 , a network module 802 , an audio output unit 803 , an input unit 804 , a sensor 805 , a display unit 806 , a user input unit 807 , an interface unit 808 , a memory 809 , and a processor 810 .

[0088] Those skilled in the art will understand that the electronic device 800 may also include a power source (such as a battery) to power each component, and the power source may be logically connected to the processor 880 through a power management system, thereby implementing functions such as charging, discharging, and power consumption management through the power management system. Figure 8 The electronic device structure shown in the figure does not constitute a limitation of the electronic device, and the electronic device may include more or fewer parts than shown in the figure.

[0089] It should be understood that in the embodiment of the present application, the input unit 804 may include a graphics processing unit (GPU) 8041 and a microphone 8042. The graphics processor 8041 processes image data of still images or videos obtained by an image capture device (such as a camera) in video capture mode or image capture mode. The display unit 806 may include a display panel 8061, which may be configured in the form of a liquid crystal display, an organic light-emitting diode, etc. The user input unit 807 includes a touch panel 8071 and other input devices 8072. The touch panel 8071 is also called a touch screen. The touch panel 8071 may include two parts: a touch detection device and a touch controller. Other input devices 8072 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, power keys, etc.), a trackball, a mouse, and a joystick, which will not be described in detail here. The memory 809 can be used to store software programs and various data, including but not limited to applications and operating systems. The processor 810 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and applications, etc., and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 810.

[0090] The electronic device in the aforementioned embodiment similarly obtains request data, which is the data to be sent to the interface; inputs the request data into a threshold prediction model to determine a threshold for the number of requests sent to the interface; and then performs flow control on the interface based on the threshold. Using model prediction to adjust the data input threshold allows for smooth threshold flow control, fully utilizing the resources of the request-receiving host and ensuring maximum service processing while maintaining stable system operation.

[0091] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by the processor, the various processes of the above-mentioned interface flow control method embodiment are implemented and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0092] The processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (ROM), random access memory (RAM), a magnetic disk, or an optical disk.

[0093] When executed, the program stored in the storage medium of the above embodiment similarly obtains request data, which is data to be sent to the interface; inputs the request data into a threshold prediction model to determine a threshold for the number of requests sent to the interface; and performs flow control on the interface based on the threshold for the number of requests. Using model prediction to adjust the data input threshold allows for smooth threshold flow control, fully utilizing the resources of the request-receiving host and ensuring maximum service processing while maintaining stable system operation.

[0094] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned interface flow control method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0095] The program executed by the chip in the above embodiment similarly obtains request data, which is the data to be sent to the interface; inputs the request data into a threshold prediction model to determine the threshold number of requests sent to the interface; and then controls the interface flow based on the threshold number of requests. Using model prediction to adjust the data input threshold, this method provides smooth threshold flow control, fully utilizing the resources of the request receiving host, and ensuring maximum service processing while maintaining stable system operation.

[0096] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0097] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0098] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0099] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. An interface flow control method, characterized in that: include: Obtaining request data, where the request data is data to be sent to the interface; Inputting the request data into a threshold prediction model to obtain a threshold value of the number of requests sent to the interface; Performing flow control on the interface based on the request quantity threshold; The threshold prediction model is constructed by the following method: Obtain target data, including: CPU usage, memory usage, I / O usage, number of threads, and response time; Preprocessing the target data to obtain comprehensive indicator data with labels; Using the labeled comprehensive indicator data to train and test the machine learning model to obtain a threshold prediction model; Wherein, the machine learning model includes: an attention mechanism function; Generating an attention allocation coefficient for the labeled comprehensive indicator data using the attention mechanism function; The weights and biases of the threshold prediction model are updated by training using a back propagation algorithm, and the internal state is updated according to a forward propagation algorithm, wherein the threshold prediction model is a three-layer LSTM network model.

2. The method according to claim 1, characterized in that The target data is preprocessed to obtain comprehensive indicator data with labels, including: Performing labeling processing on the target data to obtain labeled data; The labeled data is subjected to dimensionality reduction processing to obtain labeled comprehensive indicator data.

3. The method according to claim 2, characterized in that The dimensionality reduction process of the labeled data is performed to obtain the labeled comprehensive indicator data, including: The principal component analysis technology is used to reduce the dimension of the labeled data to obtain labeled comprehensive indicator data.

4. The method according to any one of claims 1 to 3, characterized in that The method further comprises: The Adam optimization algorithm is used to minimize the loss function.

5. An interface flow control device, characterized in that: include: A data acquisition module is used to obtain request data, where the request data is data to be sent to the interface; A prediction module is used to input the request data into a threshold prediction model to obtain a threshold value of the number of requests sent to the interface, The threshold prediction model is constructed by the following method: Obtain target data, including: CPU usage, memory usage, I / O usage, number of threads, and response time; Preprocessing the target data to obtain comprehensive indicator data with labels; Using the labeled comprehensive indicator data to train and test the machine learning model to obtain a threshold prediction model; Wherein, the machine learning model includes: an attention mechanism function; Generating an attention allocation coefficient for the labeled comprehensive indicator data using the attention mechanism function; The threshold prediction model weights and biases are updated by training using a back-propagation algorithm, and the internal state is updated according to a forward propagation algorithm, wherein the threshold prediction model is a three-layer LSTM network model; A flow control module is used to perform flow control on the interface based on the request quantity threshold.

6. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instruction to implement the interface flow control method according to any one of claims 1 to 4.

7. A storage medium, characterized in that: When the instructions in the storage medium are executed by a processor of an information processing device or a server, the information processing device or the server implements the interface flow control method according to any one of claims 1 to 4.

Images