Request processing method, baseboard management controller, electronic device and storage medium

Abstract

The present application discloses a request processing method, a baseboard management controller, an electronic device and a storage medium, which relate to the field of computer technology and are applied to a baseboard management controller. The baseboard management controller includes a first processor, a second processor and at least two types of models, and the number of each type of model is at least one; the method includes: the first processor obtains target request data input by a user; when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the first processor sends the target request data to the second processor; the second processor processes the target request data based on the target model; the first processing condition is a condition for determining whether to use a model to process the target request data, and the target model is one of at least two types of models, which solves the technical problem of low functional flexibility of the baseboard management controller.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a request processing method, a baseboard management controller, an electronic device, and a storage medium. Background Art

[0002] The baseboard management controller (BMC) is an integral part of a server, responsible for monitoring and managing its health. Faced with complex business demands such as data analysis, predictive maintenance, and adaptive adjustment, traditional BMCs are difficult to customize and optimize based on these needs, resulting in limited BMC flexibility. Summary of the Invention

[0003] The present application provides a request processing method, a baseboard management controller, an electronic device and a storage medium, so as to at least solve the technical problem of low functional flexibility of the baseboard management controller in the related art.

[0004] The present application provides a request processing method, which is applied to a baseboard management controller, wherein the baseboard management controller includes a first processor, a second processor, and at least two types of models, with the number of each type of model being at least one; the method includes:

[0005] The first processor obtains target request data input by the user;

[0006] When the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the first processor sends the target request data to the second processor;

[0007] The second processor processes the target request data based on the target model; the first processing condition is a condition for determining whether to use the model to process the target request data, and the target model is one of at least two types of models.

[0008] The present application also provides a baseboard management controller, which is used to execute the above request processing method.

[0009] Through the present application, the first processor can obtain the target request data input by the user, and when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the target request data is sent to the second processor, and the second processor processes the target request data based on the target model, where the target model is one of at least two types of models, so as to realize the flexible processing of the request data based on multiple models by collaboratively calling the first processor and the second processor in the baseboard management controller, so as to achieve the purpose of improving the functional flexibility of the baseboard management controller. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0011] Figure 1 One of the flowcharts of a request processing method provided in an embodiment of the present application;

[0012] Figure 2 This is a second flowchart of a request processing method according to an embodiment of the present application;

[0013] Figure 3 This is a block diagram of a server logic structure according to one embodiment of the present application;

[0014] Figure 4 This is a schematic structural diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0016] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such 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 elements inherent to such process, method, article, or device. The terms "first," "second," etc., in this application are used to distinguish similar objects, and are not used to describe a particular order or sequence.

[0017] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0018] An embodiment of the present application provides a request processing method applied to a baseboard management controller (BMC), the BMC including a first processor, a second processor, and at least two types of models, each type of model having at least one. The method is described in detail in conjunction with the execution flow of the request processing method.

[0019] Specifically, Figure 1The present invention provides a flowchart of a request processing method according to an embodiment of the present application.

[0020] like Figure 1 As shown, the request processing method includes the following steps:

[0021] In step S110 , the first processor obtains target request data input by a user.

[0022] In actual implementation, the baseboard management controller may be stored in an eMMC (Embedded MultiMediaCard) circuit chip.

[0023] In some embodiments, the first processor may be a Cortex-A multi-core processor.

[0024] In actual implementation, the target request data can be request data for querying information or performing maintenance operations, or any other theoretically feasible request data. The target request data can include data in any format, such as text, images, etc., and this application does not impose specific restrictions on this.

[0025] In some embodiments, the first processor in the baseboard management controller may obtain target request data input by a user through a basic input output system (BIOS).

[0026] In some embodiments, the first processor in the baseboard management controller and the BIOS may interact based on H2B (Host to BMC) shared memory. The baseboard management controller may obtain target request data input by the user from the Basic Input Output System (BIOS) based on the H2B method.

[0027] In step S120, when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the first processor sends the target request data to the second processor; the first processing condition is a condition for determining whether to use the model to process the target request data.

[0028] In some embodiments, the second processor may be a Cotex-M4 helper processor.

[0029] In some embodiments, the first processor may determine whether to use the model to process the target request data based on a preset question library, that is, determine whether the target request data meets the first processing condition.

[0030] In some embodiments, the first processor may determine the target model based on the second processing condition when the target request data does not meet the first processing condition.

[0031] In some embodiments, the first processor may update the target threshold before determining whether the target request data meets the first processing condition.

[0032] In some embodiments, the first processor may determine whether its own occupancy rate is greater than or equal to a target threshold when the target request data does not meet the first processing condition.

[0033] In step S130 , the second processor processes the target request data based on the target model, where the target model is one of at least two types of models.

[0034] In some embodiments, the baseboard management controller may run a hierarchical heterogeneous architecture, which may include a preset question library and at least two types of models, with the number of models in each type being at least one.

[0035] In some embodiments, the baseboard management controller can adopt a collaborative operation of the first processor and the second processor to determine a target model based on the second processing condition and process the target request data based on the target model when the target request data does not meet the first processing condition.

[0036] In some embodiments, the hierarchical heterogeneous architecture of the baseboard management controller may include a model for resolving query requests input by a user, and may also include a model for resolving control requests input by a user. The first processor may determine, based on the second processing condition, a target model corresponding to the request type of the target request data input by the user from the hierarchical heterogeneous architecture.

[0037] In some embodiments, after processing the target request data based on the target model, the first processor may store and display the processing results of the target request data.

[0038] In some embodiments, the second processor may include at least one assisting processing module, and the second processor may process the target request data based on the target model through the at least one assisting processing module.

[0039] In some embodiments, the first processor may process the target request data based on the target model when the target request data does not meet the first processing condition and its own occupancy rate is less than the target threshold.

[0040] In some embodiments, after determining whether the target request data meets the first processing condition based on a preset question library, if the target request data does not meet the first processing condition, the first processor and the second processor are called to determine the target model based on the second processing condition, and the target request data is processed based on the target model.

[0041] In some embodiments, the baseboard management controller can call the target resources of the second processor to process the target request data based on the target model. The target resources may include memory, storage, bandwidth, computing resources, and any other theoretically feasible resources.

[0042] According to the request processing method of an embodiment of the present application, the first processor can obtain the target request data input by the user, and when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the target request data is sent to the second processor, and the second processor processes the target request data based on the target model, where the target model is one of at least two types of models, so as to realize the flexible processing of the request data based on multiple models by collaboratively calling the first processor and the second processor in the baseboard management controller, so as to achieve the purpose of improving the functional flexibility of the baseboard management controller.

[0043] In some embodiments, when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, before the first processor sends the target request data to the second processor, the first processor may determine whether the target request data meets the first processing condition based on a preset question library; when the target request data meets the first processing condition, the first processor processes the target request data based on the preset question library.

[0044] In some embodiments, the hierarchical heterogeneous architecture of the baseboard management controller may include a pre-processing module, and the first processor may determine whether the target request data meets the first processing condition based on a preset question library through the pre-processing module.

[0045] In some embodiments, the baseboard management controller can obtain target data from the server operating system and store it in a target database. The target data may include server operating data, such as CPU operating data, or any other theoretically feasible data. The target database may be a Redis database that supports high-speed caching and temporary data storage. The target database can be deployed on the baseboard management controller.

[0046] In some embodiments, a preset question library can be constructed based on the target data stored in the target database. The preset question library can be a quick response portal that can store the target data and preset requests corresponding to each piece of data in the target data.

[0047] In some embodiments, the first processor can determine whether the target request data meets the first processing condition based on the preset question library and the target request data, such as searching whether a preset request similar to the target request data is stored in the preset question library, or whether feedback data for the target request data can be generated based on the target data currently stored in the preset question library.

[0048] In some embodiments, the first processor may process the target request data based on a preset question library if the target request data meets the first processing condition. For example, the first processor may generate feedback data for the target request data based on the target data in the preset question library.

[0049] According to the request processing method of the embodiment of the present application, when the target request data does not meet the first processing condition, the target model can be determined based on the second processing condition, and before processing the target request data based on the target model, the first processor can determine whether the target request data meets the first processing condition based on the preset question library; when the target request data meets the first processing condition, the first processor processes the target request data based on the preset question library to avoid processing the target request data through the model when the target request data meets the first processing condition, so as to minimize the model's occupancy of the baseboard management controller resources.

[0050] In some embodiments, the first processor may search for feedback data that matches the target request data based on a preset question library; if feedback data exists, it may determine that the target request data meets the first processing condition.

[0051] In actual execution, the feedback data can be any data among the pre-stored target data in the preset question library.

[0052] In some embodiments, based on the target request data, a preset question library can be searched to determine whether there is feedback data matching the target request data, that is, whether the preset question library can process the target request data. If there is feedback data matching the target request data in the preset question library, it is determined that the target request data meets the first processing condition, and the target request data is then processed based on the preset question library to obtain a processing result.

[0053] In some embodiments, the first processor may search the preset question library to see whether there is feedback data matching the target request data based on cosine similarity, Jaccard Index, or any other theoretically feasible algorithm.

[0054] In some embodiments, the first processor can compare the existing target data and target request data in the preset question library one by one based on cosine similarity, Jaccard Index or any theoretically feasible algorithm to retrieve whether there is feedback data in the preset question library that matches the target request data.

[0055] In some embodiments, the first processor may determine that the target request data does not meet the first processing condition when there is no feedback data matching the target request data in the preset question library.

[0056] According to the request processing method of the embodiment of the present application, when the target request data does not meet the first processing condition, the target model can be determined based on the second processing condition, and before processing the target request data based on the target model, the first processor can determine whether the target request data meets the first processing condition based on the preset question library; when the target request data meets the first processing condition, the first processor processes the target request data based on the preset question library to avoid processing the target request data through the model when the target request data meets the first processing condition, so as to minimize the model's occupancy of the baseboard management controller resources.

[0057] In some embodiments, after the first processor determines whether the target request data meets the first processing condition based on a preset question library, if the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold value.

[0058] In actual execution, the occupancy rate of the first processor may be the occupancy rate of memory, bandwidth, storage, computing, or any theoretically feasible resource.

[0059] In some embodiments, the first processor may monitor its own occupancy rate in real time, so as to obtain its own occupancy rate when the target request data does not meet the first processing condition.

[0060] In some embodiments, the first processor may obtain its own occupancy rate when there is no feedback data matching the target request data in the preset question library (ie, the target request data does not meet the first processing condition).

[0061] In actual execution, the target threshold may be a pre-set value, and the target threshold may be set according to environmental operation data.

[0062] In some embodiments, the first processor may compare the acquired occupancy rate with a numerical value of a target threshold to determine whether the occupancy rate is greater than or equal to the target threshold.

[0063] According to the request processing method of the embodiment of the present application, the first processor can obtain the target request data input by the user. When the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold. When the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the target request data is sent to the second processor. The second processor processes the target request data based on the target model, and the target model is one of at least two types of models, so that the first processor can use its own occupancy rate to collaboratively call the first processor and the second processor in the baseboard management controller, reduce the occupancy rate of its own resources by processing the request data through the model, so as to realize flexible processing of request data based on multiple models, thereby achieving the purpose of improving the functional flexibility of the baseboard management controller.

[0064] In some embodiments, when the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold. When the occupancy rate is less than the target threshold, the first processor processes the target request data based on the target model and obtains the processing result.

[0065] In some embodiments, if the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to a target threshold. Then, the first processor may process the target request data based on the target model if its own occupancy rate is less than the target threshold. For example, the target threshold may be 80%, and the first processor may process the target request data based on the target model if its own occupancy rate is less than 80%.

[0066] In some embodiments, the first processor may process the target request data based on the target model by calling its own main processing module and obtain the processing result.

[0067] According to the request processing method of the embodiment of the present application, the first processor can obtain the target request data input by the user. When the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold. When the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the target request data is sent to the second processor. The second processor processes the target request data based on the target model, and the target model is one of at least two types of models, so that the first processor can use its own occupancy rate to collaboratively call the first processor and the second processor in the baseboard management controller, reduce the occupancy rate of its own resources by processing the request data through the model, so as to realize flexible processing of request data based on multiple models, thereby achieving the purpose of improving the functional flexibility of the baseboard management controller.

[0068] In some embodiments, the first processor includes at least one main processing module; when the occupancy rate is less than a target threshold, each main processing module in the at least one main processing module is called to process the target request data based on the target model.

[0069] In some embodiments, the first processor may include a first main processing module, a second main processing module, a third main processing module, and a fourth main processing module.

[0070] In some embodiments, the first processor may, when the occupancy rate is less than the target threshold, call at least one of the first main processing module, the second main processing module, the third main processing module, and the fourth main processing module to process the target request data based on the target model. For example, the first processor may, when the occupancy rate is less than the target threshold, call the first main processing module to process the target request data based on the target model. The first processor may, when the occupancy rate is less than the target threshold, call the second main processing module and the third main processing module to process the target request data based on the target model.

[0071] In some embodiments, when the occupancy rate is less than a target threshold, the first processor may invoke target resources of at least one of the first main processing module, the second main processing module, the third main processing module, and the fourth main processing module to process target request data based on the target model. For example, the first processor may invoke bandwidth resources of the first main processing module, memory resources of the second main processing module, and computing resources of the third main processing module to process target request data based on the target model.

[0072] According to the request processing method of an embodiment of the present application, the first processor can obtain target request data input by the user, and when the target request data does not meet the first processing condition and the occupancy rate is less than the target threshold, call each main processing module in at least one main processing module to process the target request data based on the target model, thereby further optimizing the allocation of resources of the first processor, freely scheduling each main processing module in at least one main processing module to process the target request data, improving the resource utilization of each main processing module in the first processor, and improving the efficiency of the first processor in processing the target request data.

[0073] In some embodiments, after the first processor determines whether the target request data meets the first processing condition based on a preset question library, if the target request data does not meet the first processing condition, the first processor determines the request type corresponding to the target request data based on the second processing condition; the first processor determines the target model based on the request type.

[0074] In actual execution, the target request data may be text, image, or any theoretically feasible form of data. Based on the second processing condition, the first processor may identify and judge the target request data to determine the request type corresponding to the target request data.

[0075] In some embodiments, the hierarchical heterogeneous architecture of the baseboard management controller may include multiple types of models (e.g., models for answering query requests input by users, models for resolving control requests input by users, or any other theoretically feasible models). In some embodiments, the first processor may determine the request type of the target request data (e.g., a query request, a control request, or any other theoretically feasible request type) based on the second processing condition, and then, based on the determined request type, determine, from the multiple types of models in the hierarchical heterogeneous architecture of the baseboard management controller, a model corresponding to the request type of the target request data as a target model.

[0076] According to the request processing method of an embodiment of the present application, the first processor can obtain the target request data input by the user, and determine the request type corresponding to the target request data based on the second processing condition if the target request data does not meet the first processing condition; based on the request type, determine the target model from at least two types of models and process the target request data based on the target model, so as to determine the target model corresponding to the request type of the target request data from multiple models of the baseboard management controller, process the target request data, and realize flexible processing of the request data through multiple models in the baseboard management controller, so as to achieve the purpose of improving the functional flexibility of the baseboard management controller.

[0077] In some embodiments, the request type includes a query type request; when the request type is a query type request, the first processor determines the target model from the first type of models.

[0078] In actual execution, the request type of the target request data may be a query type request, for example, the target request data may be request data indicating a query of the current CPU resource occupancy rate, for example, the target request data may be request data indicating a query of a certain type of data in a database of a device.

[0079] In some embodiments, the first processor may determine the target model from the first type of model when it is determined based on the second processing condition that the request type corresponding to the target request data is a query type request.

[0080] In practice, the first type of model can be a natural language model, which can be a pre-trained language model. This generally involves designing a language model training task based on a large-scale corpus (including language training materials such as sentences and paragraphs), and then training a large-scale neural network algorithm structure to learn and implement it. The resulting large-scale neural network algorithm structure and parameters are the pre-trained language model. Subsequent tasks can use this model for feature extraction or task fine-tuning to achieve specific task objectives. The idea behind pre-training is to first train a task to obtain a set of model parameters, then use this set of model parameters to initialize the network model parameters. The initialized network model is then used to train other tasks to obtain a model adapted for the other task. By pre-training on a large-scale corpus, the neural language representation model can acquire powerful language representation capabilities and extract rich syntactic and semantic information from text. The pre-trained language model can provide token- and sentence-level features containing rich semantic information for use in downstream tasks. Fine-tuning can also be performed directly on the pre-trained model for downstream tasks, quickly and easily obtaining a dedicated downstream model. The neural network algorithm structure used for pre-trained language model training can be CNN, RNN, LSTM, etc., or it can be a model built with an attention network, such as Transformer, BERT, GPT, Clip, etc., which is not limited in this application. An attention network refers to a network model that uses an attention mechanism for training. The model assigns different weights to each part of the input sequence, thereby extracting more important feature information from the input sequence, so that the model ultimately obtains a more accurate output. Fine-tuning refers to further training on a dataset for a specific task based on the use of a pre-trained model to adjust the model parameters to better adapt to the target task. During the fine-tuning process, most of the layers of the pre-trained model are usually frozen, and only newly added layers are trained or a small number of key layers are adjusted. This can not only retain the useful features learned by the pre-trained model, but also quickly adapt to the specific requirements of the new task. In addition, choosing an appropriate learning rate and training rounds is also the key to successful fine-tuning.

[0081] In actual implementation, the first type of model can be used to perform information retrieval, and the number of the first type of model is at least one. In some embodiments, when the request type is a query type request, a target model can be determined from multiple first type models.

[0082] According to the request processing method of an embodiment of the present application, the first processor can obtain the target request data input by the user, and determine the request type corresponding to the target request data based on the second processing condition if the target request data does not meet the first processing condition; based on the request type, determine the target model from at least two types of models and process the target request data based on the target model, so as to determine the target model corresponding to the request type of the target request data from multiple models of the baseboard management controller, process the target request data, and realize flexible processing of the request data through multiple models in the baseboard management controller, so as to achieve the purpose of improving the functional flexibility of the baseboard management controller.

[0083] In some embodiments, the request type includes a control type request; when the request type is a control type request, the first processor determines the target model from the second type of model.

[0084] In actual execution, the request type of the target request data may also be a control request. For example, the target request data may be request data for requesting image processing, or request data for requesting prediction of the cause of equipment failure or the remaining service life of a component of the equipment.

[0085] In actual implementation, the second type of model can be a dedicated model for various control functions, such as a model for predicting the remaining useful life of an equipment or a fault diagnosis model. The second type of model can be used to perform control operations. The number of second type models is at least one. In some embodiments, when the request type is a control request, a target model can be determined from multiple second type models.

[0086] In some embodiments, the first processor may determine a target model that matches the functional requirements of the target request data from multiple second-type models when it is determined based on the second processing condition that the request type corresponding to the target request data is a control-type request.

[0087] According to the request processing method of the embodiment of the present application, the first processor can obtain the target request data input by the user, and when the target request data does not meet the first processing condition, determine the request type corresponding to the target request data based on the second processing condition; when the request type is a query type request, determine the target model from the first type of model; or, when the request type is a control type request, determine the target model from the second type of model, and process the target request data based on the target model to determine the target model corresponding to the request type of the target request data from multiple models of the baseboard management controller, process the target request data, and realize flexible processing of the request data through multiple models in the baseboard management controller, so as to achieve the purpose of improving the functional flexibility and extensibility of the baseboard management controller.

[0088] In some embodiments, when the request type is a control type request, the first processor identifies the content of the target request data and determines the target model from the second type of model based on the identification result.

[0089] In actual execution, the recognition results may include the functional requirements requested by the target request data, such as equipment remaining service life prediction, fault diagnosis, etc.

[0090] In some embodiments, the text, images and other contents in the target request data may be recognized to obtain a recognition result indicating the functional requirements requested by the target request data.

[0091] In some embodiments, based on the recognition result, a second-category model that matches the functional requirement represented by the recognition result can be determined as a target model from multiple second-category models.

[0092] According to the request processing method of the embodiment of the present application, the first processor can obtain the target request data input by the user, and when the target request data does not meet the first processing condition, determine the request type corresponding to the target request data based on the second processing condition; when the request type is a query type request, determine the target model from the first type of model; or, when the request type is a control type request, identify the content of the target request data, determine the target model from the second type of model based on the identification result, and process the target request data based on the target model, so as to determine the target model corresponding to the functional requirements of the target request data from multiple models of the baseboard management controller according to different functional requirements identified from the target request data, and process the target request data, thereby improving the adaptability and scalability of the baseboard management controller.

[0093] In some embodiments, when the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate, and before determining whether the occupancy rate is greater than or equal to the target threshold value, the first processor updates the target threshold value based on at least one of the environmental operation data and the historical request processing results.

[0094] In some embodiments, the target threshold may represent a criterion for the first processor to determine, based on its own occupancy rate, whether to process the target request data based on the target model or to have the second processor process the target request data based on the target model. The target threshold may be a set value.

[0095] In actual execution, the environmental operating data may include the resource occupancy rate of the first processor or any other theoretically feasible operating data.

[0096] In some embodiments, when the environmental operating data indicates that the resource occupancy status of the first processor is poor, the target threshold can be increased (for example, the target threshold can be increased from 80% to 90%), so that the first processor can determine the standard for processing target request data based on the target model based on its own occupancy rate, thereby reducing the possibility of the first processor processing target request data based on the target model when its own resources are insufficient.

[0097] In some embodiments, when historical request data indicates that the probability of processing historical request data based on a preset question library is low, the target threshold can be increased (for example, the target threshold can be increased from 80% to 90%), so that the first processor determines the standard improvement of processing target request data based on the target model based on its own occupancy rate, so as to improve the possibility of the first processor processing target request data based on the preset question library when its own resources are insufficient.

[0098] According to the request processing method of an embodiment of the present application, when the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and, before determining whether the occupancy rate is greater than or equal to the target threshold, can update the target threshold based on at least one of the environmental operating data and historical request processing results. This allows the frequency of request data processing through the model to be dynamically adjusted based on the environmental operating data and historical request processing results, thereby enabling the preprocessing module to reduce the load pressure on the first processor. Furthermore, the mechanism of integrating the preprocessing module with the model not only improves the level of intelligence, but also enhances robustness and stability, enabling rapid response even in the face of unknown failure modes.

[0099] In some embodiments, when the environment operation data and / or the historical request processing results meet the first target update condition, the first processor reduces the target threshold.

[0100] In actual execution, the first target update condition may be that the environmental operation data is greater than a first preset value and / or the historical request processing result is greater than a second preset value.

[0101] In actual execution, if the environment operating data satisfies the first target update condition (i.e., the environment operating data is greater than the first preset value), it can indicate that the resource utilization of the first processor is in a good state. If the historical request processing result satisfies the first target update condition (i.e., the historical request processing result is greater than the second preset value), it can indicate that the probability of processing the historical request data based on the preset question library is high.

[0102] In some embodiments, when the environmental operating data indicates that the resource occupancy status of the first processor is good, the target threshold can be reduced (for example, the target threshold can be reduced from 80% to 70%), so that the first processor can determine based on its own occupancy rate to lower the standard for processing target request data based on the target model, thereby increasing the possibility of the first processor processing target request data based on the target model when its own resources are sufficient.

[0103] In some embodiments, when historical request data indicates that the probability of processing historical request data based on a preset question library is high, the target threshold can be reduced (for example, the target threshold can be reduced from 80% to 70%), so that the first processor determines based on its own occupancy rate that the standard for processing target request data based on the target model is lowered, thereby increasing the possibility of the first processor processing target request data based on the target model when its own resources are sufficient.

[0104] According to the request processing method of an embodiment of the present application, when the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold. If the environment operation data and / or historical request processing results meet the first target update condition, the first processor can reduce the target threshold. Based on the environment operation data and historical request processing results, the frequency of processing request data through the model is dynamically adjusted, thereby enabling the preprocessing module to reduce the load pressure on the first processor. Furthermore, the mechanism of integrating the preprocessing module with the model not only improves the level of intelligence, but also enhances robustness and stability, enabling rapid response even in the face of unknown failure modes.

[0105] In some embodiments, the second processor includes a first assisting processing module and a second assisting processing module; the target model includes a first model and a second model; when the target model is the first model, the second processor processes the target request data based on the first model through the first assisting processing module; when the target model is the second model, the second processor processes the target request data based on the second model through the second assisting processing module.

[0106] In actual implementation, the first model may be one of the first type of models or one of the second type of models. The second model may be one of the first type of models or one of the second type of models.

[0107] In actual implementation, the first model may be a model that executes an addition (ADD) algorithm, and the second model may be a model that executes a multi-layer multiplication neural network (eg, MUL, MUL_MAT, multiplication algorithm, multiplication matrix algorithm) algorithm.

[0108] In some embodiments, after determining the target model, the first processor may send information about the target model to the second processor. If the target model information indicates that the target model is the first model, the second processor may process the target request data based on the first model via the first assisting processing module. Correspondingly, if the target model information indicates that the target model is the second model, the second processor may process the target request data based on the second model via the second assisting processing module.

[0109] In some embodiments, the target model can be a model for executing multiple algorithms simultaneously. For example, the target model can execute an addition (ADD) algorithm, a multilayer multiplication neural network (such as MUL, MUL_MAT, a multiplication algorithm, a multiplication matrix algorithm) algorithm, and any theoretically feasible algorithm.

[0110] In some embodiments, the second processor can execute the first algorithm of the target model through the first assisting processing module, and the first algorithm can be a partial algorithm constituting the target model; the second processor can execute the second algorithm of the target model through the second assisting processing module, and the second algorithm can be another partial algorithm constituting the target model in addition to the first algorithm, so as to enable the second processor to process the target request data based on the target model.

[0111] According to the request processing method of an embodiment of the present application, the first processor can obtain the target request data input by the user, and send the target request data to the second processor when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold. When the target model is the first model, the second processor processes the target request data based on the first model through the first assisting processing module; when the target model is the second model, the second processor processes the target request data based on the second model through the second assisting processing module, so as to improve the efficiency of obtaining the processing results of the target request data by processing the target request data through the first assisting processing module or the second assisting processing module.

[0112] In some embodiments, after the second processor processes the target request data based on the target model, the first processor may capture the processing result for the target request data from the second processor to avoid the first processor waiting for the second processor to send the processing result.

[0113] In some embodiments, after the second processor processes the target request data based on the target model and obtains the processing result, the first processor can capture the processing result for the target request data based on the inter-core communication between itself and the second processor.

[0114] In some embodiments, after the first processor captures the processing result for the target request data, the first processor may store or display the processing result.

[0115] According to the request processing method of an embodiment of the present application, the first processor can obtain the target request data input by the user, and send the target request data to the second processor if the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold. After the second processor processes the target request data based on the target model, the first processor can capture the processing results for the target request data from the second processor to avoid waiting for processing results due to communication congestion, thereby improving the processing efficiency and response speed of the target request data.

[0116] In some embodiments, after the first processor captures the processing results for the target request data from the second processor, the first processor updates the target request data and the processing results to a preset question library.

[0117] In some embodiments, when the target request data does not meet the first processing condition, after the first processor or the second processor processes the target request data based on the target model, the first processor obtains the processing result for the target request data and updates the target request data and the processing result to the preset question library.

[0118] In some embodiments, after acquiring the processing result of the target request data from the second processor, the first processor may update the target request data and the processing result to the preset question library. When the first processor acquires the target request data again, the first processor may process the acquired target request data again based on the first processing condition and the preset question library.

[0119] According to the request processing method of an embodiment of the present application, after processing the target request data based on the target model, the first processor obtains the processing result for the target request data, and updates the target request data and the processing result to the preset question library. By caching, the frequency of model calls is reduced to reduce the load pressure through the preprocessing module.

[0120] In some embodiments, after the first processor captures the processing result for the target request data from the second processor, the first processor may display the processing result for the target request data based on a visual interface.

[0121] In some embodiments, after obtaining the processing result for the target request data, the first processor may display the processing result for the target request data based on a visual interface through a message prompt, a pop-up window, or any theoretically feasible form.

[0122] In some actual executions, the first processor may obtain further operations from the user after displaying the processing result for the target request data.

[0123] According to the request processing method of an embodiment of the present application, after obtaining the processing results for the target request data, the first processor can display the processing results for the target request data based on a visual interface to quickly discover and solve potential problems, thereby reducing downtime and maintenance costs caused by equipment failure.

[0124] In order to better understand the request processing method provided in the embodiments of the present application, further explanation is given below. It should be understood that the discussion below is only exemplary.

[0125] This application provides a request processing method, the specific steps can be as follows Figure 2 As shown:

[0126] Step S210: The first processor obtains target request data input by the user.

[0127] In some embodiments, the request processing method of this embodiment may be applied to a baseboard management controller, where the baseboard management controller includes at least two types of models, and the number of models in each type is at least one.

[0128] In some embodiments, the first processor may be a Cortex-A multi-core processor.

[0129] In actual implementation, the baseboard management controller may be stored in an eMMC (Embedded MultiMediaCard) circuit chip.

[0130] In actual implementation, the target request data can be request data for querying information or performing maintenance operations, or any other theoretically feasible request data. The target request data can include data in any format, such as text, images, etc., and this application does not impose specific restrictions on this.

[0131] In some embodiments, the first processor in the baseboard management controller may obtain target request data input by a user through a basic input output system (BIOS).

[0132] In some embodiments, as Figure 3 As shown, the first processor in the baseboard management controller and the basic input and output system can interact based on a shared memory method (e.g., H2B, i.e., Host to BMC). The baseboard management controller can obtain target request data input by the user from the basic input and output system (BIOS) based on the H2B method.

[0133] In step S213 , the first processor determines whether the target request data meets the first processing condition based on the preset question library.

[0134] In some embodiments, as Figure 3 As shown, the hierarchical heterogeneous architecture of the baseboard management controller may include a pre-processing module, and the first processor may determine whether the target request data meets the first processing condition based on a preset question library through the pre-processing module.

[0135] In some embodiments, the baseboard management controller can obtain target data from the server operating system and store it in a target database. The target data may include server operating data, such as CPU operating data, or any other theoretically feasible data. The target database may be a Redis database that supports caching and temporary data storage. The target database may be deployed on the baseboard management controller. The baseboard management controller can obtain the target data from the server operating system via the Platform Environment Control Interface (PECI).

[0136] In some embodiments, the baseboard management controller may also include memory, such as double data rate SDRAM (DDR), and the baseboard management controller may also include flash memory (FLASH) and a baseboard management controller common base module (BMC CBB). The pre-processing module and the flash memory may communicate based on a serial peripheral interface (Serial Peripheral Interface).

[0137] In some embodiments, the baseboard management controller can communicate with the central processing unit (e.g., central processing unit 0 and central processing unit 1) based on the platform environment control interface. The server may also include an operating system, memory (e.g., DDR), a solid-state drive (e.g., NVME (Non-Volatile Memory Express), a non-volatile storage device accessed through a PCIe interface), and a hard disk drive (HDD).

[0138] In some embodiments, a preset question library can be constructed based on the target data stored in the target database. The preset question library can be a quick response portal that can store the target data and preset requests corresponding to each piece of data in the target data.

[0139] In some embodiments, the first processor can determine whether the target request data meets the first processing condition based on the preset question library and the target request data, such as searching whether a preset request similar to the target request data is stored in the preset question library, or whether feedback data for the target request data can be generated based on the target data currently stored in the preset question library.

[0140] In some embodiments, the first processor may search for feedback data that matches the target request data based on a preset question library; if feedback data exists, it may determine that the target request data meets the first processing condition.

[0141] In actual execution, the feedback data can be any data among the pre-stored target data in the preset question library.

[0142] In some embodiments, based on the target request data, a preset question library can be searched to determine whether there is feedback data matching the target request data, that is, whether the preset question library can process the target request data. If there is feedback data matching the target request data in the preset question library, it is determined that the target request data meets the first processing condition, and the target request data is then processed based on the preset question library to obtain a processing result.

[0143] In some embodiments, the first processor may search the preset question library to see whether there is feedback data matching the target request data based on cosine similarity, Jaccard Index, or any other theoretically feasible algorithm.

[0144] In some embodiments, the first processor can compare the existing target data and target request data in the preset question library one by one based on cosine similarity, Jaccard Index or any theoretically feasible algorithm to retrieve whether there is feedback data in the preset question library that matches the target request data.

[0145] In some embodiments, the first processor may determine that the target request data does not meet the first processing condition when there is no feedback data matching the target request data in the preset question library.

[0146] Step S216: When the target request data meets the first processing condition, the first processor processes the target request data based on the preset question library.

[0147] In some embodiments, the first processor may process the target request data based on a preset question library if the target request data meets the first processing condition. For example, the first processor may generate feedback data for the target request data based on the target data in the preset question library.

[0148] Step S219: When the target request data does not meet the first processing condition, the first processor determines the request type corresponding to the target request data based on the second processing condition; the first processor determines the target model based on the request type.

[0149] In actual execution, the target request data may be text, image, or any theoretically feasible form of data. Based on the second processing condition, the first processor may identify and judge the target request data to determine the request type corresponding to the target request data.

[0150] In some embodiments, the hierarchical heterogeneous architecture of the baseboard management controller may include multiple types of models (e.g., models for answering query requests input by users, models for resolving control requests input by users, or any other theoretically feasible types). In some embodiments, the first processor may determine the request type of the target request data (e.g., a query request, a control request, or any other theoretically feasible request type) based on the second processing condition, and then, based on the determined request type, determine, from the multiple types of models in the hierarchical heterogeneous architecture of the baseboard management controller, a model corresponding to the request type of the target request data as a target model.

[0151] In some embodiments, the request type includes a query type request; when the request type is a query type request, the first processor determines the target model from the first type of models.

[0152] In actual execution, the request type of the target request data may be a query type request, for example, the target request data may be request data indicating a query of the current CPU resource occupancy rate, for example, the target request data may be request data indicating a query of a certain type of data in a database of a device.

[0153] In some embodiments, the first processor may determine the target model from the first type of model when it is determined based on the second processing condition that the request type corresponding to the target request data is a query type request.

[0154] In actual execution, the first type of model can be used to perform information retrieval, the number of the first type of model is at least one, and the first type of model can be a natural language model. In some embodiments, when the request type is a query type request, a target model can be determined from multiple first type models.

[0155] In some embodiments, the request type includes a control type request; when the request type is a control type request, the first processor determines the target model from the second type of model.

[0156] In actual execution, the request type of the target request data may also be a control request. For example, the target request data may be request data for requesting image processing, or request data for requesting prediction of the cause of equipment failure or the remaining service life of a component of the equipment.

[0157] In actual implementation, the second type of model can be a dedicated model for various control functions, such as a model for predicting the remaining useful life of an equipment or a fault diagnosis model. The second type of model can be used to perform control operations. The number of second type models is at least one. In some embodiments, when the request type is a control request, a target model can be determined from multiple second type models.

[0158] In some embodiments, the first processor may determine a target model that matches the functional requirements of the target request data from multiple second-type models when it is determined based on the second processing condition that the request type corresponding to the target request data is a control-type request.

[0159] In some embodiments, when the request type is a control type request, the first processor identifies the content of the target request data and determines the target model from the second type of model based on the identification result.

[0160] In actual execution, the recognition results may include the functional requirements requested by the target request data, such as equipment remaining service life prediction, fault diagnosis, etc.

[0161] In some embodiments, the text, images and other contents in the target request data may be recognized to obtain a recognition result indicating the functional requirements requested by the target request data.

[0162] In some embodiments, based on the recognition result, a second-category model that matches the functional requirement represented by the recognition result can be determined as a target model from multiple second-category models.

[0163] In some embodiments, after determining the target model, the first processor may send information about the target model to the second processor.

[0164] Step S222: When the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold value.

[0165] In actual execution, the occupancy rate of the first processor may be the occupancy rate of memory, bandwidth, storage, computing, or any theoretically feasible resource.

[0166] In some embodiments, the first processor may monitor its own occupancy rate in real time, so as to obtain its own occupancy rate when the target request data does not meet the first processing condition.

[0167] In some embodiments, the first processor may obtain its own occupancy rate when there is no feedback data matching the target request data in the preset question library (ie, the target request data does not meet the first processing condition).

[0168] In actual execution, the target threshold may be a pre-set value, and the target threshold may be set according to environmental operation data.

[0169] In some embodiments, the target threshold may represent a criterion for the first processor to determine, based on its own occupancy rate, whether to process the target request data based on the target model or to have the second processor process the target request data based on the target model. The target threshold may be a set value.

[0170] In actual execution, the environmental operating data may include the resource occupancy rate of the first processor or any other theoretically feasible operating data.

[0171] In some embodiments, when the environmental operating data indicates that the resource occupancy status of the first processor is poor, the target threshold can be increased (for example, the target threshold can be increased from 80% to 90%), so that the first processor can determine the standard for processing target request data based on the target model based on its own occupancy rate, thereby reducing the possibility of the first processor processing target request data based on the target model when its own resources are insufficient.

[0172] In some embodiments, when historical request data indicates that the probability of processing historical request data based on a preset question library is low, the target threshold can be increased (for example, the target threshold can be increased from 80% to 90%), so that the first processor determines the standard improvement of processing target request data based on the target model based on its own occupancy rate, so as to improve the possibility of the first processor processing target request data based on the preset question library when its own resources are insufficient.

[0173] In some embodiments, when the environment operation data and / or the historical request processing results meet the first target update condition, the first processor reduces the target threshold.

[0174] In actual execution, the first target update condition may be that the environmental operation data is greater than a first preset value and / or the historical request processing result is greater than a second preset value.

[0175] In actual execution, if the environment operating data satisfies the first target update condition (i.e., the environment operating data is greater than the first preset value), it can indicate that the resource utilization of the first processor is in a good state. If the historical request processing result satisfies the first target update condition (i.e., the historical request processing result is greater than the second preset value), it can indicate that the probability of processing the historical request data based on the preset question library is high.

[0176] In some embodiments, when the environmental operating data indicates that the resource occupancy status of the first processor is good, the target threshold can be reduced (for example, the target threshold can be reduced from 80% to 70%), so that the first processor can determine based on its own occupancy rate to lower the standard for processing target request data based on the target model, thereby increasing the possibility of the first processor processing target request data based on the target model when its own resources are sufficient.

[0177] In some embodiments, when historical request data indicates that the probability of processing historical request data based on a preset question library is high, the target threshold can be reduced (for example, the target threshold can be reduced from 80% to 70%), so that the first processor determines based on its own occupancy rate that the standard for processing target request data based on the target model is lowered, thereby increasing the possibility of the first processor processing target request data based on the target model when its own resources are sufficient.

[0178] In some embodiments, the first processor may compare the acquired occupancy rate with a numerical value of a target threshold to determine whether the occupancy rate is greater than or equal to the target threshold.

[0179] Step S225 : When the occupancy rate is less than the target threshold, the first processor processes the target request data based on the target model.

[0180] In some embodiments, if the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to a target threshold. Then, the first processor may process the target request data based on the target model if its own occupancy rate is less than the target threshold. For example, the target threshold may be 80%, and the first processor may process the target request data based on the target model if its own occupancy rate is less than 80%.

[0181] In some embodiments, the first processor may process the target request data based on the target model by calling its own main processing module and obtain the processing result.

[0182] In some embodiments, the first processor includes at least one main processing module; when the occupancy rate is less than a target threshold, each main processing module in the at least one main processing module is called to process the target request data based on the target model.

[0183] In some embodiments, the first processor may include a first main processing module, a second main processing module, a third main processing module, and a fourth main processing module.

[0184] In some embodiments, the first processor may, when the occupancy rate is less than the target threshold, call at least one of the first main processing module, the second main processing module, the third main processing module, and the fourth main processing module to process the target request data based on the target model. For example, the first processor may, when the occupancy rate is less than the target threshold, call the first main processing module to process the target request data based on the target model. The first processor may, when the occupancy rate is less than the target threshold, call the second main processing module and the third main processing module to process the target request data based on the target model.

[0185] In some embodiments, when the occupancy rate is less than a target threshold, the first processor may invoke target resources of at least one of the first main processing module, the second main processing module, the third main processing module, and the fourth main processing module to process target request data based on the target model. For example, the first processor may invoke bandwidth resources of the first main processing module, memory resources of the second main processing module, and computing resources of the third main processing module to process target request data based on the target model.

[0186] Step S228: When the occupancy rate is greater than or equal to the target threshold, the first processor sends the target request data to the second processor.

[0187] Step S231: The second processor processes the target request data based on the target model.

[0188] In some embodiments, the second processor includes a first assisting processing module and a second assisting processing module; the target model includes a first model and a second model; when the target model is the first model, the second processor processes the target request data based on the first model through the first assisting processing module; when the target model is the second model, the second processor processes the target request data based on the second model through the second assisting processing module.

[0189] In actual implementation, the first model may be one of the first type of models or one of the second type of models. The second model may be one of the first type of models or one of the second type of models.

[0190] In actual implementation, the first model may be a model that executes an addition (ADD) algorithm, and the second model may be a model that executes a multi-layer multiplication neural network (eg, MUL, MUL_MAT) algorithm.

[0191] In some embodiments, after determining the target model, the first processor may send information about the target model to the second processor. If the target model information indicates that the target model is the first model, the second processor may process the target request data based on the first model via the first assisting processing module. Correspondingly, if the target model information indicates that the target model is the second model, the second processor may process the target request data based on the second model via the second assisting processing module.

[0192] In some embodiments, after the second processor processes the target request data based on the target model, the second processor obtains a processing result for the target request data and sends the processing result to the first processor.

[0193] In some embodiments, after the second processor processes the target request data based on the target model and obtains the processing result, the second processor may send the processing result to the first processor based on inter-core communication between itself and the first processor.

[0194] Step S234: The first processor obtains the processing result.

[0195] In some embodiments, after the second processor processes the target request data based on the target model and obtains the processing result, the first processor can capture the processing result for the target request data based on the inter-core communication between itself and the second processor.

[0196] In step S237 , the first processor obtains the processing result for the target request data, and updates the target request data and the processing result to the preset question library.

[0197] In some embodiments, after obtaining the processing result for the target request data, the first processor may update the target request data and the processing result to the preset question library. When the first processor obtains the target request data again, the first processor may process the again obtained target request data using the preset question library based on the first processing condition.

[0198] In some embodiments, the first processor obtains the processing results for the target request data, and updates the target request data and the processing results to a preset question library, thereby reducing the frequency of model calls through caching, so as to reduce the load pressure through the preprocessing module.

[0199] In step S240 , the first processor displays the processing result of the target request data based on the visual interface.

[0200] In some embodiments, after obtaining the processing result for the target request data, the first processor may display the processing result for the target request data based on a visual interface through a message prompt, a pop-up window, or any theoretically feasible form.

[0201] In some actual executions, the first processor may obtain further operations from the user after displaying the processing result for the target request data.

[0202] According to the request processing method of the embodiment of the present application, the first processor can obtain the target request data input by the user, and when the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the target request data is sent to the second processor, and the second processor processes the target request data based on the target model, which is one of at least two types of models, so as to realize the flexible processing of the request data based on multiple models by collaboratively calling the first processor and the second processor in the baseboard management controller, so as to achieve the purpose of improving the functional flexibility of the baseboard management controller. At the same time, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to the target threshold. When the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to the target threshold, the target request data is sent to the second processor, and the second processor processes the target request data based on the target model, which is one of at least two types of models, so as to realize the first processor collaboratively calling the first processor and the second processor in the baseboard management controller through its own occupancy rate, thereby improving the reliability and stability of the processing process of the target request data.

[0203] The embodiments of the present application further provide a baseboard management controller, which is used to execute the request processing method in any of the above embodiments.

[0204] In some embodiments, as Figure 4 As shown, an embodiment of the present application also provides an electronic device 400, including a processor 401, a memory 402, and a computer program stored in the memory 402 and executable on the processor 401. When the program is executed by the processor 401, the various processes of the above-mentioned request processing method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0205] It should be noted that the computer devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0206] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above request processing method embodiments when running.

[0207] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, 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.

[0208] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of any of the above request processing method embodiments are implemented.

[0209] An embodiment of the present application further provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the above request processing method embodiments are implemented.

[0210] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0211] The above is a detailed introduction to a request processing method and program product provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only intended to help understand the method and core ideas of the present application. It should be noted that, for those skilled in the art, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

[0212] The above describes in detail the disk array control method and program product provided by this application. This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is intended only to facilitate understanding of the method and core concepts of this application. It should be noted that those skilled in the art may make various improvements and modifications to this application without departing from the principles of this application, and such improvements and modifications also fall within the scope of protection of the claims of this application.

Claims

1. A request processing method, characterized in that: Applied to a baseboard management controller, the baseboard management controller includes a first processor, a second processor, and at least two types of models, with the number of each type of model being at least one; the method includes: The first processor obtains target request data input by a user; The first processor searches, based on a preset question library, for feedback data that matches the target request data; In the case where the feedback data exists, determining that the target request data meets the first processing condition; When the target request data meets the first processing condition, the first processor processes the target request data based on the preset question library; If the target request data does not meet the first processing condition and the occupancy rate of the first processor is greater than or equal to a target threshold, the first processor sends the target request data to the second processor; The second processor processes the target request data based on a target model; the first processing condition is a condition for determining whether to use a model to process the target request data, and the target model is one of the at least two types of models.

2. The method according to claim 1, characterized in that After determining that the target request data meets the first processing condition in the presence of the feedback data, the method includes: In a case where the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to a target threshold value.

3. The method according to claim 2, characterized in that If the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to a target threshold value, the method further includes: When the occupancy rate is less than the target threshold, the first processor processes the target request data based on a target model and obtains a processing result.

4. The method according to claim 3, characterized in that The first processor includes at least one main processing module; when the occupancy rate is less than a target threshold, the first processor processes the target request data based on a target model, including: When the occupancy rate is less than the target threshold, each of the at least one main processing module is called to process the target request data based on a target model.

5. The method according to claim 1, wherein After determining that the target request data meets the first processing condition in the presence of the feedback data, the method further includes: If the target request data does not meet the first processing condition, the first processor determines the request type corresponding to the target request data based on the second processing condition; The first processor determines a target model based on the request type.

6. The method according to claim 5, characterized in that The request type includes query type request; The first processor determines a target model based on the request type, including: When the request type is a query request, the first processor determines a target model from the first type of models.

7. The method according to claim 5, characterized in that The request type includes a control request; The first processor determines a target model based on the request type, including: When the request type is a control request, the first processor determines a target model from the second type of models.

8. The method according to claim 7, characterized in that When the request type is a control request, the first processor determines the target model from the second type of model, including: In the case where the request type is a control request, the first processor identifies the content of the target request data; Based on the recognition results, the target model is determined from the second type of models.

9. The method according to claim 2, characterized in that When the target request data does not meet the first processing condition, the first processor obtains its own occupancy rate and determines whether the occupancy rate is greater than or equal to a target threshold value, the method includes: The first processor updates the target threshold based on at least one of environmental operation data and historical request processing results.

10. The method according to claim 9, characterized in that The first processor updates the target threshold based on at least one of the environmental operation data and the historical request processing result, including: When the environment operation data and / or the historical request processing result meets the first target update condition, the first processor reduces the target threshold.

11. The method according to claim 1, wherein The second processor includes a first assisting processing module and a second assisting processing module; The target model includes a first model and a second model; The second processor processes the target request data based on the target model, including: In a case where the target model is the first model, the second processor processes the target request data based on the first model through the first auxiliary processing module; When the target model is the second model, the second processor processes the target request data based on the second model through a second auxiliary processing module.

12. The method according to claim 1, characterized in that After the second processor processes the target request data based on the target model, the method further includes: The first processor fetches a processing result for the target request data from the second processor, so as to avoid the first processor waiting for the second processor to send the processing result.

13. The method according to any one of claims 1 to 12, characterized in that After the first processor captures the processing result of the target request data from the second processor, the method includes: The first processor updates the target request data and the processing result to a preset question library.

14. The method according to any one of claims 1 to 12, characterized in that After the first processor captures the processing result of the target request data from the second processor, the method further includes: The first processor displays the processing result of the target request data based on a visual interface.

15. A baseboard management controller, characterized in that: The baseboard management controller is used to execute the request processing method according to any one of claims 1 to 14.

16. An electronic device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the request processing method according to any one of claims 1 to 14.

17. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the request processing method according to any one of claims 1 to 14 are implemented.

18. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the request processing method according to any one of claims 1 to 14 are implemented.

Images