Equipment configuration regression method and related device

Abstract

The embodiment of the invention discloses an equipment configuration regression method and a related device. The method comprises the following steps: firstly, acquiring basic container configuration information, basic system configuration information and reference network configuration information; then, container configuration regression processing is carried out on a target container in the target computing equipment based on the basic container configuration information, and system configuration regression processing is carried out on system parameter configuration of the target computing equipment based on the basic system configuration information; and performing network configuration regression processing on the network parameter configuration of the target computing device based on the reference network configuration information. Thus, before the target computing equipment performs computing power sharing circulation, container environment configuration, system parameter configuration and network parameter configuration of the target computing equipment are all returned to corresponding initial states, so that interference of basic environment configuration of an AI training task executed on the target computing equipment previously on an AI training task executed later is avoided; the related configuration of the target computing device can be restored more efficiently, and the loss of computing power resources is reduced.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a device configuration regression method and related apparatus. Background Art

[0002] With the continuous development of artificial intelligence (AI) technology, AI models are widely used in many businesses; and as the business requirements for related services increase, AI large models with large parameters and excellent performance are gradually emerging. Since AI large models have a large number of training parameters and a large amount of data exchanged during training, it is necessary to use computing devices with large computing power to perform such training tasks. Such computing devices generally have high single-machine configurations and high usage costs.

[0003] In order to make full use of the computing power resources of the above-mentioned computing devices and avoid cost waste due to idle computing power, it is often necessary to realize the sharing and circulation of computing power resources on such computing devices, that is, after using the computing device to complete the training task of a certain AI model, continue to use the computing power resources of the computing device to perform training tasks for other AI models.

[0004] However, the basic environment configurations relied on when training different AI models may be different, and different basic environment configurations may interfere with the normal training of other AI models. Therefore, how to avoid the basic environment configuration used in the previous AI model training from affecting the AI ​​model training tasks to be performed later when realizing the shared circulation of computing power resources of computing devices is an urgent problem to be solved. Summary of the invention

[0005] The embodiments of the present application provide a device configuration regression method and related apparatus, which can avoid the influence of the basic environment configuration used in the previous AI model training on the AI ​​model training tasks to be performed later when realizing the shared circulation of the computing power resources of the computing device.

[0006] A first aspect of the present application provides a device configuration regression method, the method comprising:

[0007] Obtaining basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters;

[0008] For the target computing device to share computing power resources, compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, and perform container configuration regression processing on the target container according to the comparison result;

[0009] Based on the basic system configuration information, performing system configuration regression processing on the system parameter configuration of the target computing device;

[0010] Based on the reference network configuration information, a network configuration regression process is performed on the network parameter configuration of the target computing device.

[0011] A second aspect of the present application provides a device configuration regression apparatus, the apparatus comprising:

[0012] A basic information acquisition module, used to acquire basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters;

[0013] A container regression module is used to compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information for the target computing device to share computing resources, and perform container configuration regression processing on the target container according to the comparison result;

[0014] A system parameter regression module, configured to perform system configuration regression processing on the system parameter configuration of the target computing device based on the basic system configuration information;

[0015] A network parameter regression module is used to perform network configuration regression processing on the network parameter configuration of the target computing device based on the reference network configuration information.

[0016] A third aspect of the present application provides a computer device, the device comprising a processor and a memory:

[0017] The memory is used to store computer programs;

[0018] The processor is used to execute the steps of the device configuration regression method as described in the first aspect according to the computer program.

[0019] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store a computer program, and the computer program is used to execute the steps of the device configuration regression method described in the first aspect above.

[0020] In a fifth aspect, the present application provides a computer program product or a computer program, the computer program product or the computer program includes a computer instruction, the computer instruction is stored in a computer-readable storage medium. A processor of a computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device performs the steps of the device configuration regression method described in the first aspect above.

[0021] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0022] In the device configuration regression method provided in the embodiment of the present application, before the target computing device performs computing power sharing and circulation, the target computing device is subjected to configuration regression processing based on the corresponding configuration information. Specifically, in this method, the basic container configuration information, the basic system configuration information and the reference network configuration information can be obtained first, wherein the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined on the basis of the configuration information of the original network parameters; then, based on the basic container configuration information, the target container in the target computing device is subjected to container configuration regression processing, based on the basic system configuration information, the system configuration parameter configuration of the target computing device is subjected to system configuration regression processing, and based on the reference network configuration information, the network configuration parameter configuration of the target computing device is subjected to network configuration regression processing. In this way, before the target computing device carries out computing power sharing and circulation, the container environment configuration, system parameter configuration and network parameter configuration of the target computing device are all returned to the corresponding initial states based on the above-mentioned basic container configuration information, basic system configuration information and reference network configuration information, so as to avoid the basic environment configuration of the AI ​​training task previously executed on the target computing device from interfering with the AI ​​training task executed thereafter; in addition, by performing configuration regression processing on the target computing device based on the corresponding configuration information, the relevant configuration of the target computing device can be restored more efficiently, reducing the loss of computing power resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of an application scenario of a device configuration regression method provided in an embodiment of the present application;

[0024] Figure 2 A method flow chart of a device configuration regression method provided in an embodiment of the present application;

[0025] Figure 3 A flow chart of a network configuration regression process for a network parameter configuration of a target computing device provided in an embodiment of the present application;

[0026] Figure 4A schematic diagram of a process for cleaning residual data provided in an embodiment of the present application;

[0027] Figure 5 An overall architecture diagram of a device configuration regression method provided in an embodiment of the present application;

[0028] Figure 6 A flowchart of a configuration regression method for a computing device provided in an embodiment of the present application;

[0029] Figure 7 A schematic diagram of the structure of a device configuration regression device provided in an embodiment of the present application;

[0030] Figure 8 is a schematic diagram of the structure of a terminal device provided in an embodiment of the present application;

[0031] Fig. 9 A schematic diagram of the structure of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] The terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products or apparatuses.

[0034] In the related art, in order to avoid the impact of the basic environment configuration used in the previous AI model training on the AI ​​model training tasks to be performed later when realizing the shared circulation of computing power resources of computing devices, the computing device is usually re-installed, that is, the operating system is reinstalled for the computing device, so that the container environment configuration, system parameter configuration and network parameter configuration can be initialized; however, reinstalling the operating system for the computing device takes time at the granularity of hours, and the computing power resources during the installation of the operating system for the computing device are also wasted. The longer it takes, the more computing power resources are wasted.

[0035] In view of this, an embodiment of the present application provides a device configuration regression method, in which basic container configuration information, basic system configuration information and reference network configuration information are obtained; wherein the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined on the basis of the configuration information of the original network parameters; then, for the target computing device of the computing power resources to be shared, the target container configuration information corresponding to the target container in the target computing device and the basic container configuration information are compared, and according to the comparison result, the container configuration regression processing is performed on the target container; and based on the basic system configuration information, the system configuration regression processing is performed on the system parameter configuration of the target computing device; based on the reference network configuration information, the network configuration regression processing is performed on the network parameter configuration of the target computing device.

[0036] In the above-mentioned device configuration regression method, before the target computing device performs computing power sharing flow, the target computing device is subjected to configuration regression processing based on the corresponding configuration information. Specifically, before the target computing device performs computing power sharing flow, based on the above-mentioned basic container configuration information, basic system configuration information and reference network configuration information, the container environment configuration, system parameter configuration and network parameter configuration of the target computing device are all returned to the corresponding initial state, thereby avoiding the basic environment configuration of the AI ​​training task previously executed on the target computing device from interfering with the AI ​​training task executed thereafter. In addition, compared with the method of reinstalling the operating system for the computing device in the related art, the embodiment of the present application performs configuration regression processing on the target computing device based on the corresponding configuration information, which can more efficiently restore the relevant configuration of the target computing device and reduce the loss of computing power resources.

[0037] The device configuration regression method provided in the embodiment of the present application can be executed by a computer device with data processing capabilities, and the computer device can be a server or a terminal device. Among them, the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, a car terminal, etc., but is not limited to this. The terminal and the server can be directly or indirectly connected via wired or wireless communication, and this application is not limited here.

[0038] In order to facilitate understanding of the device configuration regression method provided in the embodiment of the present application, the application scenario of the device configuration regression method is exemplarily introduced below by taking the server as the execution subject of the device configuration regression method provided in the embodiment of the present application as an example.

[0039] See also Figure 1 , Figure 1 A schematic diagram of an application scenario of a device configuration regression method provided in an embodiment of the present application. Figure 1 As shown, the application scenario includes a server 101, which can be any computing device used to perform AI model training tasks. It is used to execute the device configuration regression method provided in the embodiment of the present application to regress its own related environmental configuration before sharing the computing power resources.

[0040] Before the server 101 performs the shared circulation of computing power resources, it can first obtain the basic container configuration information, the basic system configuration information and the reference network configuration information; wherein, the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined on the basis of the configuration information of the original network parameters. Then, based on the basic container configuration information, the server 101 can compare the target container configuration information corresponding to its own target container with the basic container configuration information, and perform container configuration regression processing on the target container according to the comparison result. Next, the server 101 can perform system configuration regression processing on its own system parameter configuration based on the basic system configuration information. Furthermore, the server 101 can perform network configuration regression processing on its own network parameter configuration based on the reference network configuration information.

[0041] In this way, before server 101 shares and circulates computing resources, the container environment configuration, system parameter configuration and network parameter configuration of server 101 are all returned to the corresponding initial states based on the above-mentioned basic container configuration information, basic system configuration information and reference network configuration information, thereby avoiding the basic environment configuration of the AI ​​training task previously executed on server 101 from interfering with the AI ​​training task executed thereafter; in addition, by performing configuration regression processing on server 101 based on the corresponding configuration information, the relevant configuration of server 101 can be restored more efficiently, thereby reducing the loss of computing resources.

[0042] The device configuration regression method provided by the present application is introduced below in conjunction with a method embodiment.

[0043] See also Figure 2 , Figure 2A method flow chart of a device configuration regression method provided in an embodiment of the present application, the method can be executed by a target computing device, it should be understood that the target computing device can be any computing device in a computing system that can perform a model training task, the method specifically includes the following steps:

[0044] S201. Obtain basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters.

[0045] In an embodiment of the present application, before the target computing device performs power sharing and circulation, the target computing device first obtains basic container configuration information, basic system configuration information and reference network configuration information. It is understandable that the target computing device obtains the basic container configuration information and the basic system configuration information directly from the initialized computing device. It should be understood that the initial basic container configuration information and the basic system configuration information of the target computing device can be obtained directly on the initialized device; or the basic container configuration information and the basic system configuration information of the target computing device can be autonomously configured according to empirical parameters.

[0046] The reference network configuration information is determined based on the original network parameter configuration information. The reference network configuration information may include the original network parameter configuration information and may also include configuration information having an optimization effect on network performance.

[0047] As the basis for determining the reference network configuration information, the original network parameter configuration information (also referred to as basic network configuration information) can be directly obtained on the initialized computing device, or the original network parameter configuration information of the target computing device can be autonomously configured based on empirical parameters, and then the reference network configuration information is determined based on the original network parameter configuration information. S202. For the target computing device of the computing power resources to be shared, the target container configuration information corresponding to the target container in the target computing device is compared with the basic container configuration information, and the container configuration regression processing is performed on the target container according to the comparison result.

[0048] Any container environment on the target computing device can be used as the target container; the configuration information of the target container is the target container configuration information. After the target computing device obtains the basic container configuration information, the target container configuration information corresponding to the target container in the target computing device can be compared with the basic container configuration information to obtain a comparison result.

[0049] According to the comparison result, container configuration regression processing is performed on the target container. Specifically, if the comparison result is that the target container configuration information corresponding to the target container is the same as the basic container configuration information, there is no need to adjust the configuration information of the target container; if the comparison result is that the target container configuration information corresponding to the target container is different from the basic container configuration information, it is necessary to adjust the configuration information of the target container; specifically, the target computing device can perform an automated difference correction operation to achieve container configuration regression processing on the target container.

[0050] S203: Based on the basic system configuration information, perform system configuration regression processing on the system parameter configuration of the target computing device.

[0051] Specifically, after the target computing device obtains the basic system configuration information, it can perform system configuration regression processing on the system parameter configuration of the target computing device based on the basic system configuration information. Specifically, the basic system configuration information can be used to overwrite the system parameter configuration information of the target computing device, thereby restoring the system configuration information of the target computing device to the original system parameter configuration information.

[0052] S204. Based on the reference network configuration information, perform network configuration regression processing on the network parameter configuration of the target computing device.

[0053] As introduced above, the reference network configuration information is determined based on the original network parameter configuration information. The reference network configuration information may include the original network parameter configuration information of the target computing device, and may also include configuration items that have an optimization effect on network performance.

[0054] Specifically, after the target computing device first obtains the reference network configuration information, it performs network configuration information regression processing on the network configuration information of the target computing device based on the reference network configuration information. Specifically, the reference network configuration information can be used to overwrite the configuration information of the network parameters in the target computing device, thereby restoring the network configuration information of the target computing device to the original network configuration information.

[0055] One thing that needs to be explained is that the execution order of step S202 to step S204 cannot be disordered, that is, when performing configuration regression processing on the target computing device, the container configuration regression processing must be performed first, and then the system configuration regression processing, and finally the network configuration regression processing. Otherwise, it is easy for the regressed configuration information to be overwritten again. For example, if the system parameter configuration is reset first and then the container configuration information is reset, then when the container configuration information is corrected, the system configuration information may be modified again. In addition, since the network configuration information is updated and iterated, it needs to be executed last.

[0056] The device configuration regression method provided in the embodiment of the present application performs configuration regression processing on the target computing device based on the corresponding configuration information before the target computing device performs computing power sharing and circulation. Specifically, in this method, the basic container configuration information, the basic system configuration information and the reference network configuration information can be obtained first, wherein the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined on the basis of the configuration information of the original network parameters; then, based on the basic container configuration information, the target container in the target computing device is subjected to container configuration regression processing, based on the basic system configuration information, the system configuration parameter configuration of the target computing device is subjected to system configuration regression processing, and based on the reference network configuration information, the network configuration parameter configuration of the target computing device is subjected to network configuration regression processing. In this way, before the target computing device carries out computing power sharing and circulation, the container environment configuration, system parameter configuration and network parameter configuration of the target computing device are all returned to the corresponding initial states based on the above-mentioned basic container configuration information, basic system configuration information and reference network configuration information, so as to avoid the basic environment configuration of the AI ​​training task previously executed on the target computing device from interfering with the AI ​​training task executed thereafter; in addition, by performing configuration regression processing on the target computing device based on the corresponding configuration information, the relevant configuration of the target computing device can be restored more efficiently, reducing the loss of computing power resources.

[0057] In view of the fact that the relevant technology usually adopts the mode of re-installing the computing device, that is, reinstalling the operating system for the computing device, so as to realize the initialization of the container environment configuration and system parameter configuration. Reinstalling the operating system for the computing device, on the one hand, is time-consuming, and on the other hand, reinstalling the operating system will also initialize the network parameters of the computing device, which will overwrite the network tuning parameters of the computing device. Every time the AI ​​model training is performed, the parameters of the tuning network need to be reconfigured, resulting in a waste of computing resources.

[0058] In view of this, the embodiment of the present application provides a method for performing network configuration regression processing on the network parameter configuration of the target computing device, which can quickly restore the network parameter configuration of the target computing device, and at the same time retain the configuration items of the network performance tuning, thereby retaining the platform performance obtained based on the previous network performance tuning operation, avoiding re-performing the tuning operation on the network, resulting in a waste of resources. The specific implementation process of the method is as follows:

[0059] First of all, it should be noted that the reference network configuration information is determined based on the network tuning operation performed based on the original network parameter configuration information. It is understandable that when the target computing device is training the AI ​​model, it will perform a tuning operation on the network of the target computing device based on the original network parameter configuration information of the target computing device, and save the corresponding tuning parameters; the reference network configuration information includes the tuning parameters obtained by tuning the network performance based on the original network parameter configuration information.

[0060] Specifically, the reference network configuration information includes fixed configuration information corresponding to each network configuration item and temporary configuration information corresponding to at least one network configuration item; the fixed configuration information corresponding to each network configuration item is the network configuration information fixedly stored in the reference network configuration information, which can be the configuration information of the original network parameters, or the configuration information of the network parameters determined to have a network performance optimization effect after double-layer verification; the temporary configuration information is the configuration information that can optimize the network performance determined through historical network tuning operations, and the temporary configuration information can also be understood as the configuration information of the network parameters determined to have a network performance optimization effect after a single-layer verification. If it is determined through the second-layer verification that the temporary configuration information does have a network performance optimization effect, it can be determined as the fixed configuration information.

[0061] On this basis, based on the reference network configuration information, the network parameter configuration of the target computing device is processed by network configuration regression. The specific implementation process is as follows:

[0062] First, the reference network configuration information is imported into a reference network configuration file, and then the reference network configuration file is transmitted to the target computing device; specifically, the reference network configuration information can be obtained on an initialized computing device, and then the reference network configuration information obtained on the initialized computing device is imported into the reference network configuration file.

[0063] If the reference network configuration file includes temporary configuration information corresponding to a certain network configuration item, the temporary configuration information corresponding to the network configuration item is used to overwrite the configuration information of the corresponding network configuration item in the target computing device. If the reference network configuration file does not include temporary configuration information corresponding to a certain network configuration item, but only includes fixed configuration information corresponding to the network configuration item, the fixed configuration information corresponding to the network configuration item is used to overwrite the configuration information of the corresponding network configuration item in the target computing device.

[0064] When performing network configuration regression processing on the network parameter configuration of the target computing device, the embodiment of the present application can retain the network tuning parameters of the network configuration items by using the reference network configuration information to overwrite the configuration information of the network configuration items in the target computing device, thereby retaining the platform performance obtained based on the previous network performance tuning operation, avoiding re-performing tuning operations on the network and causing waste of resources.

[0065] In an optional embodiment of the present application, after the temporary configuration information corresponding to the network configuration item is used to overwrite the configuration information of the network configuration item in the target computing device, it is also possible to verify whether the temporary configuration information has an optimization effect on the network performance of the target computing device. If it can, the temporary configuration information can be used as the new fixed configuration information corresponding to the network configuration item to overwrite the fixed configuration information of the network configuration item, so as to ensure that the computing device has good platform performance when the subsequent related device configuration regression processing is performed based on the fixed configuration information. The specific implementation process is as follows:

[0066] When the configuration information of the corresponding network configuration item in the target computing device is overwritten with the temporary configuration information corresponding to a certain network configuration item, it is detected whether the network performance of the target computing device produces an optimization effect; if an optimization effect is produced, the temporary configuration information corresponding to the network configuration item is used as the new fixed configuration information corresponding to the network configuration item, and the new fixed configuration information is used to overwrite the fixed configuration information corresponding to the network configuration item in the reference network configuration information.

[0067] Specifically, during the AI ​​model training process of the target computing device, if the tuning configuration is performed for the network performance, and in this model training, the network performance is optimized, the configuration information obtained by the tuning configuration is temporarily recorded in the database as temporary configuration information. During the next model training process, it is verified whether the temporary configuration information has the effect of optimizing the network performance for this model training. If so, it is considered that the temporary configuration information generally has the effect of optimizing the network performance for model training, and the temporary configuration information is used as the new fixed configuration information corresponding to the network configuration item, and the new fixed configuration information is used to overwrite the fixed configuration information corresponding to the network configuration item in the reference network configuration information.

[0068] In this way, the network configuration regression processing of the target computing device can be realized, and the temporary configuration information that can improve the network performance of the target computing device can be retained during the subsequent AI model training. By double-verifying the configuration information generated based on the network performance tuning operation, it is ensured that the saved solidified configuration information has a better network performance optimization effect, which can bring better platform performance to the computing device.

[0069] The following example illustrates the specific implementation process of performing network configuration regression processing on the network parameter configuration of the target computing device based on the reference network configuration information:

[0070] See also Figure 3 , Figure 3 A process framework diagram for performing network configuration regression processing on the network parameter configuration of a target computing device provided in an embodiment of the present application.

[0071] In an optional embodiment of the present application, a database may be used to save the original network configuration file of the target computing device. Specifically, the initialized computing device will import the original network parameter configuration information into the original network configuration file. It can be understood that the original network parameter configuration information includes the commonly used configuration information of the initialized network and the configuration information transmitted by the Transmission Control Protocol (TCP). Among them, the commonly used network configuration information includes information such as device, type, Domain Name System (DNS), gateway, etc. Before the target computing device performs computing power sharing and circulation, the target computing device will obtain the original network configuration file in the database, and use the original network configuration file to initialize the network configuration of the target computing device.

[0072] When the AI ​​model is being trained, if the optimization operation of the tuning configuration is performed on the network performance of the target computing device, and the network optimization effect is verified during the AI ​​model training process, the tuning parameters corresponding to the network configuration item will be temporarily recorded in the database, but the corresponding network configuration item of the original network configuration file in the database will not be changed.

[0073] A double verification process needs to be performed for the tuning parameters corresponding to the network configuration item temporarily recorded in the database; specifically, before the next AI model training, that is, when the target computing device is shared and circulated, the tuning parameters corresponding to the network configuration item temporarily recorded in the database will be written into the reference network configuration file returned to the target computing device. If the network configuration item also has the effect of optimizing network performance during the next AI model training process, the tuning parameters corresponding to the network configuration item temporarily recorded in the database will be used as the new solidified configuration information corresponding to the network configuration item, and the original network configuration file of the target computing device will be updated to generate a reference network configuration file, and the network configuration information in the reference network configuration file will be used to update the network configuration item corresponding to the target computing device.

[0074] In this way, the network configuration regression processing of the target computing device can be realized, and the temporary configuration information that can improve the network performance of the target computing device can be retained during subsequent AI model training, thereby improving the efficiency of AI model training and further avoiding the problem of waste of computing power resources of the target computing device.

[0075] Before the target computing device performs the container configuration regression operation, it is necessary to uninstall the container environment and clean up the remaining files. When performing the container uninstallation operation, since the container environment performs the uninstallation, there may be residual files or residual process information of the container that have not been cleaned up, and it is necessary to perform the residual data cleaning operation. The specific implementation of the residual data cleaning operation is as follows:

[0076] An uninstall operation is performed on the target container; after the target container completes the uninstall operation, data may remain. In the embodiment of the present application, the remaining data can be divided into two categories: one is residual process data; the other is residual file data. Then, the residual process data in the target container and the residual file data in the target container are cleaned up.

[0077] Specifically, the process of cleaning up residual data is as follows: Figure 4 As shown, see Figure 4 , Figure 4 A schematic diagram of a process for cleaning residual data provided in an embodiment of the present application.

[0078] For the residual process data in the target container, you can send a kill signal by specifying the process number to complete the cleanup operation of the residual process data. Specifically, the cgroup of the container is a file group where the container information is aggregated and saved. The proc directory under the cgroup contains the process pid list in the container; pid can be understood as the identifier of the process. Then execute the ls operation of the directory. In Linux, ls stands for "list", which means "list". Its main function is to display the contents under the specified working directory. It can list the files and subdirectories contained in the working directory, and can also view file permissions. After executing the ls operation of the directory, you can get the identifier pid of all processes in the container. Then use the pid information of the file as input to verify whether the process of each pid number exists one by one. If it exists, execute the delete kill operation to clear the residual process data corresponding to the process.

[0079] For the residual file data in the target container, the residual file data mainly includes empty container process file directories and other information. You can also perform a del operation by reusing the pid list. After the container environment is uninstalled and the residual data is cleaned up, you can perform a container configuration regression operation on the target container.

[0080] In this way, before performing container configuration regression processing on the target container, the residual process data and residual file data in the target container can be cleaned up, so as to avoid the container configuration of the initialized target container being affected by the residual data, and further avoid affecting the training of the AI ​​model.

[0081] In some possible implementations, the target container configuration information corresponding to the target container in the target computing device is compared with the basic container configuration information, and according to the comparison result, the container configuration regression processing is performed on the target container, which can be specifically implemented in the following manners:

[0082] Import the basic container configuration information into the basic container configuration file, and then transmit the basic container configuration file to the target computing device. Specifically, the basic container configuration file can be obtained in the initialized computing device, and the basic container configuration file can be saved in a database. When it is necessary to perform container configuration regression processing on the target container, the basic container configuration file in the database is imported into the target computing device.

[0083] Then, for each basic container configuration information in the basic container configuration file, determine the target container configuration information in the target container that corresponds to the basic container configuration information. It can be understood that the basic container configuration information is the initial container configuration information of the target computing device; the target container configuration information is the container configuration information after the container configuration information is adjusted during the training of the AI ​​model based on the initial container configuration information.

[0084] Before the target computing device performs computing power resource sharing and circulation, the container configuration information of the target computing device needs to be restored to the initialized state. Therefore, the target container configuration information is compared with the basic container configuration information to see if they are consistent. If they are inconsistent, the target container configuration information is overwritten with the basic container configuration information, so that the container information of the target computing device can be restored to the initialized state. If the target container configuration information is consistent with the basic container configuration information, it means that the target container information is already in the initialized state, and the operation of overwriting the target container configuration information with the basic container configuration information is not performed.

[0085] In this way, there is no need to reinstall the operating system on the target computing device. By comparing the target container configuration information with the basic container configuration information to see if they are consistent, if they are inconsistent, the basic container configuration information is used to overwrite the target container configuration information, thereby realizing the initialization operation of the container environment of the target computing device. This takes very little time and can avoid wasting the computing power resources of the target computing device.

[0086] In some possible implementations, based on the basic system configuration information, a system configuration regression process is performed on the system parameter configuration of the target computing device, which can be specifically implemented in the following manners:

[0087] Import the basic system configuration information into the basic system configuration file, and transmit the basic system configuration file to the target computing device. Specifically, the basic system configuration information includes the configuration under the proc, etc and bin directories of the system, as well as the binary name of the library file and the message digest algorithm (md5) signature value and other information. The format of the basic system configuration file is shown in Table 1.

[0088] See Table 1, which is a schematic diagram of a basic system configuration file provided by an embodiment of the present application. Table 1 shows the storage form of information such as the configuration of the system proc, etc and bin directories, as well as the binary name and md5 signature value of the library file included in the basic system configuration file.

[0089] Table 1

[0090] File name / configuration item md5 signature value Configuration Values File Name A xxxx -- Configuration item B -- xxx

[0091] A database can be used to save basic system configuration files. When it is necessary to perform system configuration regression processing on the target system parameter configuration, the basic system configuration file in the database is imported into the target computing device. It can be understood that the basic system configuration information is the initial system configuration information of the target computing device; the system parameter configuration of the target computing device is the system configuration information after the system parameter configuration information is adjusted during the training of the AI ​​model based on the initial system configuration information.

[0092] Before the target computing device performs computing power resource sharing and circulation, the system configuration information of the target computing device needs to be restored to an initialized state. Therefore, the target system configuration information needs to be overwritten with the basic system configuration information, so that the system information of the target computing device can be restored to an initialized state.

[0093] In this way, there is no need to reinstall the operating system on the target computing device. By using the basic system configuration information to overwrite the target system configuration information, the system environment initialization operation of the target computing device is realized, which takes very short time and can avoid wasting the computing power resources of the target computing device.

[0094] Considering that the target library file is large, the target library file can be a secondary library file; and usually, the target library file is less modified, so for the target library file, the target signature value comparison operation can be performed first, where the target signature value can be an md5 signature value. When the target signature value is inconsistent, the copy and overwrite operation of the target library file is performed. This can be implemented in the following ways:

[0095] For the target library file in the target computing device, first calculate the target signature value corresponding to the target library file; then compare whether the target signature value is consistent with the basic signature value corresponding to the basic library file in the basic system configuration file. If they are inconsistent, the target library file is overwritten with the basic library file; if they are consistent, skip the operation of overwriting the target library file with the basic library file. In this way, before overwriting the target library file with the basic library file, compare whether the target signature value corresponding to the target library file is consistent with the basic signature value corresponding to the basic library file in the basic system configuration file. If they are consistent, skip the operation of overwriting the target library file with the basic library file.

[0096] In this way, the time for system configuration regression processing of the system parameter configuration of the target computing device can be reduced, and the utilization efficiency of the computing power resources of the target computing device can be improved.

[0097] The above embodiment describes the specific implementation process of the device configuration regression method. Starting from how to improve the utilization rate of the computing power resources of the computing device, the embodiment of this application proposes the idea of ​​mirroring configuration from the perspective of saving the full configuration. By mirroring the container environment, system kernel parameters and network configuration parameters, the original security configuration file is constructed to support the configuration regression operation before the computing power of the computing device is transferred, so as to ensure the initialization of the device environment configuration. The overall architecture diagram of the device configuration regression method is as follows: Figure 5 shown.

[0098] See also Figure 5 , Figure 5 An overall architecture diagram of a device configuration regression method provided for an embodiment of the present application. The architecture diagram is divided into three processes. Before the target computing device performs power sharing and circulation, the container configuration regression processing operation is first executed in order. After the container configuration regression processing is executed; then the system configuration regression processing operation is executed; and finally the network configuration regression processing operation is executed. The order of these three processes cannot be disordered, otherwise the configuration will be overwritten. For example, if the system configuration information is reset first, and then the container environment configuration is executed, then the system configuration information may be modified again when the container configuration information is corrected; since the network configuration information is updated and iterated, it needs to be executed last.

[0099] In the solution provided in the embodiment of the present application, the container configuration regression processing, since the container environment is an incremental installation and there is an uninstall command, this solution is for the container configuration regression processing, first uninstalling the container environment and deleting the residual file data and process data; then using the initialization computing device and the target computing device to perform a comparison and verification of the container configuration information, if the comparison is the same, skipping, if there is a difference in the comparison, performing an automated difference correction operation.

[0100] For the system configuration regression processing of the target computing device, since different models may modify or reconfigure system parameters when executing training, for example, the system, proc, etc. related parameters are modified. Different model system configurations are different, and even mutually exclusive. The embodiment of the present application constructs a configuration file without any modification for the system parameters. After each model training is completed, the system configuration of the device is reinitialized through the constructed basic system configuration file.

[0101] For the parameter configuration of the network of the computing device, the embodiment of the present application adopts a two-stage strategy. In the first stage, a reference network configuration file is constructed to facilitate the reinitialization operation of the network parameters of the target computing device. For example, the network parameters include the maximum data packet (Maximum Transmission Unit, mut) value transmitted on the network, the optimization of kernel resources and the occupancy configuration of floppy disk fd resources, etc. In the second stage, for the configuration items of network performance tuning, these items will be summarized into the reference network configuration file to realize the platform capability of high-performance network communication.

[0102] By performing container configuration regression processing, system configuration regression processing and network configuration regression processing on the target computing device, the flow of computing power resources of the large model computing power device can be completed. There is no need to restart and reinstall the operating system of the computing device. It is only necessary to perform programmed verification and reinitialization operations on the configuration execution, so that the computing power of the computing device can be efficiently circulated, avoiding the waste of computing power caused by frequent re-listing operations; and the container configuration regression processing, system configuration regression processing and network configuration regression processing allow the computing power of the computing device to flow seamlessly between model tasks, which helps to improve the R&D efficiency of model personnel and avoid the long environment initialization cycle and the waiting of R&D personnel; further, the embodiment of the present application is conducive to the improvement of the communication efficiency of the computing device by regularizing the configuration execution of network performance tuning on a platform, avoiding the need to reconfigure the parameters of the tuning network every time, and improving the efficiency and effectiveness of model training; further, the embodiment of the present application reduces the traditional restart and reinstallation process through mirror configuration, which also reduces the triggering amount of problems in the process, and helps to improve the stability of the system.

[0103] In order to more clearly describe the implementation process of the configuration regression method of the computing device provided in the embodiment of the present application, see Figure 6 , Figure 6 A flowchart of a configuration regression method implementation process for a computing device provided in an embodiment of the present application.

[0104] First, perform a copy operation of the environment configuration, that is, obtain the basic container configuration information, basic system configuration information and reference network configuration information in the initialized computing device; then import the basic container configuration information into the basic container configuration file, import the basic system configuration information into the basic system configuration file, import the reference network configuration information into the reference network configuration file, and save the reference network configuration file to the database. When performing the transfer of large model computing power, perform the uninstallation operation of the containerized environment in sequence; and determine whether there is residual data information. If there is residual data information, after cleaning the residual data information, compare the target container configuration information with the basic container configuration information. If the two are inconsistent, overwrite the target container configuration information by using the basic container configuration information.

[0105] For the system configuration, this solution uses the original initialized system configuration to perform operations covered by the system configuration to ensure strong consistency of the system configuration.

[0106] For the network configuration parameters, if double performance verification is performed, the performance configuration parameters will be written into the reference network configuration file of the database as a platform configuration. Other non-performance configuration information will be overwritten to achieve strong consistency protection of network configuration.

[0107] The present application also provides a device configuration regression device, see Figure 7 , Figure 7 A schematic diagram of a device configuration regression apparatus provided in an embodiment of the present application, the apparatus comprising:

[0108] The acquisition module 701 is used to obtain basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters.

[0109] The comparison module 702 is used to compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information for the target computing device to share the computing resources.

[0110] The configuration regression processing module 703 is used to perform container configuration regression processing on the target container according to the comparison result; perform system configuration regression processing on the system parameter configuration of the target computing device based on the basic system configuration information; and perform network configuration regression processing on the network parameter configuration of the target computing device based on the reference network configuration information.

[0111] The device configuration regression device provided in the embodiment of the present application is used to execute the corresponding method. Before the computing device performs computing power sharing and circulation, the container environment configuration, system parameter configuration and network parameter configuration of the computing device are respectively regressed using the corresponding original configuration information, so that the container environment configuration, system parameter configuration and network parameter configuration of the computing device are returned to the initial state. Among them, the original configuration information includes the original container configuration information, the original system configuration information and the reference network configuration information determined based on the original network configuration information. In this way, it is ensured that when computing power is shared and circulated, the basic environment configuration of the AI ​​training task previously executed on the computing device will not interfere with the AI ​​training task executed thereafter; and, compared with the method of reinstalling the operating system in the prior art, performing regression operations based on the configuration information can more efficiently restore the basic environment configuration and reduce computing power loss.

[0112] In an optional embodiment of the present application, the reference network configuration information is determined according to a network tuning operation performed based on the configuration information of the original network parameters, the reference network configuration information includes fixed configuration information corresponding to each network configuration item, and temporary configuration information corresponding to at least one network configuration item, the temporary configuration information is configuration information that can optimize network performance determined by historical network tuning operations;

[0113] The network configuration regression processing module 703 specifically includes:

[0114] An import unit, used for importing reference network configuration information into a reference network configuration file;

[0115] a transmission unit, transmitting the reference network configuration file to a target computing device;

[0116] An overwriting unit is used to, for each network configuration item in the target computing device, overwrite the configuration information of the network configuration item in the target computing device with the temporary configuration information corresponding to the network configuration item if the reference network configuration file includes the temporary configuration information corresponding to the network configuration item; and overwrite the configuration information of the network configuration item in the target computing device with the fixed configuration information corresponding to the network configuration item if the reference network configuration file does not include the temporary configuration information corresponding to the network configuration item.

[0117] In an optional embodiment of the present application, the network configuration regression processing module 703 further includes a detection unit;

[0118] In the case where the configuration information of the network configuration item in the target computing device is overwritten with the temporarily stored configuration information corresponding to the network configuration item, a detection unit is used to detect whether the network performance of the target computing device produces an optimization effect;

[0119] If yes, the overwriting unit is used to determine the temporarily stored configuration information corresponding to the network configuration item as the new fixed configuration information corresponding to the network configuration item, and use the new fixed configuration information to overwrite the fixed configuration information corresponding to the network configuration item in the reference network configuration information.

[0120] In an optional embodiment of the present application, the device configuration regression device also includes an unloading device and a cleaning device; before comparing the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, the unloading device is used to perform an unloading operation on the target container; the cleaning device is used to clean up the residual process data in the target container and the residual file data in the target container.

[0121] In an optional embodiment of the present application, the regression processing module 703 is configured to:

[0122] An import unit, specifically used to import basic container configuration information into a basic container configuration file;

[0123] A transmission unit, specifically used to transmit the basic container configuration file to a target computing device;

[0124] The overwriting unit is specifically used to determine, for each basic container configuration information in the basic container configuration file, the target container configuration information corresponding to the basic container configuration information in the target container, and compare whether the target container configuration information is consistent with the basic container configuration information. If they are inconsistent, the target container configuration information is overwritten with the basic container configuration information; if they are consistent, the target container configuration information is not overwritten with the basic container configuration information.

[0125] In an optional embodiment of the present application, the regression processing module 703 is configured to:

[0126] An import unit, specifically used to import basic system configuration information into a basic system configuration file;

[0127] A transmission unit, specifically used to transmit the basic system configuration file to a target computing device;

[0128] The overwriting unit is specifically used to overwrite various target system configuration information in the target computing device using various basic system configuration information in the basic system configuration file.

[0129] In an optional embodiment of the present application, the configuration regression processing module further includes a calculation unit;

[0130] The calculation unit is specifically used to calculate the target signature value corresponding to the target library file in the target computing device;

[0131] The overwriting unit is specifically used to compare whether the target signature value is consistent with the basic signature value corresponding to the basic library file in the basic system configuration file. If they are inconsistent, the target library file is overwritten by the basic library file. If they are consistent, the execution of overwriting the target library file with the basic library file is skipped.

[0132] The device configuration regression device provided in the embodiment of the present application performs configuration regression processing on the target computing device based on the corresponding configuration information before the target computing device performs computing power sharing and circulation. In this way, before the target computing device performs computing power sharing and circulation, based on the above-mentioned basic container configuration information, basic system configuration information and reference network configuration information, the container environment configuration, system parameter configuration and network parameter configuration of the target computing device are all regressed to the corresponding initial state, thereby avoiding the basic environment configuration of the AI ​​training task previously executed on the target computing device from interfering with the AI ​​training task executed thereafter; in addition, the configuration regression processing of the target computing device based on the corresponding configuration information can more efficiently restore the relevant configuration of the target computing device and reduce the loss of computing power resources.

[0133] The embodiment of the present application also provides a computer device, which may specifically be a terminal device or a server. The terminal device and the server provided in the embodiment of the present application will be introduced below from the perspective of hardware entity.

[0134] See also Figure 8 , Figure 8 Schematic diagram of the structure of the terminal device provided in the embodiment of the present application. Figure 8 For the sake of convenience, only the parts related to the embodiments of the present application are shown. For specific technical details not disclosed, please refer to the method part of the embodiments of the present application. The terminal can be any terminal device including a mobile phone, a tablet computer, a personal digital assistant, a point of sales (POS), a car computer, etc., taking the terminal as a computer as an example:

[0135] Figure 8 FIG. 1 is a block diagram showing a partial structure of a computer related to a terminal provided in an embodiment of the present application. Figure 8 The computer includes: a radio frequency (RF) circuit 1210, a memory 1220, an input unit 1230 (including a touch panel 1231 and other input devices 1232), a display unit 1240 (including a display panel 1241), a sensor 1250, an audio circuit 1260 (which can be connected to a speaker 1261 and a microphone 1262), a wireless fidelity (WiFi) module 1270, a processor 1280, and a power supply 1290. Those skilled in the art can understand that Figure 8The computer structure shown in the figure does not constitute a limitation of the computer, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0136] The memory 1220 can be used to store software programs and modules. The processor 1280 executes various functional applications and data processing of the computer by running the software programs and modules stored in the memory 1220. The memory 1220 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the computer (such as audio data, a phone book, etc.), etc. In addition, the memory 1220 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0137] The processor 1280 is the control center of the computer, which uses various interfaces and lines to connect various parts of the entire computer, and executes various functions of the computer and processes data by running or executing software programs and / or modules stored in the memory 1220, and calling data stored in the memory 1220. Optionally, the processor 1280 may include one or more processing units; preferably, the processor 1280 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 1280.

[0138] In the embodiment of the present application, the processor 1280 included in the terminal also has the following functions:

[0139] Obtaining basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters;

[0140] For the target computing device to share computing power resources, compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, and perform container configuration regression processing on the target container according to the comparison result;

[0141] Based on the basic system configuration information, performing system configuration regression processing on the system parameter configuration of the target computing device;

[0142] Based on the reference network configuration information, a network configuration regression process is performed on the network parameter configuration of the target computing device.

[0143] Optionally, the processor 1280 is further used to execute the steps of any implementation method of the device configuration regression method provided in the embodiments of the present application.

[0144] See also Fig. 9 , Fig. 9 A schematic diagram of the structure of a server provided for an embodiment of the present application. The server 1300 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 1322 (for example, one or more processors) and memory 1332, and one or more storage media 1330 (for example, one or more mass storage devices) storing application programs 1342 or data 1344. Among them, the memory 1332 and the storage medium 1330 may be short-term storage or permanent storage. The program stored in the storage medium 1330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the server. Furthermore, the central processing unit 1322 may be configured to communicate with the storage medium 1330 to execute a series of instruction operations in the storage medium 1330 on the server 1300.

[0145] The server 1300 may also include one or more power supplies 1326, one or more wired or wireless network interfaces 1350, one or more input and output interfaces 1358, and / or one or more operating systems, such as Windows Server 2003. TM , Mac OS X TM , Unix TM ,Linux TM , FreeBSD TM etc.

[0146] The steps performed by the server in the above embodiment can be based on the Fig. 9 The server structure shown.

[0147] The CPU 1322 is used to execute the following steps:

[0148] Obtaining basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters;

[0149] For the target computing device to share computing power resources, compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, and perform container configuration regression processing on the target container according to the comparison result;

[0150] Based on the basic system configuration information, performing system configuration regression processing on the system parameter configuration of the target computing device;

[0151] Based on the reference network configuration information, a network configuration regression process is performed on the network parameter configuration of the target computing device.

[0152] Optionally, the CPU 1322 may also be used to execute steps of any implementation of the device configuration regression method provided in the embodiments of the present application.

[0153] An embodiment of the present application also provides a computer-readable storage medium for storing a computer program, wherein the computer program is used to execute any one of the implementation methods of a device configuration regression method described in the aforementioned embodiments.

[0154] The embodiment of the present application also provides a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. The processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes any one of the implementations of the device configuration regression method described in the above embodiments.

[0155] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0156] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0157] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0158] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0159] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store computer programs.

[0160] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0161] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A device configuration regression method, characterized in that: The method comprises: Obtaining basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters; For the target computing device to share computing power resources, compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, and perform container configuration regression processing on the target container according to the comparison result; Based on the basic system configuration information, performing system configuration regression processing on the system parameter configuration of the target computing device; Based on the reference network configuration information, a network configuration regression process is performed on the network parameter configuration of the target computing device.

2. The method according to claim 1, characterized in that The reference network configuration information is determined according to a network tuning operation performed based on the configuration information of the original network parameters, the reference network configuration information includes fixed configuration information corresponding to each network configuration item, and temporary configuration information corresponding to at least one network configuration item, the temporary configuration information is configuration information that can optimize network performance determined by the historical network tuning operation; The performing network configuration regression processing on the network parameter configuration of the target computing device based on the reference network configuration information includes: importing the reference network configuration information into a reference network configuration file, and transmitting the reference network configuration file to the target computing device; For each network configuration item in the target computing device, if the reference network configuration file includes temporary configuration information corresponding to the network configuration item, the temporary configuration information corresponding to the network configuration item is used to overwrite the configuration information of the network configuration item in the target computing device; if the reference network configuration file does not include the temporary configuration information corresponding to the network configuration item, the fixed configuration information corresponding to the network configuration item is used to overwrite the configuration information of the network configuration item in the target computing device.

3. The method according to claim 2, characterized in that The method further comprises: In a case where the configuration information of the network configuration item in the target computing device is overwritten by the temporarily stored configuration information corresponding to the network configuration item, detecting whether the network performance of the target computing device produces an optimization effect; If so, determine the temporarily stored configuration information corresponding to the network configuration item as the new fixed configuration information corresponding to the network configuration item, and use the new fixed configuration information to overwrite the fixed configuration information corresponding to the network configuration item in the reference network configuration information.

4. The method according to claim 1, characterized in that Before comparing the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, the method further includes: Performing an unloading operation on the target container; Residual process data in the target container and residual file data in the target container are cleaned up.

5. The method according to claim 1 or 4, characterized in that: The step of comparing the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information, and performing container configuration regression processing on the target container according to the comparison result, includes: Importing the basic container configuration information into a basic container configuration file, and transmitting the basic container configuration file to the target computing device; For each item of the basic container configuration information in the basic container configuration file, determine the target container configuration information in the target container that corresponds to the basic container configuration information, and compare whether the target container configuration information is consistent with the basic container configuration information; if they are inconsistent, use the basic container configuration information to overwrite the target container configuration information; if they are consistent, do not perform the overwriting of the target container configuration information with the basic container configuration information.

6. The method according to claim 1, characterized in that The performing system configuration regression processing on the system parameter configuration of the target computing device based on the basic system configuration information includes: importing the basic system configuration information into a basic system configuration file, and transmitting the basic system configuration file to the target computing device; Utilize the various items of the basic system configuration information in the basic system configuration file to overwrite various items of the target system configuration information in the target computing device.

7. The method according to claim 6, characterized in that The method further comprises: For the target library file in the target computing device, calculating the target signature value corresponding to the target library file; Compare the target signature value with the basic signature value corresponding to the basic library file in the basic system configuration file to see if they are consistent. If they are inconsistent, use the basic library file to overwrite the target library file. If they are consistent, skip the execution of overwriting the target library file with the basic library file.

8. A device configuration regression device, characterized in that: The device comprises: A basic information acquisition module, used to acquire basic container configuration information, basic system configuration information and reference network configuration information; the basic container configuration information is the configuration information of the original container environment, the basic system configuration information is the configuration information of the original system parameters, and the reference network configuration information is determined based on the configuration information of the original network parameters; A container regression module is used to compare the target container configuration information corresponding to the target container in the target computing device with the basic container configuration information for the target computing device to share computing resources, and perform container configuration regression processing on the target container according to the comparison result; A system parameter regression module, configured to perform system configuration regression processing on the system parameter configuration of the target computing device based on the basic system configuration information; A network parameter regression module is used to perform network configuration regression processing on the network parameter configuration of the target computing device based on the reference network configuration information.

9. A computer device, characterized in that: The computer device includes a processor and a memory; The memory is used to store computer programs; The processor is configured to execute the device configuration regression method according to any one of claims 1 to 7 according to the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store a computer program, and the computer program is used to execute the device configuration regression method according to any one of claims 1 to 7.

Images