A configuration method of an operating system kernel and a related product

Abstract

The application provides a configuration method of an operating system kernel and related products, and relates to the technical field of computer networks. The method can ensure that the output parameter configuration value is matched with the hardware resources of the computing node and is a parameter configuration value for a specific running task by inputting the obtained hardware configuration information of the computing node and the running characteristics of the computing node in a running task state into a kernel parameter configuration model. Furthermore, the method can realize automatic dynamic optimization of the operating kernel of the computing node by generating the operating system kernel of the computing node according to the parameter configuration value output by the kernel parameter configuration model and deploying the generated operating system kernel to the computing node, thereby reducing the workload caused by manual configuration and manual optimization, improving the adaptability of the operating system kernel to hardware resources and running task requirements, and thus improving the configuration rate and accuracy of the operating system kernel.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer networks, and in particular to a configuration method of an operating system kernel and related products. BACKGROUND

[0002] A High Performance Computing (HPC) cluster is usually composed of multiple computing nodes for processing large-scale and complex running tasks. The operating system (OS) kernel of the computing node, as a bridge connecting hardware and application programs, its performance and efficiency directly affect the running efficiency of the entire HPC cluster. However, the currently widely used OS kernel is generally designed, which undoubtedly limits the full play of the performance of the HPC cluster.

[0003] In the related art, the OS kernel can be customized for specific HPC cluster environments and application requirements through artificial configuration and manual optimization. However, this artificial configuration and manual optimization method is not only complex, time-consuming and laborious, but also highly dependent on the professional skills and experience of the operator, and is prone to configuration errors due to human factors, affecting the rate and accuracy of kernel configuration. SUMMARY

[0004] To solve the above problems, the embodiments of the present application provide a configuration method of an operating system kernel and related products, which can improve the rate and accuracy of kernel configuration.

[0005] In a first aspect, the embodiments of the present application provide a configuration method of an operating system kernel, which comprises: obtaining hardware configuration information of a computing node and running characteristics of the computing node in a running task state; inputting the hardware configuration information and the running characteristics into a kernel parameter configuration model to obtain a parameter configuration value of an operating system kernel of the computing node; generating the operating system kernel of the computing node according to the parameter configuration value; and deploying the generated operating system kernel to the computing node. Thus, the method can ensure that the output parameter configuration value is matched with the hardware resources of the computing node and is a parameter configuration value for a specific running task by inputting the obtained hardware configuration information of the computing node and the running characteristics of the computing node in the running task state into the kernel parameter configuration model. Moreover, the method can realize automatic dynamic optimization of the operating kernel of the computing node by generating the operating system kernel of the computing node through the parameter configuration value output by the kernel parameter configuration model and deploying the generated operating system kernel to the computing node, which reduces the workload caused by artificial configuration and manual optimization, improves the adaptability of the operating system kernel to hardware resources and running task requirements, and thus improves the configuration rate and accuracy of the operating system kernel.

[0006] In some specific implementation manners, the kernel parameter configuration model comprises a parameter updating sub-model, and the parameter configuration value of the operating system kernel of the computing node is obtained by inputting the hardware configuration information and the running feature into the kernel parameter configuration model, comprising: determining the to-be-updated parameter item of the operating system kernel of the computing node by inputting the hardware configuration information and the running feature into the parameter updating sub-model of the kernel parameter configuration model, and determining the parameter configuration value for the to-be-updated parameter item.

[0007] In some specific implementation manners, the kernel parameter configuration model comprises a clustering sub-model and a plurality of parameter updating sub-models, and the to-be-updated parameter item of the operating system kernel of the computing node is determined by inputting the hardware configuration information and the running feature into the updating sub-model of the kernel parameter configuration model, comprising: determining the grouping result corresponding to the hardware configuration information and the running feature by inputting the hardware configuration information and the running feature into the clustering sub-model of the kernel parameter configuration model; and determining the to-be-updated parameter item of the operating system kernel of the computing node by inputting the hardware configuration information and the running feature into the parameter updating sub-model corresponding to the grouping result.

[0008] In some specific implementation manners, the method further comprises: if the running feature of the computing node in the running task state after deploying the operating system kernel is superior to the running feature of the computing node in the running task state before deploying the operating system kernel, updating the model parameters of the kernel parameter configuration model according to the hardware configuration information, the running feature and the parameter configuration value corresponding to the operating system kernel.

[0009] In some specific implementation manners, the operating system kernel of the computing node is generated according to the parameter configuration value, comprising: determining the non-updated parameter item of the operating system kernel according to the original parameter configuration value of the non-updated parameter item and the parameter configuration value output by the kernel parameter configuration model; and generating the operating system kernel of the computing node according to the original parameter configuration value of the non-updated parameter item and the parameter configuration value output by the kernel parameter configuration model.

[0010] In some specific implementation manners, the generated operating system kernel is deployed to the computing node, comprising: when the running task is not running, or when the running feature represents that the real-time load of the running task is lower than the preset load threshold, the generated operating system kernel is deployed to the computing node.

[0011] In some specific implementations, the hardware configuration information of the computing node and the running features of the computing node in the running task state are obtained, and the obtaining includes: obtaining the hardware configuration information of a plurality of computing nodes in a computing cluster and the running features of the plurality of computing nodes in the running task state; inputting the hardware configuration information and the running features into a kernel parameter configuration model to obtain parameter configuration values of an operating system kernel of the computing node, including: inputting the hardware configuration information and the running features of the plurality of computing nodes into clustering sub-models of the kernel parameter configuration model to determine grouping results corresponding to the plurality of computing nodes respectively; inputting the hardware configuration information and the running features of the plurality of computing nodes into parameter updating sub-models corresponding to the grouping results corresponding to the plurality of computing nodes respectively to determine to-be-updated parameter items of the operating system kernel of the plurality of computing nodes and determine parameter configuration values for the to-be-updated parameter items; and generating the operating system kernel of the computing node according to the parameter configuration values, including: generating corresponding operating system kernels for the plurality of computing nodes according to the parameter configuration values of the plurality of computing nodes; and deploying the generated operating system kernels to the computing nodes, including: deploying the operating system kernels corresponding to the plurality of computing nodes to the corresponding plurality of computing nodes respectively.

[0012] In a second aspect, the embodiments of the present application disclose a model training method, which includes: obtaining a sample data set, the sample data set including hardware configuration information of a plurality of computing nodes in a historical time period, running features of the plurality of computing nodes in the running task state in the historical time period, and parameter configuration values of operating system kernels of the plurality of computing nodes in the historical time period; and training a machine learning model according to the sample data set to obtain a kernel parameter configuration model. Thus, the method trains the machine learning model by taking the hardware configuration information of the plurality of computing nodes in the historical time period, the running features of the plurality of computing nodes in the running task state in the historical time period, and the parameter configuration values of the operating system kernels of the plurality of computing nodes in the historical time period as the sample data set, so as to ensure that the parameter configuration values output in the actual application of the kernel parameter configuration model trained are matched with the hardware resources of the computing nodes and are parameter configuration values for specific running tasks, thereby reducing the workload caused by manual configuration and manual optimization, improving the adaptability of the operating system kernel to the hardware resources and the running task requirements, and further improving the configuration rate and accuracy of the operating system kernel.

[0013] In some specific implementations, the obtaining the sample data set comprises: obtaining hardware configuration information of the plurality of computing nodes in a historical time period and running features of the plurality of computing nodes in the running task state in the historical time period; clustering the plurality of computing nodes according to the hardware configuration information of the plurality of computing nodes in the historical time period and the running features of the plurality of computing nodes in the running task state in the historical time period to obtain a plurality of grouping results, wherein the similarity of the hardware configuration information of the computing nodes in each grouping result is higher than a first similarity threshold, and the similarity of the running features of the computing nodes in each grouping result is higher than a second similarity threshold; and determining, for each grouping result, the parameter item to be updated and determining the parameter configuration value for the parameter item to be updated.

[0014] In a third aspect, an embodiment of the present application provides a configuration device of an operating system kernel, which comprises: an information obtaining module, a parameter obtaining module, a kernel updating module, and a kernel deploying module; the information obtaining module is configured to obtain hardware configuration information of a computing node and running features of the computing node in a running task state; the parameter obtaining module is configured to input the hardware configuration information and the running features into a kernel parameter configuration model to obtain a parameter configuration value of an operating system kernel of the computing node; the kernel updating module is configured to generate the operating system kernel of the computing node according to the parameter configuration value; and the kernel deploying module is configured to deploy the generated operating system kernel to the computing node. Thus, the device can ensure that the output parameter configuration value is matched with the hardware resources of the computing node and is a parameter configuration value for a specific running task by inputting the obtained hardware configuration information of the computing node and the running features of the computing node in the running task state into the kernel parameter configuration model. Moreover, the device can realize automatic dynamic optimization of the operating kernel of the computing node by generating the operating system kernel of the computing node according to the parameter configuration value output by the kernel parameter configuration model and deploying the generated operating system kernel to the computing node, thereby reducing the workload caused by manual configuration and manual optimization, improving the adaptability of the operating system kernel to the hardware resources and the running task requirements, and thus improving the configuration rate and accuracy of the operating system kernel.

[0015] In a fourth aspect, an embodiment of the present application discloses a model training device, which comprises an acquisition module and a training module; the acquisition module is configured to acquire a sample data set, the sample data set comprising hardware configuration information of a plurality of computing nodes in a historical time period, running features of the plurality of computing nodes in a running task state in the historical time period, and parameter configuration values of an operating system kernel of the plurality of computing nodes in the historical time period; and the training module is configured to train a machine learning model according to the sample data set to obtain a kernel parameter configuration model. Thus, the device trains the machine learning model by taking the hardware configuration information of the plurality of computing nodes in the historical time period, the running features of the plurality of computing nodes in the running task state in the historical time period, and the parameter configuration values of the operating system kernel of the plurality of computing nodes in the historical time period as the sample data set, so as to ensure that the parameter configuration values output in actual application of the kernel parameter configuration model trained are matched with hardware resources of the computing nodes and are parameter configuration values for specific running tasks, thereby reducing workload caused by manual configuration and manual optimization, improving adaptability of the operating system kernel to hardware resources and running task requirements, and further improving configuration speed and accuracy of the operating system kernel.

[0016] In a fifth aspect, an embodiment of the present application provides an electronic device, which comprises a memory configured to store a computer program or computer instructions, and a processor configured to execute the computer program or computer instructions stored in the memory, so that the electronic device executes the method of the first aspect or the second aspect.

[0017] In a sixth aspect, an embodiment of the present application provides a computer readable storage medium, which is characterized by storing a computer program or instructions, and when the computer program or instructions are executed, the method of the first aspect or the second aspect is implemented. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 A scenario architecture diagram of the configuration method of the operating system kernel provided by an embodiment of the present application;

[0019] Figure 2 A flowchart of the configuration method of the operating system kernel provided by an embodiment of the present application;

[0020] Figure 3 A flowchart of another configuration method of the operating system kernel provided by an embodiment of the present application;

[0021] Figure 4 A flowchart of the training method of the kernel parameter configuration model provided by an embodiment of the present application;

[0022] Figure 5 A flowchart of the model training method provided by an embodiment of the present application;

[0023] Figure 6 A schematic diagram of a configuration device of an operating system kernel provided for an embodiment of the present application;

[0024] Figure 7 A schematic diagram of a model training device provided for an embodiment of the present application. DETAILED DESCRIPTION

[0025] The terms “first”, “second” in the embodiments of the present application are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with “first”, “second” can explicitly or implicitly include one or more of the features.

[0026] First, the technical terms related to the embodiments of the present application are explained:

[0027] A high performance computing (HPC) cluster is formed by tens to thousands of computing nodes (e.g. multiple computers) interconnected by high-speed networks, each computing node having high-performance computing capability. These computing nodes work together to solve complex running tasks.

[0028] The operating system (OS) kernel is the core of the computing node, which is responsible for managing the hardware resources of the computing node, providing the program running environment corresponding to the running task, and serving as a bridge between the user program and the hardware device.

[0029] As described above, the OS kernel can be customized for specific HPC cluster environment and application requirements by manual configuration and manual optimization. However, this manual configuration and manual optimization method is not only complex, time-consuming and laborious, but also highly dependent on the professional skills and experience of the operator, which is prone to configuration errors due to human factors, affecting the rate and accuracy of kernel configuration.

[0030] Therefore, the embodiments of the present application provide a method for configuring an operating system kernel and related products, the method comprising: obtaining hardware configuration information of a computing node and running characteristics of the computing node in a running task state; inputting the hardware configuration information and the running characteristics into a kernel parameter configuration model to obtain a parameter configuration value of an operating system kernel of the computing node; generating the operating system kernel of the computing node according to the parameter configuration value; and deploying the generated operating system kernel to the computing node.

[0031] Therefore, the configuration method of the operating system kernel can ensure that the output parameter configuration value is matched with the hardware resources of the computing node and is a parameter configuration value for a specific running task by obtaining the hardware configuration information of the computing node and the running characteristics of the computing node in the running task state, and inputting the hardware configuration information and the running characteristics into the kernel parameter configuration model.

[0032] Moreover, the operating system kernel of the computing node can be updated by the parameter configuration value output by the kernel parameter configuration model, and the updated operating system kernel can be deployed to the computing node, so that automatic and dynamic optimization of the operating kernel of the computing node can be implemented, the workload caused by traditional manual configuration and manual optimization is reduced, the adaptability of the operating system kernel to the hardware resources and the running task requirements is improved, and therefore the rate and accuracy of kernel configuration are improved.

[0033] Further, since the parameter configuration value is dynamically generated according to the hardware configuration information of the computing node and the running characteristics of the computing node in the running task state, the parameter configuration value can better adapt to the changes of the hardware resources and the running task requirements, and the adaptability ensures that the computing node can maintain efficient operation under different loads and task requirements. Moreover, the updated operating system kernel can better utilize the hardware resources of the computing node, so that the computing efficiency is improved, and the risk of computing node crash or performance degradation caused by insufficient hardware resources is reduced.

[0034] Next, the execution subject of the configuration method of the operating system kernel provided in the embodiments of the present application is specifically introduced:

[0035] The execution subject of the configuration method of the operating system kernel provided in the embodiments of the present application can be a computer device with data processing capability, and the computer device can specifically be a terminal device or a server. As an example, the terminal device can specifically include but is not limited to a mobile phone, a desktop computer, a tablet computer, a notebook computer, a palm computer, a smart voice interaction device, a smart home appliance, a vehicle-mounted terminal, an aircraft, etc. The server can be a standalone physical server, or a server cluster or a distributed system composed of multiple physical servers. In addition, the server can also be a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs, and basic cloud computing services such as big data and artificial intelligence platforms. Furthermore, the configuration method of the operating system kernel provided in the embodiments of the present application can also be executed by the terminal device and the server in cooperation. The terminal and the server can be directly or indirectly connected through wired or wireless communication, which is not limited in the present application. Therefore, the implementation subject executing the technical solutions of the present application is not limited in the embodiments of the present application.

[0036] Referring toFigure 1 FIG. 1 is a schematic diagram of a scenario of a configuration method of an operating system kernel according to an embodiment of the present application, comprising a computer device 100. The computer device 100 can be a terminal device or a server in various forms as described above. Specifically, the following embodiments take the computer device 100 as a server A for exemplary illustration.

[0037] The server A is configured to obtain hardware configuration information of the computing node and running features of the computing node in a running task state. It can be understood that obtaining the hardware configuration information is crucial for understanding the physical capabilities of the computing node, because the hardware configuration information determines the types and scales of tasks that the computing node can handle. Obtaining the running features helps to understand the actual pressure of the running tasks on the computing node and the bottlenecks or optimization points that the computing node may encounter when processing these running tasks.

[0038] The server A is further configured to input the hardware configuration information and the running features into a kernel parameter configuration model to obtain a parameter configuration value of an operating system kernel of the computing node. It can be understood that inputting the hardware configuration information and the running features into the pre-trained kernel parameter configuration model can ensure that the parameter configuration value output by the kernel parameter configuration model not only matches the hardware resources of the computing node, but also is a parameter configuration value optimized for the specific running task and the operating system kernel of the computing node, that is, the parameter configuration value is specially designed to maximize the computing performance of the computing node when processing the specific running task under the specific hardware resources.

[0039] The server A is further configured to update the operating system kernel of the computing node according to the generated parameter configuration value. This usually involves modifying the kernel configuration file, adjusting the kernel startup parameters, or sending configuration update instructions to the kernel through a specific management interface, etc.

[0040] The server A is further configured to deploy the updated operating system kernel to the computing node. This means that the new operating system kernel will take effect and the computing node will run according to the new operating system kernel. It can be understood that updating the operating system kernel of the computing node by the parameter configuration value output by the kernel parameter configuration model and deploying the updated operating system kernel to the computing node can achieve automatic and dynamic optimization of the operating kernel of the computing node, which reduces the workload caused by manual configuration and manual optimization, improves the adaptability of the operating system kernel to hardware resources and running task requirements, and thus improves the speed and accuracy of kernel configuration.

[0041] Next, the server A is taken as an execution subject to specifically introduce the configuration method of the operating system kernel provided by the embodiments of the present application:

[0042] Referring to Figure 2FIG. 1 is a flowchart of a method for configuring an operating system kernel according to an embodiment of the present application. As shown in FIG. 1, the method for configuring an operating system kernel comprises the following steps: Figure 2

[0043] S201: Obtain hardware configuration information of the computing node and running characteristics of the computing node in a running task state.

[0044] The purpose of obtaining the hardware configuration information of the computing node is to understand the physical capabilities of the computing node, including CPU model and quantity, memory size, storage device type (such as SSD or HDD) and its capacity, network interface type and speed, etc. In some specific implementation manners, system commands (such as lscpu command, free-m command, lsblk command, and ethtool command, etc.) can be used to query the hardware configuration information of the computing node, or the hardware configuration information can be obtained by reading the operating system configuration file of the computing node, or a special hardware detection tool (such as dmidecode, lshw) can be used to obtain the hardware configuration information, etc. It can be understood that the hardware configuration information determines the type and scale of the tasks that the computing node can handle. For example, a computing node with a hardware configuration of high-performance CPU and sufficient memory is more suitable for running compute-intensive tasks, and a computing node with a hardware configuration of fast storage and network interface is more suitable for running I / O-intensive tasks.

[0045] The purpose of obtaining the running characteristics of the computing node in the running task state is to understand the actual pressure of the running task on the computing node, and the bottlenecks or optimization points that the computing node may encounter when processing the running task, including CPU usage, memory usage, cache usage, network bandwidth, hard disk IO and read / write rate, etc. In some specific implementation manners, the running characteristics of the computing node in the running task state can be observed in real time through the monitoring tools (such as top command, htop command, vmstat command, and iostat command, etc.) set on the computing node, or the historical data in the log file can be parsed using log analysis tools (such as awk, sed, grep, etc.) to extract the running characteristics of the computing node in the running task state. It can be understood that obtaining the running characteristics helps to identify potential bottlenecks of the computing node, such as CPU overload, insufficient memory, disk I / O bottleneck, etc., which are crucial for generating parameter configuration values for optimizing specific running tasks.

[0046] Therefore, by understanding the actual requirements of the running task and the physical capabilities of the computing node, more reasonable and effective parameter configuration values can be generated, thereby improving the efficiency of kernel configuration.

[0047] ​S202: input the hardware configuration information and the running characteristics into the kernel parameter configuration model to obtain a parameter configuration value of an operating system kernel of the computing node.

[0048] The kernel parameter configuration model is a pre-trained machine learning model. By inputting the hardware configuration information and the running characteristics into the kernel parameter configuration model, the kernel parameter configuration model can output the parameter configuration value of the operating system kernel according to the input hardware configuration information and the running characteristics.

[0049] The parameter configuration value output by the kernel parameter configuration model is generated based on the physical capabilities of the computing node and the requirements of the specific running task. Therefore, the parameter configuration value helps to maximize the hardware performance of the computing node while minimizing the resource consumption and potential performance bottlenecks of the computing node. For example, the parameter configuration value can be a parameter configuration value related to the settings of the kernel scheduler (such as the type of the scheduler, the scheduling strategy, the thread priority, etc.), a parameter configuration value related to the memory management strategy (such as the memory allocation strategy, the page replacement algorithm, the memory compression and deduplication, etc.), a parameter configuration value related to the network stack configuration (such as the optimization of the network interface, the management and control of the network traffic, etc.), a parameter configuration value related to the file system (such as the type of the file system, the mounting option, the I / O scheduler, etc.), etc. The specific parameter configuration value is not limited in the present application.

[0050] S203: generating the operating system kernel of the computing node according to the parameter configuration value.

[0051] It should be noted that before generating the operating system kernel of the computing node according to the parameter configuration value, a pre-validation can be performed to ensure that the obtained parameter configuration value will not cause instability or performance degradation of the operating system.

[0052] In some specific implementations, a test environment (such as a virtual machine, a standby physical server, etc.) similar to the hardware configuration of the computing node can be created first, and the parameter configuration value obtained from the kernel parameter configuration model can be simulated in the test environment. Then, a performance monitoring tool (such as top, htop, vmstat, iostat, etc.) can be used to evaluate the running characteristics (such as execution speed, throughput, response time, etc.). If the running characteristics are better (such as the execution speed increases, the throughput increases, and the response time decreases), the operating system kernel of the computing node is generated according to the parameter configuration value.

[0053] In another specific implementation, the official documents of the operating system and the hardware can also be referred to to ensure that the parameter configuration value is compatible and within a safe range. If the parameter configuration value is compatible and within a safe range, the operating system kernel of the computing node is generated according to the parameter configuration value.

[0054] It should be noted that before generating the operating system kernel of the computing node according to the parameter configuration value, the original parameter configuration value of the current operating system kernel can also be backed up so as to roll back to the original state in case of problems, thereby avoiding the risk of unstable or degraded performance of the operating system.

[0055] It should be noted that for the specific method of generating the kernel configuration, there will be differences according to the type and version of the operating system. In one example, a brand new operating system kernel can be generated directly according to the parameter configuration value output by the kernel parameter configuration model. The advantage of this method is that the functions and characteristics of the kernel can be completely customized to meet specific needs. In another example, the un-updated parameter items of the operating system kernel can be determined first according to the original parameter configuration value of the operating system kernel and the parameter configuration value output by the kernel parameter configuration model, and then the operating system kernel of the computing node can be generated according to the original parameter configuration value of the un-updated parameter items and the parameter configuration value output by the kernel parameter configuration model. This method can update on the basis of preserving the existing kernel, reduce workload, avoid unnecessary changes, and ensure the stability of the system. For this, the present application is not limited.

[0056] S204: Deploy the generated operating system kernel to the computing node.

[0057] The generated operating system kernel file can be uploaded to the specified location of the computing node through network transmission methods such as Secure copy (SCP), Secure File Transfer Protocol (SFTP), or local copy method, thereby completing the deployment of the operating system kernel.

[0058] It should be noted that during the process of deploying the new kernel, choosing the right time is crucial. Because restarting the computing node to load the new operating system kernel will cause all running tasks on the computing node to be interrupted. In order to minimize the impact on the availability and performance of the operating system, the deployment is usually performed during the following time periods: first, choose to deploy during the gap between running tasks, i.e. during the time period when no running task is being executed, thereby ensuring that the ongoing running tasks will not be disturbed during the deployment process. This usually occurs during off-peak hours such as night or weekend when the operating system load is low and user activity is less. Second, choose to deploy during the low peak period of HPC cluster running, for example, when the real-time load of running tasks is lower than the preset load threshold. This can be determined by monitoring the load and resource usage of the HPC cluster. Deploying when the cluster load is low can minimize the impact on the overall performance of the cluster.

[0059] To sum up, the embodiment of the present application discloses a configuration method of an operating system kernel. The configuration method of the operating system kernel can ensure that the output parameter configuration value is matched with the hardware resources of the computing node and is a parameter configuration value for a specific running task by inputting the obtained hardware configuration information of the computing node and the running characteristics of the computing node in a running task state into a kernel parameter configuration model. Moreover, the method can realize automatic dynamic optimization of the operating kernel of the computing node by generating the operating system kernel of the computing node according to the parameter configuration value output by the kernel parameter configuration model and deploying the generated operating system kernel to the computing node, thereby reducing the workload caused by manual configuration and manual optimization, improving the adaptability of the operating system kernel to the hardware resources and the running task requirements, and thus improving the configuration rate and accuracy of the operating system kernel.

[0060] Referring to Figure 3 The figure is a flowchart of another configuration method of an operating system kernel provided by the embodiment of the present application. As shown in the configuration method of the operating system kernel, Figure 3 includes the following steps:

[0061] S301: Obtain the hardware configuration information of a plurality of computing nodes in an HPC cluster and the running characteristics of the plurality of computing nodes in a running task state.

[0062] It can be understood that by obtaining the hardware configuration information of a plurality of computing nodes in a computing cluster and the running characteristics of the plurality of computing nodes in a running task state, it can be ensured that the subsequently generated operating system kernel is not only matched with the hardware resources of the computing node, but also is an operating system kernel for a specific running task, thereby improving the adaptability of the operating system kernel to the hardware resources and the running task requirements, and further improving the kernel configuration and accuracy.

[0063] S302: Input the hardware configuration information and the running characteristics of the plurality of computing nodes into a kernel parameter configuration model to obtain the parameter configuration value of the operating system kernel of the plurality of computing nodes.

[0064] The kernel parameter configuration model is a pre-trained machine learning model. The hardware configuration information and the running characteristics of the plurality of computing nodes can be input into the kernel parameter configuration model, so that the kernel parameter configuration model outputs the parameter configuration value of the operating system kernel of the plurality of computing nodes according to the input hardware configuration information and running characteristics.

[0065] It should be noted that the kernel parameter configuration model can include a clustering sub-model and a plurality of parameter updating sub-models. The purpose of the clustering sub-model is to cluster the plurality of computing nodes into a plurality of grouping results with similar characteristics according to the hardware configuration information and the running characteristics, so that more personalized operating system kernel parameter setting suggestions can be provided for different hardware configuration information and running characteristics. The purpose of the parameter updating sub-model is to determine which parameters need to be adjusted according to the specific hardware configuration information and running characteristics, and to provide recommended parameter configuration values for these parameters.

[0066] Specifically, first, by inputting the hardware configuration information and the running characteristics of the plurality of computing nodes into the clustering sub-model of the kernel parameter configuration model, the clustering sub-model will classify the plurality of computing nodes according to a preset algorithm (such as K-means, DBSCAN, etc.), thereby determining the grouping results corresponding to the plurality of computing nodes respectively. Subsequently, by inputting the hardware configuration information and the running characteristics of the plurality of computing nodes into the parameter updating sub-models corresponding to the grouping results respectively, the operating system kernel parameter items to be updated of the plurality of computing nodes are determined, and parameter configuration values for the operating system kernel parameter items to be updated are determined, in order to achieve the purpose of optimizing performance.

[0067] Next, a training method of a kernel parameter configuration model provided by an embodiment of the present application is introduced. Referring to Figure 4 The figure is a flowchart of the training method of the kernel parameter configuration model provided by an embodiment of the present application. Specifically, the training method of the kernel parameter configuration model can be shown in the following steps:

[0068] A1: Obtain the hardware configuration information of the plurality of computing nodes in the historical time period, and the running characteristics of the plurality of computing nodes in the historical time period under the running task state.

[0069] A2: Cluster the plurality of computing nodes according to the hardware configuration information of the plurality of computing nodes in the historical time period and the running characteristics of the plurality of computing nodes in the historical time period under the running task state, to obtain a plurality of grouping results.

[0070] The purpose of cluster analysis (CA) is to find which computing nodes in the HPC cluster in the historical time period have similar characteristics (i.e., similar hardware configuration information and similar running characteristics).

[0071] In the configuration method of the operating system kernel provided by the embodiment of the present application, the computing nodes with similar hardware configuration information and running characteristics can be clustered by statistical methods to obtain a plurality of grouping results.

[0072] Specifically, first, the collected hardware configuration information and running features are preprocessed. In one example, outliers, missing values, and the like in the hardware configuration information and running features can be removed to ensure the integrity and accuracy of the hardware configuration information and running features. In another example, the hardware configuration information and running features can also be standardized to remove the influence of different feature dimensions, so that the hardware configuration information between different computing nodes and the running features between different computing nodes are more comparable.

[0073] Subsequently, the similarity of the preprocessed hardware configuration information and running features between different computing nodes is calculated. For example, the similarity of the preprocessed hardware configuration information between different computing nodes can be compared by calculating the cosine similarity, Euclidean distance, and the like, and the computing nodes with a similarity higher than a first similarity threshold are preliminarily clustered, and then the similarity of the preprocessed running features between the preliminarily clustered computing nodes is compared, and the computing nodes with a similarity higher than a second similarity threshold are secondarily clustered, thereby obtaining the final grouping result.

[0074] Finally, a suitable clustering algorithm (such as K-means, hierarchical clustering, DBSCAN, etc.) is selected. The computing nodes are clustered according to the calculated similarity. These clustering algorithms can automatically classify computing nodes with high similarity into a class, obtaining several grouping results.

[0075] It should be noted that after obtaining several grouping results, the grouping results can also be comprehensively evaluated to ensure that the grouping results meet the expectations. In one example, the compactness of each grouping result, i.e., the similarity of the computing nodes within the same grouping result, can be evaluated. If the similarity of the hardware configuration information of the computing nodes within the same grouping result is higher than the first similarity threshold, and the similarity of the running features of the computing nodes within the same grouping result is higher than the second similarity threshold, the subsequent A3 step is performed. In another example, the separation of each grouping result, i.e., the similarity of the computing nodes between different grouping results, can be evaluated. If the similarity between the hardware configuration information corresponding to multiple grouping results is lower than a third similarity threshold, and the similarity between the running features corresponding to multiple grouping results is lower than a fourth similarity threshold, the subsequent A3 step is performed. The specific evaluation method is not limited by the present application. Thus, through clustering analysis, all computing nodes in the HPC cluster will be divided into several grouping results with similar characteristics. The computing nodes within each grouping result have high similarity in hardware configuration and running features, while the computing nodes between different grouping results have great difference.

[0076] It can be understood that by clustering the computing nodes with similar features, the machine learning model can learn the features of the computing nodes within each grouping result more easily during subsequent model training, thereby improving the learning efficiency and accuracy of the model and providing strong support for customized optimization of the HPC cluster.

[0077] A3: For each grouping result, determine the parameter item to be updated by a decision tree model or a random forest model.

[0078] Determining the parameter item to be updated refers to identifying the influencing factors of the performance bottleneck. The decision tree (DT) model is used to analyze the hardware configuration information and running features of the internal computing nodes of the grouping result. By recursively selecting the optimal features for division, the decision tree model can gradually build a model that can accurately predict the performance bottleneck. Specifically, each internal computing node of the decision tree model represents a test of a feature attribute, each branch represents the output of the feature attribute in a certain value range, and each leaf node corresponds to a performance bottleneck category or a specific numerical value. During the training process, the decision tree model will continuously try to find the features that best distinguish different performance states, thereby building a decision path that accurately reflects the performance bottleneck.

[0079] The random forest (RF) model is used to further verify and refine the prediction results of the decision tree. Specifically, the random forest model will build multiple decision trees and integrate the prediction results of each decision tree through voting or averaging. This method can significantly reduce the overfitting risk of a single decision tree model and improve the generalization ability for unknown data, thereby more accurately identifying the main influencing factors of the performance bottleneck.

[0080] A4: For each grouping result, determine the parameter configuration value for the parameter item to be updated.

[0081] The parameter configuration value is used to optimize the running features of the running task, such as improving execution speed, throughput, and reducing response time, etc.

[0082] In some specific implementations, more suitable parameter configuration values of the operating system kernel can be generated for each grouping result directly according to the to-be-updated parameter item (e.g., hardware resource, software configuration, algorithm efficiency, etc.) of each grouping result. In an example, the to-be-updated parameter item corresponding to the grouping result 1 can be a memory management parameter item, and then the generated parameter configuration values for this grouping result can include increasing the size of the memory page, optimizing the memory fragmentation algorithm, or enabling a more efficient memory compression technology, etc. In another example, the to-be-updated parameter item corresponding to the grouping result 2 can be a CPU scheduling parameter item. Then the generated parameter configuration values for this grouping result can include optimizing the CPU scheduling strategy, such as adjusting the priority scheduling algorithm of the process, enabling multi-core parallel processing, or optimizing the power management strategy of the CPU, etc.

[0083] A5: training a machine learning model according to the hardware configuration information of the plurality of computing nodes in the historical time period, the running features of the plurality of computing nodes in the running task state in the historical time period, and the parameter configuration values of the operating system kernel of the plurality of computing nodes in the historical time period, to obtain a kernel parameter configuration model.

[0084] Next, taking an HPC cluster including 1000 computing nodes, of which 500 computing nodes mainly run memory-intensive tasks (type A) and 500 computing nodes mainly run CPU-intensive tasks (type B) as an embodiment for explanation and illustration:

[0085] First, collect the hardware configuration information (such as CPU model, memory size, etc.) and running features (such as execution speed, response time, throughput, etc.) of all computing nodes. Second, cluster the computing nodes with similar hardware configuration information and running features, i.e., distinguish type A and type B, to obtain several grouping results of type A and several grouping results of type B. The purpose of this is to group nodes with similar characteristics for more efficient management and optimization. Next, for the several grouping results of type A and the several grouping results of type B, determine the to-be-updated parameter item and determine the parameter configuration value for the to-be-updated parameter item through the decision tree model or the random forest model respectively. These parameter configuration values aim to optimize the performance of the computing nodes to adapt to their specific hardware and running features. Finally, according to the hardware configuration information, running features and generated parameter configuration values of type A and type B nodes, respectively train a machine learning model. For type A nodes, a kernel parameter configuration model A is trained; for type B nodes, a kernel parameter configuration model B is trained. The kernel parameter configuration model A and the kernel parameter configuration model B can be used in the future to predict and optimize newly added computing nodes of the HPC cluster, or dynamically adjust the kernel configuration according to the changes of the current computing nodes, to improve the overall performance and resource utilization.

[0086] S303: Generate corresponding operating system kernels for the plurality of computing nodes according to the parameter configuration values of the plurality of computing nodes.

[0087] Generate corresponding operating system kernels for the plurality of computing nodes according to the parameter configuration values of the plurality of computing nodes, and ensure that the generated operating system kernels can provide optimal performance for specific hardware configuration information and running characteristics.

[0088] It should be noted that the specific method of generating kernel configuration may vary depending on the type and version of the operating system. In one example, a brand new operating system kernel can be generated directly according to the parameter configuration values output by the kernel parameter configuration model. This method has the advantage of being able to completely customize the functions and characteristics of the kernel to meet specific needs. In another example, the operating system kernel of the computing node can be generated according to the original parameter configuration values of the operating system kernel and the parameter configuration values output by the kernel parameter configuration model, and then according to the original parameter configuration values of the non-updated parameter items and the parameter configuration values output by the kernel parameter configuration model. This method can update the existing kernel while reducing the workload and avoiding unnecessary changes, while ensuring the stability of the system. For this purpose, the present application is not limited.

[0089] S304: Deploy the generated plurality of operating system kernels to the corresponding plurality of computing nodes, and restart the plurality of computing nodes to obtain the plurality of computing nodes after deploying the operating system kernels.

[0090] In some specific implementations, the generated plurality of operating system kernels needs to be deployed to the corresponding plurality of computing nodes by means of network transmission file, installation using automation script or management tool, etc. And since the operating system kernel is the core of the OS, any change to the operating system kernel needs to be effective through restarting the operating system. This means that after deploying the new operating system kernel, the computing node needs to be restarted so that the new operating system kernel can take control of the operating system and start running, i.e. obtaining the computing node after deploying the operating system kernel.

[0091] It should be noted that in the process of deploying the new operating system kernel, it is crucial to choose the right time. Because restarting the computing node will cause all running tasks on the computing node to be interrupted. Therefore, it is usually done during the task gap (i.e. the period when no running task is running) or during the low peak period of cluster running (i.e. the period when the real-time load of running task is lower than the preset load threshold), in order to minimize the impact on the availability and performance of the OS.

[0092] S305: Monitor the running characteristics of each computing node after deploying the operating system kernel of the HPC cluster.

[0093] The monitoring of the running features of each deployed operating system kernel behind the computing node includes monitoring of execution speed, response time, and throughput of each deployed operating system kernel behind the computing node.

[0094] S306: updating the model parameters of the kernel parameter configuration model according to the running features of each deployed operating system kernel behind the computing node.

[0095] First, the running features of each deployed operating system kernel behind the computing node are analyzed in depth. If the computing node before the deployment of the operating system kernel and the computing node behind the deployment of the operating system kernel run the same running task, the running feature of the computing node behind the deployment of the operating system kernel in the running task state is better than that of the computing node before the deployment of the operating system kernel in the running task state (i.e. any one or more of the following conditions is met: the execution speed corresponding to the computing node behind the deployment of the operating system kernel is greater than the execution speed corresponding to the computing node before the deployment of the operating system kernel, or the throughput corresponding to the computing node behind the deployment of the operating system kernel is greater than the throughput corresponding to the computing node before the deployment of the operating system kernel, or the response time corresponding to the computing node behind the deployment of the operating system kernel is less than the response time corresponding to the computing node before the deployment of the operating system kernel). According to the hardware configuration information, running features and parameter configuration values corresponding to the operating system kernel, the model parameters of the kernel parameter configuration model are updated, so that the kernel parameter configuration model can better learn which parameters can bring the improvement of running performance.

[0096] It can be understood that the traditional configuration method of the operating system kernel is usually a one-time setting when the OS is installed or configured. This method may provide good performance in the initial stage, but it is difficult to adapt to the dynamic changes of the running environment and task requirements. With the upgrading of hardware, the updating of software version and the change of work load, these static settings may gradually become unsuitable, and even may become a performance bottleneck. In contrast, the configuration method of the operating system kernel provided by the present application can dynamically adjust according to the hardware configuration information and running features of the HPC cluster. By monitoring the changes of the hardware configuration information and running features of the HPC cluster in real time, performance problems can be found and solved in time, and performance bottlenecks can be avoided, so that the operating system can better adapt to various application scenarios.

[0097] Further, the method for configuring an operating system kernel can quickly identify the physical capabilities and performance requirements of the new computing node when the new computing node joins the HPC cluster, and automatically adjust the OS kernel configuration according to the physical capabilities and performance requirements of the new computing node, to ensure that the new computing node can be efficiently and stably integrated into the cluster. This means that the new computing node does not need to go through a tedious manual configuration and testing process, and can be immediately put into use, thereby greatly shortening the time cost of cluster expansion.

[0098] In summary, the present application provides a method for configuring an operating system kernel, whereby the method for configuring an operating system kernel can ensure that the output parameter configuration value is matched with the hardware resources of the computing node, and is a parameter configuration value for a specific running task, by obtaining the hardware configuration information of the computing node, obtaining the running characteristics of the computing node in the running task state, and inputting the hardware configuration information and the running characteristics into the kernel parameter configuration model. And the operating system kernel of the computing node is updated by the parameter configuration value output by the kernel parameter configuration model, and the updated operating system kernel is deployed to the computing node, which can realize the automatic and dynamic optimization of the operating kernel of the computing node. At the same time, the workload caused by traditional manual configuration and manual optimization is reduced, the adaptability of the operating system kernel to hardware resources and running task requirements is improved, and the rate and accuracy of kernel configuration are improved. Further, since the parameter configuration value is dynamically generated according to the hardware configuration information of the computing node and the running characteristics of the computing node in the running task state, it can better adapt to the changes of hardware resources and running task requirements. This adaptability ensures that the computing node can maintain efficient operation under different loads and task requirements. And the updated operating system kernel can better utilize the hardware resources of the computing node, thereby improving the computing efficiency while reducing the risk of computing node crash or performance degradation caused by insufficient hardware resources.

[0099] Next, a model training method provided by an embodiment of the present application is introduced. Referring to Figure 5 , which is a flowchart of a model training method provided by an embodiment of the present application. Specifically, the model training method can be shown in the following steps:

[0100] S501: Obtain a sample data set, which includes hardware configuration information of a plurality of computing nodes in a historical time period, running characteristics of the plurality of computing nodes in the running task state in the historical time period, and parameter configuration values of the operating system kernel of the plurality of computing nodes in the historical time period.

[0101] In some specific implementations, the obtaining the sample data set comprises: obtaining hardware configuration information of a plurality of computing nodes in a historical time period and running features of the plurality of computing nodes in the running task state in the historical time period; clustering the plurality of computing nodes according to the hardware configuration information of the plurality of computing nodes in the historical time period and the running features of the plurality of computing nodes in the running task state in the historical time period to obtain a plurality of grouping results, wherein the similarity of the hardware configuration information of the computing nodes in each grouping result is higher than a first similarity threshold, and the similarity of the running features of the computing nodes in each grouping result is higher than a second similarity threshold; and determining, for each grouping result, a parameter item to be updated and determining a parameter configuration value for the parameter item to be updated.

[0102] S502: training the machine learning model according to the sample data set to obtain the kernel parameter configuration model.

[0103] In summary, the embodiment of the present application provides a model training method. The method trains a machine learning model by using the hardware configuration information of a plurality of computing nodes in a historical time period, the running features of the plurality of computing nodes in the running task state in the historical time period, and the parameter configuration values of the operating system kernel of the plurality of computing nodes in the historical time period as a sample data set. The method can ensure that the parameter configuration values output by the kernel parameter configuration model in actual application are matched with the hardware resources of the computing nodes and are parameter configuration values for specific running tasks. Therefore, the workload caused by manual configuration and manual optimization is reduced, the adaptability of the operating system kernel to hardware resources and running task requirements is improved, and the configuration rate and accuracy of the operating system kernel are improved.

[0104] Based on the operating system kernel configuration method provided in the foregoing embodiments, the present application also provides an operating system kernel configuration device. The operating system kernel configuration device provided in the embodiments of the present application is described in detail as follows:

[0105] Referring to Figure 6 FIG. 1 is a schematic diagram of an operating system kernel configuration device provided in an embodiment of the present application. The operating system kernel configuration device 600 comprises an information obtaining module 601, a parameter obtaining module 602, a kernel updating module 603, and a kernel deploying module 604.

[0106] The information obtaining module 601 is configured to obtain the hardware configuration information of the computing nodes and the running features of the computing nodes in the running task state.

[0107] The parameter obtaining module 602 is configured to input the hardware configuration information and the running features into the kernel parameter configuration model to obtain the parameter configuration values of the operating system kernel of the computing nodes.

[0108] The kernel updating module 603 is configured to generate the operating system kernel of the computing node according to the parameter configuration value.

[0109] The kernel deployment module 604 is configured to deploy the generated operating system kernel to the computing node.

[0110] In some specific implementation manners, the kernel parameter configuration model comprises a parameter updating sub-model, and the parameter obtaining module 602 is specifically configured to: input the hardware configuration information and the running feature into the parameter updating sub-model of the kernel parameter configuration model, to determine the to-be-updated parameter item of the operating system kernel of the computing node, and determine the parameter configuration value for the to-be-updated parameter item.

[0111] In some specific implementation manners, the kernel parameter configuration model comprises a clustering sub-model and a plurality of parameter updating sub-models, and the parameter obtaining module 602 is specifically configured to: input the hardware configuration information and the running feature into the clustering sub-model of the kernel parameter configuration model, to determine a grouping result corresponding to the hardware configuration information and the running feature; and input the hardware configuration information and the running feature into the parameter updating sub-model corresponding to the grouping result, to determine the to-be-updated parameter item of the operating system kernel of the computing node.

[0112] In some specific implementation manners, the configuration apparatus 600 of the operating system kernel further comprises a parameter updating module, which is configured to: if the running feature of the computing node in the running task state after the operating system kernel is deployed is superior to the running feature of the computing node in the running task state before the operating system kernel is deployed, update the model parameter of the kernel parameter configuration model according to the hardware configuration information, the running feature and the parameter configuration value corresponding to the operating system kernel.

[0113] In some specific implementation manners, the kernel updating module 603 is specifically configured to: determine the non-updated parameter item of the operating system kernel according to the original parameter configuration value of the operating system kernel and the parameter configuration value output by the kernel parameter configuration model; and generate the operating system kernel of the computing node according to the original parameter configuration value of the non-updated parameter item and the parameter configuration value output by the kernel parameter configuration model.

[0114] In some specific implementation manners, the kernel deployment module 604 is specifically configured to: deploy the generated operating system kernel to the computing node when the running task is not running, or when the running feature represents that the real-time load of the running task is lower than a preset load threshold.

[0115] In some specific implementation manners, the information obtaining module 601 is specifically configured to: obtain hardware configuration information of a plurality of computing nodes in a computing cluster, and running features of the plurality of computing nodes in a running task state; the parameter obtaining module 602 is specifically configured to: determine a grouping result corresponding to each of the plurality of computing nodes by inputting the hardware configuration information and the running features of the plurality of computing nodes into a clustering submodel of a kernel parameter configuration model; determine a to-be-updated parameter item of an operating system kernel of the plurality of computing nodes, and determine a parameter configuration value for the to-be-updated parameter item by inputting the hardware configuration information and the running features of the plurality of computing nodes into a parameter updating submodel corresponding to the grouping result corresponding to each of the plurality of computing nodes; the kernel updating module 603 is specifically configured to: generate a corresponding operating system kernel for each of the plurality of computing nodes according to the parameter configuration value of each of the plurality of computing nodes; and the kernel deploying module 604 is specifically configured to: deploy the corresponding operating system kernel generated for each of the plurality of computing nodes to the corresponding computing node.

[0116] In summary, the application provides a configuration device of an operating system kernel, so that the configuration device of the operating system kernel can ensure that the output parameter configuration value is matched with the hardware resources of the computing node and is a parameter configuration value for a specific running task by obtaining the hardware configuration information of the computing node and the running features of the computing node in a running task state, and inputting the hardware configuration information and the running features into a kernel parameter configuration model. Moreover, the operating system kernel of the computing node is updated by the parameter configuration value output by the kernel parameter configuration model, and the updated operating system kernel is deployed to the computing node, which can realize automatic and dynamic optimization of the operating kernel of the computing node, reduce the workload caused by traditional manual configuration and manual optimization, improve the adaptability of the operating system kernel to hardware resources and running task requirements, and thus improve the rate and accuracy of kernel configuration. Furthermore, since the parameter configuration value is dynamically generated according to the hardware configuration information of the computing node and the running features of the computing node in a running task state, the parameter configuration value can better adapt to changes in hardware resources and running task requirements, and this adaptability ensures that the computing node can maintain efficient operation under different loads and task requirements. Moreover, the updated operating system kernel can better utilize the hardware resources of the computing node, thereby improving the computing efficiency and reducing the risk of computing node crash or performance degradation caused by insufficient hardware resources.

[0117] Based on the model training method provided in the foregoing embodiments, the application also provides a model training device. The model training device provided in the embodiments of the application is specifically introduced as follows:

[0118] Referring to Figure 7FIG. 7 is a schematic diagram of a model training device according to an embodiment of the present application. The model training device 700 includes an obtaining module 701 and a training module 702.

[0119] The obtaining module 701 is configured to obtain a sample data set, the sample data set including hardware configuration information of a plurality of computing nodes in a historical time period, running features of the plurality of computing nodes in a running task state in the historical time period, and parameter configuration values of an operating system kernel of the plurality of computing nodes in the historical time period.

[0120] The training module 702 is configured to train a machine learning model according to the sample data set to obtain a kernel parameter configuration model.

[0121] In some specific implementations, the obtaining module 701 is specifically configured to: obtain the hardware configuration information of the plurality of computing nodes in the historical time period and the running features of the plurality of computing nodes in the running task state in the historical time period; cluster the plurality of computing nodes according to the hardware configuration information of the plurality of computing nodes in the historical time period and the running features of the plurality of computing nodes in the running task state in the historical time period to obtain a plurality of grouping results, wherein the similarity of the hardware configuration information of the computing nodes in each grouping result is higher than a first similarity threshold, and the similarity of the running features of the computing nodes in each grouping result is higher than a second similarity threshold; and for each grouping result, determine a to-be-updated parameter item and determine a parameter configuration value for the to-be-updated parameter item.

[0122] To sum up, the embodiment of the present application provides a model training device, which trains a machine learning model by using the hardware configuration information of a plurality of computing nodes in a historical time period, running features of the plurality of computing nodes in a running task state in the historical time period, and parameter configuration values of an operating system kernel of the plurality of computing nodes in the historical time period as a sample data set, so as to ensure that the parameter configuration values output in the actual application of the kernel parameter configuration model trained are matched with the hardware resources of the computing nodes and are parameter configuration values for specific running tasks, thereby reducing the workload caused by manual configuration and manual optimization, improving the adaptability of the operating system kernel to hardware resources and running task requirements, and further improving the configuration rate and accuracy of the operating system kernel.

[0123] The embodiment of the present application also provides an electronic device, which includes a memory for storing computer programs or computer instructions, and a processor for executing the computer programs or computer instructions stored in the memory, so that the electronic device executes the configuration method of the operating system kernel or the model training method in any of the above embodiments.

[0124] The embodiment of the present application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program or instructions. When the computer program or instructions are executed, the configuration method of the operating system kernel or the model training method in any of the above embodiments is implemented.

[0125] The embodiment of the present application further provides a computer program product. When the computer program product is executed on a server, the server implements the configuration method of the operating system kernel or the model training method in any of the above embodiments.

[0126] It should be understood that, in the embodiments of the present application, “at least one” refers to one or more, and “multiple” refers to two or more. “And / or” is used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, “A and / or B” can represent three cases of only A, only B, and A and B existing at the same time, wherein A and B can be singular or plural. The character “ / ” generally represents that the associated objects before and after it are in an “or” relationship. “At least one of the following” or the like means any combination of these items, including single item or any combination of multiple items. For example, at least one of a, b or c can represent a, b, c, “a and b”, “a and c”, “b and c”, or “a and b and c”, wherein a, b, and c can be single or multiple.

[0127] It should be further noted that, in this paper, relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the terms “include”, “contain” or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement “including a…” does not exclude the presence of another identical element in the process, method, article or device including the element.

[0128] The steps of the method or algorithm described in combination with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0129] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A configuration method of an operating system kernel, characterized by, The method comprises: obtaining hardware configuration information of a computing node and running features of the computing node in a running task state; the running features represent actual pressure of the running task on the computing node and bottlenecks or optimization points encountered by the computing node when processing the running task; inputting the hardware configuration information and the running features into a kernel parameter configuration model to obtain a parameter configuration value of an operating system kernel of the computing node; generating an operating system kernel of the computing node according to the parameter configuration value; deploying the generated operating system kernel to the computing node; the kernel parameter configuration model comprises a parameter update sub-model, and the inputting the hardware configuration information and the running features into the kernel parameter configuration model to obtain the parameter configuration value of the operating system kernel of the computing node comprises: determining a to-be-updated parameter item of the operating system kernel of the computing node by inputting the hardware configuration information and the running features into the parameter update sub-model of the kernel parameter configuration model, and determining a parameter configuration value for the to-be-updated parameter item.

2. The method of claim 1, wherein, the kernel parameter configuration model comprises a clustering sub-model and a plurality of parameter update sub-models, and the determining the to-be-updated parameter item of the operating system kernel of the computing node by inputting the hardware configuration information and the running features into the parameter update sub-model of the kernel parameter configuration model comprises: determining a grouping result corresponding to the hardware configuration information and the running features by inputting the hardware configuration information and the running features into the clustering sub-model of the kernel parameter configuration model; determining the to-be-updated parameter item of the operating system kernel of the computing node by inputting the hardware configuration information and the running features into the parameter update sub-model corresponding to the grouping result.

3. The method of claim 1, wherein, The method further comprises: if the running features of the computing node in the running task state after the operating system kernel is deployed are better than the running features of the computing node in the running task state before the operating system kernel is deployed, updating model parameters of the kernel parameter configuration model according to the hardware configuration information, the running features and the parameter configuration value corresponding to the operating system kernel.

4. The method of claim 1, wherein, the generating the operating system kernel of the computing node according to the parameter configuration value comprises: determining an un-updated parameter item of the operating system kernel according to an original parameter configuration value of the operating system kernel and a parameter configuration value output by the kernel parameter configuration model; generating the operating system kernel of the computing node according to the original parameter configuration value of the un-updated parameter item and the parameter configuration value output by the kernel parameter configuration model.

5. The method of claim 1, wherein, the deploying the generated operating system kernel to the computing node comprises: deploying the generated operating system kernel to the computing node when the running task is not running, or when the running features represent that real-time load of the running task is lower than a preset load threshold.

6. The method of claim 1, wherein, the obtaining the hardware configuration information of the computing node and the running features of the computing node in the running task state comprises: Obtaining hardware configuration information of a plurality of computing nodes in a computing cluster, and running features of the plurality of computing nodes in a running task state; The inputting the hardware configuration information and the running features into a kernel parameter configuration model to obtain parameter configuration values of an operating system kernel of the computing nodes comprises: Determine the grouping results corresponding to the plurality of computing nodes respectively by inputting the hardware configuration information and the running features of the plurality of computing nodes into clustering sub-models of the kernel parameter configuration model; Determine the to-be-updated parameter items of the operating system kernel of the plurality of computing nodes and determine parameter configuration values for the to-be-updated parameter items by inputting the hardware configuration information and the running features of the plurality of computing nodes into parameter updating sub-models corresponding to the grouping results corresponding to the plurality of computing nodes respectively; The generating the operating system kernel of the computing nodes according to the parameter configuration values comprises: Generate corresponding operating system kernels for the plurality of computing nodes according to the parameter configuration values of the plurality of computing nodes respectively; Deploy the generated operating system kernels to the computing nodes, comprising: Deploy the corresponding operating system kernels generated for the plurality of computing nodes to the corresponding plurality of computing nodes respectively.

7. An electronic device, comprising: The electronic device comprises: A memory for storing computer programs or computer instructions; A processor for executing the computer programs or computer instructions stored in the memory, so that the electronic device executes the method of any one of claims 1 to 6.

Images