Computing power resource management method, system and device, medium, and program product

Abstract

Disclosed are a computing power resource management method, system and device, a medium, and a program product. The computing power resource management system comprises a resource requirement platform, a proxy gateway, and a computing power resource pool. The resource requirement platform is configured to acquire requirement description text inputted by a user, and parse the requirement description text to obtain a service type and computing power resource parameters required by the user. The proxy gateway is configured to acquire, on the basis of the service type and the computing power resource parameters, a corresponding computing power resource node from the proxied computing power resource pool for configuration, and generate an access address of the computing power resource node. By means of the technical means of the present disclosure, appropriate computing power resources can be recommended and provided for a user on the basis of a service scenario and user requirements, thereby effectively meeting usage requirements of the user and improving use experience of the user.

Description

A computing power resource management method, system, device, medium and program product

[0001] Cross-reference to Related Applications

[0002] The present disclosure claims priority to the Chinese Patent Application No. 202410738704.2, filed on June 7, 2024, entitled "A computing power resource management method, system, device, medium and program product", by China Mobile Research Institute Co., Ltd. and China Mobile Communications Group Co., Ltd. TECHNICAL FIELD

[0003] The present disclosure relates to the field of cloud computing technology, and particularly relates to a computing power resource management method, system, device, medium and program product. BACKGROUND

[0004] In the existing computing power service providing method, a fixed computing power server parameter combination package is provided for a user, and after the user selects, the server IP address corresponding to the parameter combination is queried in the related service, so as to provide a service interface of the computing power resource.

[0005] However, the inventors have found that the existing method cannot recommend computing power resource parameters according to business scenarios, and for users who do not understand the actual effect of resource parameters, they cannot accurately select a server parameter combination package according to their own business scenarios when facing several fixed parameter packages provided by a computing power provider, and even some users only know the natural language expression of computing power demand, which may lead to the wrong selection of computing power resources, and thus the computing power resources do not meet the actual needs of the user. SUMMARY

[0006] The purpose of the embodiments of the present disclosure is to provide a computing power resource management method, system, device, medium and program product, which can recommend and provide appropriate computing power resources for users according to business scenarios and user needs, effectively meet the use needs of users, and improve the use experience of users.

[0007] To achieve the above-mentioned purpose, the embodiments of the present disclosure provide a computing power resource management system, comprising a resource demand platform, a proxy gateway and a computing power resource pool.

[0008] The resource demand platform is configured to obtain a demand description text input by a user, and parse the demand description text to obtain a business type and a computing power resource parameter of a user demand.

[0009] The proxy gateway is configured to obtain a corresponding computing power resource node in the proxyed computing power resource pool according to the business type and the computing power resource parameter, configure the computing power resource node, and generate an access address of the computing power resource node.

[0010] As an improvement of the above scheme, the resource demand platform comprises a resource demand platform interface, a business type recommendation module and a parameter analysis module;

[0011] The resource demand platform interface is configured to obtain a demand description text input by a user; wherein the demand description text is a business type description text or a resource parameter description text;

[0012] When the demand description text is a business type description text, the business type recommendation module is configured to analyze the business type description text to obtain a business type; and,

[0013] The preset business knowledge base records a corresponding relationship between a business type and a computing resource parameter;

[0014] When the demand description text is a resource parameter description text, the parameter analysis module is configured to analyze the resource parameter description text to obtain a computing resource parameter and send the computing resource parameter to the business type recommendation module;

[0015] The business type recommendation module is further configured to search the business knowledge base to obtain a business type corresponding to the computing resource parameter.

[0016] As an improvement of the above scheme, the resource demand platform further comprises an address coding module, and the proxy gateway comprises a computing resource configuration module;

[0017] The address coding module is configured to determine an address of a corresponding proxy gateway according to the business type to establish a communication link between the resource demand platform and the proxy gateway;

[0018] The business type recommendation module is further configured to send the computing resource parameter to the computing resource configuration module;

[0019] The computing resource configuration module is configured to match a corresponding computing resource node in a proxy computing resource pool according to the computing resource parameter and generate an access address of the computing resource node.

[0020] As an improvement of the above scheme, the resource demand platform interface is further configured to display the demand description text, the business type, the computing resource parameter and the access address of the computing resource node.

[0021] As an improvement of the above scheme, the proxy gateway further comprises a resource monitoring module and a calculation unloading module;

[0022] The resource monitoring module is configured to monitor running state information of each computing resource node; wherein the running state information is a normal running state or an abnormal running state;

[0023] The computing offloading module is configured to obtain an abnormal running state computing resource node as a to-be-offloaded node, and assign configured services of the to-be-offloaded node to other allocable computing resource nodes.

[0024] As an improvement of the above scheme, the computing offloading module is specifically configured to:

[0025] obtain an abnormal running state computing resource node as a to-be-offloaded node;

[0026] analyze configured services of the to-be-offloaded node, and record the configured services as an offloading task;

[0027] determine other allocable computing resource nodes as allocable computing resource nodes according to a service type of the offloading task, and obtain addresses of the allocable computing resource nodes;

[0028] generate an offloading notification data packet, and send the offloading notification data packet to each of the allocable computing resource nodes according to the addresses, so that the allocable computing resource nodes obtain the offloading task allocated to the allocable computing resource nodes through the offloading notification data packet and return a response result.

[0029] As an improvement of the above scheme, the offloading notification data packet is composed of data fields corresponding to each of the allocable computing resource nodes, and each of the data fields includes an encoding bit, a first interface bit and a second interface bit.

[0030] The encoding bit is composed of an encoding segmentation identifier and an encoding value of the offloading task, the first interface bit is configured to provide a data interface for the allocable computing resource node to obtain the offloading task allocated to the allocable computing resource node, and the second interface bit is configured to provide an interface for the allocable computing resource node to return the response result after obtaining the offloading task.

[0031] The disclosure also provides a computing resource management method, which is executed by a computing resource management system including a resource demand platform, a proxy gateway and a computing resource pool, and includes the following steps:

[0032] The resource demand platform obtains a demand description text input by a user.

[0033] The resource demand platform analyzes a service type and a computing resource parameter of a user demand according to the demand description text.

[0034] The proxy gateway configures corresponding computing resource nodes in the computing resource pool managed by the proxy gateway according to the service type and the computing resource parameter, and generates access addresses of the computing resource nodes.

[0035] As an improvement of the above scheme, the demand description text is a business type description text or a resource parameter description text;

[0036] The resource demand platform parses the business type and the computing resource parameter of the user demand according to the demand description text, including:

[0037] When the demand description text is a business type description text, the business type description text is parsed to obtain the business type;

[0038] A preset business knowledge base is searched to obtain the computing resource parameter corresponding to the business type; wherein, the preset business knowledge base records the corresponding relationship between the business type and the computing resource parameter;

[0039] When the demand description text is a resource parameter description text, the resource parameter description text is parsed to obtain the computing resource parameter;

[0040] The business knowledge base is searched to obtain the business type corresponding to the computing resource parameter.

[0041] As an improvement of the above scheme, the proxy gateway configures the corresponding computing resource node in the computing resource pool managed by the proxy gateway according to the business type and the computing resource parameter, generates the access address of the computing resource node, including:

[0042] According to the business type, the corresponding computing resource pool is determined;

[0043] According to the computing resource parameter, the corresponding computing resource node in the computing resource pool is matched;

[0044] The access address of the computing resource node is generated.

[0045] As an improvement of the above scheme, the method further includes: displaying the demand description text, the business type, the computing resource parameter and the access address of the computing resource node.

[0046] As an improvement of the above scheme, the method further includes:

[0047] The proxy gateway monitors the running state information of each computing resource node; wherein, the running state information is normal running state or abnormal running state;

[0048] The proxy gateway obtains the computing resource node in abnormal running state as a to-be-unloaded node;

[0049] The proxy gateway allocates the configured business of the to-be-unloaded node to other allocable computing resource nodes.

[0050] As an improvement of the above scheme, the method further comprises:

[0051] The proxy gateway acquires the abnormal running state computing resource node as a to-be-unloaded node;

[0052] Analyze the configured service of the to-be-unloaded node, denoted as an unloading task;

[0053] According to the service type of the unloading task, determine other allocable computing resource nodes as allocable computing resource nodes, and acquire the addresses of the allocable computing resource nodes;

[0054] Generate an unloading notification data packet, and send the unloading notification data packet to each allocable computing resource node according to the addresses, so that the allocable computing resource nodes acquire the unloading tasks allocated to them through the unloading notification data packet and return response results.

[0055] The embodiment of the disclosure further provides an algorithm resource management system, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the algorithm resource management method according to any one of the above.

[0056] The embodiment of the disclosure further provides a computer readable storage medium, comprising a stored computer program, wherein the computer readable storage medium controls the device where the computer readable storage medium is located to execute the algorithm resource management method according to any one of the above when the computer program runs.

[0057] The embodiment of the disclosure further provides a computer program product, comprising a computer program or computer instructions, and the computer program or the computer instructions implement the algorithm resource management method according to any one of the above when executed by a processor.

[0058] Compared with the prior art, the algorithm resource management method, system, device, medium and program product disclosed by the disclosure are based on natural language for algorithm demand analysis, which is beneficial for users to describe the required service type or resource parameter in natural language, and combines user demand and professional algorithm resource application scene experience to perform bidirectional speculation between resource parameters and service types, recommend and provide adaptive algorithm resources for users, realize more perfect algorithm demand analysis and more efficient resource utilization, effectively solve the situation that resource scheduling flexibility and expansibility are insufficient in the algorithm resource use process due to the relatively fixed algorithm providing mode in the prior art, can meet various application scenes, effectively meet the use demand of users, and improve the use experience of users. BRIEF DESCRIPTION OF DRAWINGS

[0059] FIG. 1 is a first structural schematic diagram of a computing resource management system according to an embodiment of the present disclosure;

[0060] FIG. 2 is a second structural schematic diagram of a computing resource management system according to an embodiment of the present disclosure;

[0061] FIG. 3 is a first schematic diagram of a resource demand platform interface according to an embodiment of the present disclosure;

[0062] FIG. 4 is a schematic diagram of the working principle of an address coding module according to an embodiment of the present disclosure;

[0063] FIG. 5 is a first working flow schematic diagram of a computing resource management system according to an embodiment of the present disclosure;

[0064] FIG. 6 is a second schematic diagram of a resource demand platform interface according to an embodiment of the present disclosure;

[0065] FIG. 7 is a third schematic diagram of a resource demand platform interface according to an embodiment of the present disclosure;

[0066] FIG. 8 is a third structural schematic diagram of a computing resource management system according to an embodiment of the present disclosure;

[0067] FIG. 9 is a first schematic diagram of an offloading notification data packet according to an embodiment of the present disclosure;

[0068] FIG. 10 is a second schematic diagram of an offloading notification data packet according to an embodiment of the present disclosure;

[0069] FIG. 11 is a first flow schematic diagram of a computing resource management method according to an embodiment of the present disclosure;

[0070] FIG. 12 is a second flow schematic diagram of a computing resource management method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present disclosure.

[0072] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, and are only for the purpose of facilitating the description of the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0073] The terms "first", "second", "third", etc. are used only for descriptive purposes and do not connote or imply relative importance or a quantity of the indicated technical features. Thus, features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specified.

[0074] Referring to FIG. 1 is a first structural schematic diagram of a computing power resource management system provided by an embodiment of the present disclosure. An embodiment of the present disclosure provides a computing power resource management system 10, comprising a resource demand platform 11, a proxy gateway 12 and a computing power resource pool 13; wherein,

[0075] The resource demand platform 11 is configured to obtain a demand description text input by a user; and parse a business type and a computing power resource parameter of a user demand according to the demand description text.

[0076] The proxy gateway 12 is configured to obtain corresponding computing power resource nodes in the proxyed computing power resource pool according to the business type and the computing power resource parameter, configure the computing power resource nodes, and generate an access address of the computing power resource nodes.

[0077] In an embodiment of the present disclosure, the computing power resource management system 10 is composed of three parts. The resource demand platform 11 provides an interactive entrance according to a demand description text of a user-proposed business scenario or resource parameter, parses a business type and a computing power resource parameter of a user demand according to the demand description text, and provides a platform for subsequent feedback of a computing power resource interface. The resource demand platform 11 exists in the form of a server device.

[0078] The proxy gateway 12 is configured to proxy a plurality of computing power resource pools, perform uplink and downlink data forwarding control and computing power resource pool management, and obtain corresponding computing power resource nodes in the proxyed computing power resource pool according to the business type and the computing power resource parameter, configure the computing power resource nodes, and generate an access address of the computing power resource nodes. The proxy gateway 12 is a server or a router, etc., and is a device set supporting gateway and computing functions.

[0079] The computing power resource pool 13 is a resource cluster formed according to different business types. The unit providing computing power resources in the computing power resource pool is a computing power resource node. Each computing power resource pool includes a plurality of computing power resource nodes. The computing power resource pool 13 is a server cluster, and also includes devices with routing capabilities for communication.

[0080] The computing power resource node is a node providing various capabilities of resources, and exists in the form of a computer, a server, a virtual machine, etc., supports IPv6 (Internet Protocol Version 6), and SRv6 (Segment Routing IPv6 based on an IPv6 forwarding plane). The computing power resource parameter is used to describe parameters of a computing node resource, including but not limited to memory, disk space, an operating system type, a number of cores, etc.

[0081] By using the technical means of the embodiments of the present disclosure, the computing power demand is analyzed based on natural language, which is beneficial for a user to describe a required business type or resource parameter in natural language, and to make bidirectional speculation between the resource parameter and the business type in combination with user demand and professional computing power resource application scenario experience, so as to recommend and provide suitable computing power resources for the user, to realize more perfect computing power demand analysis and more efficient resource utilization, and to effectively solve the problems of insufficient resource scheduling flexibility and expansibility in the use of computing power resources due to the fixed computing power providing mode in the prior art, to meet various application scenarios, to effectively meet the use demand of the user, and to improve the use experience of the user.

[0082] As a preferred embodiment, the embodiments of the present disclosure are further implemented on the basis of the above embodiments, and the structure of the computing power resource management system 10 is further optimized. The resource demand platform 11 includes a resource demand platform interface, a business type recommendation module, and a parameter analysis module.

[0083] The resource demand platform interface is used to obtain a demand description text input by a user; wherein the demand description text is a business type description text or a resource parameter description text.

[0084] When the demand description text is a business type description text, the business type recommendation module is used to analyze the business type description text to obtain a business type; and,

[0085] A preset business knowledge base is searched to obtain computing power resource parameters corresponding to the business type; wherein the preset business knowledge base records a corresponding relationship between a business type and computing power resource parameters.

[0086] When the demand description text is a resource parameter description text, the parameter analysis module is used to analyze the resource parameter description text to obtain computing power resource parameters, and to send the computing power resource parameters to the business type recommendation module.

[0087] The business type recommendation module is further used to search the business knowledge base to obtain a business type corresponding to the computing power resource parameters.

[0088] Referring to FIG. 2, it is a second structural schematic diagram of the computing resource management system in the embodiment of the present disclosure, the resource demand platform comprises a resource demand platform interface, a parameter analysis module and a business type recommendation module. Among them,

[0089] The resource demand platform interface is used to obtain the demand description text input by the user, and supports the user to input two types of descriptions, one is the text content describing the business type, that is, the business type description text, and the other is the text content describing the resource parameters, that is, the resource parameter description text.

[0090] The parameter analysis module is a module for analyzing the computing resource parameters according to the resource parameter description text.

[0091] The business type recommendation module is a module for analyzing the business type according to the business type description text, and judging the corresponding business type and characteristics according to the computing resource parameters, or providing computing resource parameter recommendation according to the business type.

[0092] In the specific application process, when the resource demand platform interface receives the business type description text, it is sent to the business type recommendation module. The business type recommendation module matches the description text with the existing business types, which include but are not limited to "image recognition model inference business", "speech recognition model training business", "video encoding and decoding business", "software development management business", "enterprise employee office business", "web service deployment business", etc. Each business type corresponds to a computing resource pool, which is an independent SRv6 domain. Thus, the business type required by the user is determined, and the corresponding computing resource parameters are recommended according to the business type. The specific means can be to determine the numerical range of the required parameters according to the data limit vocabulary. For the business type and the corresponding data limit vocabulary and other scene requirements, the business type recommendation module queries the preset business knowledge base. The business knowledge base contains the business type application scene and the computing resource parameter data used by it. Through the business knowledge base, the computing resource parameter requirements of a certain business type under a certain data limit vocabulary are compared, so as to recommend a set of computing resource parameters.

[0093] When the resource demand platform interface receives the business type description text, it is sent to the parameter analysis module. After the resource parameter analysis range in the parameter analysis module, the multiple computing resource parameters required by the user are analyzed and sent to the business type recommendation module. The business type recommendation module also searches for the computing resource parameter information in the business knowledge base and analyzes the business type to determine the corresponding business scene type described by the user.

[0094] More preferably, referring to FIG. 3, which is a first schematic diagram of a resource demand platform interface in an embodiment of the present disclosure, the resource demand platform interface is also used to display the demand description text, the business type, the computing power resource parameter, and the access address of the computing power resource node.

[0095] In some embodiments, the resource demand platform interface exists in the form of a web page, including page elements such as resource description, resource parameter, business type, resource node address, and resource entry. Among them, the resource description part is used to input the user's demand description text of computing power resources, that is, the business type description text or the resource parameter description text.

[0096] The resource parameter part is used to display the names (i.e. "Parameter 1", "Parameter 2", etc. in the figure) of the computing power resource parameters and the corresponding resource parameter values embodied in the resource parameter description text; or to display the names of the computing power resource parameters and the corresponding resource parameter values recommended by the business type recommendation module after analyzing the business type description text.

[0097] The business type part is used to display the business type embodied in the business type description text, or the business type obtained by the business type recommendation module after analyzing the resource parameters obtained by the parameter analysis module according to the user's input resource parameter description text.

[0098] The resource node address part is the IP address of the computing power resource node returned by the proxy gateway after resource configuration; the resource entry is the entry for users to access resources, which is convenient for users to click into the computing power resource use page.

[0099] As a preferred implementation, the resource demand platform 11 further includes an address coding module, and the proxy gateway 12 includes a computing power resource configuration module.

[0100] The address coding module is used to determine the address of the corresponding proxy gateway according to the business type, so as to establish a communication link between the resource demand platform and the proxy gateway;

[0101] The business type recommendation module is also used to send the computing power resource parameters to the computing power resource configuration module;

[0102] The computing power resource configuration module is used to match the corresponding computing power resource node in the proxy computing power resource pool according to the computing power resource parameters, and generate the access address of the computing power resource node.

[0103] Referring to FIG. 2, in the embodiment of the present disclosure, the address coding module is configured to calculate the IPv6 address allocation of the corresponding proxy gateway according to the service type required by the user, so as to establish the communication link between the resource demand platform 11 and the proxy gateway 12. The computing power resource configuration module in the proxy gateway 12 is configured to configure the computing power resource node corresponding to the computing power resource parameter required by the user in the computing power resource pool, form the computing power resource node, and configure the IP address and network.

[0104] Preferably, the address coding module is specifically configured to encode the text of the service type into binary data, and then convert the binary data into a hexadecimal address to obtain the address of the corresponding proxy gateway, and establish the communication link between the resource demand platform and the proxy gateway.

[0105] In some embodiments, the address coding module receives the service type text of the service type recommendation module, runs the address coding conversion method, and obtains the IPv6 address of the corresponding proxy gateway through the address coding conversion method processing, so as to open the communication link between the resource demand platform and the proxy gateway.

[0106] The address coding conversion method includes a set of encoding methods, which is divided into a text encoding method and an address encoding method. The text encoding method is used to encode the service type text into binary form, and the address encoding method is used to convert binary data into an IPv6 address (hexadecimal). The type covers various fields of encoding forms, including but not limited to text, image, audio, and for multimedia encoding methods such as image, audio, the main data (such as RGB data, grayscale data in image) is used as the input of the method, rather than the metadata of file format, parameters, etc.

[0107] In some embodiments, the set of encoding methods is represented as follows:

[0108] In the text encoding method and the address encoding method, one of each is selected to form an address coding conversion function, which can be represented as: f(service type, text encoding method, address encoding method).

[0109] Referring to FIG. 4, it is a schematic diagram of the working principle of the address coding module in the embodiment of the present disclosure. The address coding conversion function has built-in metadata required by the address coding method to provide necessary parameters in the coding process. In addition, the address coding conversion function has built-in address coding method adapter components to convert binary form data into the input format of the address coding method, and extract the output data of the address coding method at the appropriate link. After the steps of binary conversion and length arrangement, the IPv6 address is formed. For example, the input data of the JPEG coding method is in the unit of 8x8 grayscale decimal data, including discrete cosine transform, matrix quantization, Z-shaped coding, DC / AC separation, and entropy coding. The adapter component selects the decimal form output result of the Z-shaped coding step and converts it into hexadecimal data. According to the format of the 128-bit IPv6 address, the super-long part is truncated, and the insufficient part is zero-padded.

[0110] Further, referring to FIG. 5, it is a first workflow schematic diagram of the computing power resource management system in the embodiment of the present disclosure. The embodiment of the present disclosure further explains and illustrates the computing power resource providing process. When the user enters the resource demand platform interface to select the computing power resource, the resource demand platform, the proxy gateway, and the resource pool will enter the computing power resource providing process.

[0111] In some embodiments, when the user explicitly knows the computing power resource parameters required by the service, the user inputs the resource parameter description text in the resource description part of the resource demand platform interface. When the user only knows the service type scenario, the user inputs the service type description text in the resource description part of the resource demand platform interface.

[0112] For the service type description text (description type is service type description), the service type recommendation module matches the existing service types in the service knowledge base, and judges the value range of the required computing power resource parameters according to the data limitation words such as "low pixel" and "high definition". For the resource parameter description text (description type is resource parameter description), the parameter analysis module analyzes the specific computing power resource parameters (i.e., resource parameters in FIG. 5), and the service type recommendation module matches the corresponding service type. The resource demand platform interface displays the service type and the computing power resource parameters respectively (i.e., web displays the service type and the resource parameters).

[0113] The business type recommendation module delivers the business type to the address coding module, the address coding module calculates the IP address of the proxy gateway, the resource demand platform sends the computing resource parameter and the computing resource address to the computing resource configuration module (i.e., computing resource configuration in FIG. 5) in the proxy gateway, the computing resource configuration module configures the computing resource node meeting the computing resource parameter in the corresponding computing resource pool, and configures the IP address and the network. After the computing resource configuration module is configured successfully, the result is returned to the resource demand platform. The resource demand platform interface displays the computing resource address at the resource node address, and provides a web computing resource entry for the user to click and use the resource (i.e., the web display resource entry button is available in FIG. 5).

[0114] The application scenario is explained as follows.

[0115] When the user enters the resource demand platform interface to select the computing resource, the resource demand platform, the proxy gateway and the resource pool enter the computing resource providing process, and the specific steps are as follows.

[0116] Referring to FIG. 6, which is a second schematic diagram of the resource demand platform interface in the embodiment of the present disclosure, the user inputs the business type description text "WAV format speech recognition AI model training business" in the resource demand platform interface; the business type recommendation module extracts the business type description in the text, matches to the business type "speech recognition AI model training business", and queries the business knowledge base according to the data restriction word "WAV format", to obtain the common computing resource parameter in this scenario as "GPU is NVIDA, processor 8 cores, memory 64G, disk 1T, system Linux", and the web page displays the corresponding business type and computing resource parameter.

[0117] Referring to FIG. 7, which is a third schematic diagram of the resource demand platform interface in the embodiment of the present disclosure, the business type recommendation module delivers the user business type "speech recognition AI model training business" to the address coding module, the address coding module calculates the IP address "2001:DB8::1" of the proxy gateway according to the address coding conversion method of the business type "speech recognition AI model training business", the resource demand platform sends the computing resource parameter and the computing resource address to the computing resource configuration module in the proxy gateway, the computing resource configuration module configures the virtual machine meeting the computing resource parameter in the computing resource pool, forms the computing resource node, and configures the IP address "2001:DB8::45" and the network, and the computing resource configuration module returns the result to the resource demand platform after being configured successfully. The resource demand platform interface displays the computing resource address, and provides a web entry for using the resource.

[0118] By using the technical means of the embodiments of the present disclosure, the computing power demand analysis is performed based on natural language, the business type description and the resource parameter description are supported, and the user can describe the required business type or resource parameter in natural language. Through the business knowledge base and in combination with the professional computing power resource application scene experience, the bidirectional speculation between the resource parameters and the business types is performed. For the user who knows the computing power resource parameters, the corresponding computing power resources can be provided. For the user who can only provide the business type scene description, more scene related information can be provided to assist the user in selection, so that the user is recommended and provided with the corresponding computing power resources, and more perfect computing power demand analysis and more efficient resource utilization are realized.

[0119] As a preferred embodiment, the embodiments of the present disclosure are further implemented on the basis of any one of the above embodiments. It should be noted that in the prior art, the user selects the computing power resources and also selects the purchase period, which is generally several months or even several years. For the purchased computing power resources, the user may encounter insufficient computing, storage, and professional processing capacity in the process of use due to the adjustment of the business and the change of the application resource occupation. When the actual computing power demand of the user is greater than the purchased computing power resources, the application cannot normally run, and even causes service crash.

[0120] To solve this problem, in the embodiments of the present disclosure, the proxy gateway 12 further comprises a resource monitoring module and a computing offloading module.

[0121] The resource monitoring module is configured to monitor the running state information of each computing power resource node; wherein the running state information is a normal running state or an abnormal running state.

[0122] The computing offloading module is configured to obtain the computing power resource node in the abnormal running state as a to-be-offloaded node; and distribute the configured business of the to-be-offloaded node to other allocable computing power resource nodes.

[0123] Referring to FIG. 8, it is a third structural schematic diagram of the computing power resource management system in the embodiments of the present disclosure. The proxy gateway 12 further comprises a resource monitoring module and a computing offloading module. The resource monitoring module is configured to monitor the application running condition in the computing power resource node, and the computing offloading module is configured to obtain the running state information of the computing power resource node and manage the computing offloading process of the computing power resource node. For example, when the computing power resource of a certain computing power resource node is insufficient, it is considered that the computing power resource node is in an abnormal running state if it is monitored that the computing power resource node enters a bottleneck state or cannot provide services in terms of part of the resource parameters or professional applications, the resource node becomes a to-be-offloaded node, and the computing offloading is started.

[0124] As a preferred embodiment, the computing offloading module is specifically configured to:

[0125] obtain the abnormal running state computing resource node as a to-be-unloaded node;

[0126] analyze the configured service of the to-be-unloaded node, denoted as an unloading task;

[0127] determine other allocable computing resource nodes according to the service type of the unloading task as allocable computing resource nodes, and obtain the addresses of the allocable computing resource nodes;

[0128] generate an unloading notification data packet, and send the unloading notification data packet to each allocable computing resource node according to the address, so that the allocable computing resource node obtains the unloading task allocated to itself through the unloading notification data packet and returns a response result.

[0129] Preferably, the unloading notification data packet is composed of a data field corresponding to each allocable computing resource node, and each data field includes an encoding bit, a first interface bit and a second interface bit.

[0130] The encoding bit is composed of an encoding segmentation identifier and an encoding value of the unloading task, the first interface bit is used to provide a data interface for the allocable computing resource node to obtain the unloading task allocated to itself, and the second interface bit is used to provide an interface for the allocable computing resource node to return a response result after obtaining the unloading task.

[0131] In the embodiment of the present disclosure, the resource monitoring module knows the SID (Segment Identifier) of each computing resource node in the SRv6 domain, the available state, that is, whether it can support computing unloading tasks as a resource provider, and also knows the unloading task types of different resource nodes, including but not limited to intensive computing, data storage, high concurrency, GPU acceleration, image processing, and audio processing. These unloading task types are the basic units of computing unloading tasks, and each type has a corresponding type code.

[0132] When the resource monitoring module detects that the computing resource of the computing resource node is insufficient, it will calculate the unloading task type of the to-be-unloaded node and the resource parameter condition required by the application type according to the program running condition of the node, so as to compare and select the available resource nodes SID and the unloading task type code in the known available condition and resource parameter configuration condition of the current resource pool.

[0133] The offloading notification data packet is sent by the to-be-offloaded node, and a SID of a number of allocable resource nodes provided by the resource monitoring module is encapsulated in a Segment list of an extension header SRH (Segment Routing Header) of the data packet, as a forwarding path of the data packet. In the forwarding process of the data packet, each allocable resource node reads the calculation offloading related information carried by the data packet, including a calculation task data interface (that is, a first interface) and a calculation result receiving interface (that is, a second interface), both of which are opened on the to-be-offloaded node when the calculation offloading task is generated, and both of which are closed when the calculation offloading task is completed, the calculation task data interface is used for the resource node to obtain offloading task data of the to-be-offloaded node as an input data source of the offloading task, and the calculation result receiving interface is used for the resource node to feed back a response result to the to-be-offloaded node after completing the offloading task.

[0134] Referring to FIG. 9, it is a first schematic diagram of an offloading notification data packet in the embodiment of the present disclosure. The calculation offloading related information in the data packet needs to be read by each intermediate computing resource node in the forwarding process of the data packet, and the information is stored in a TLV (Type Length Value) structure in the SRH, and hop-by-hop analysis is realized in an extended form. The data field of each allocable resource node is composed of “encoding”, “first interface x-1” and “second interface x-2”. The “encoding” is composed of “encoding division identifier” and “encoding value”. The encoding division identifier is used to separate the information of different computing resource nodes, so as to inform each computing resource node that the calculation offloading related information of the next computing resource node is behind this place, so as to avoid analysis errors. The encoding value represents the encoding of the offloading task. In the TLV, the calculation offloading related information also exists in the “other” structure, which includes the Type and Length fields in the TLV. Since the total length of the TLV may have an 8-bit integer multiple requirement, the end may need a “zero padding” part. The positions of the calculation task data interface x-1 and the calculation result receiving interface x-2 are not required to be in front of and behind each other, and one implementation form is shown in the figure. The order of the interface identification when the computing resource node analyzes is agreed by the to-be-offloaded node, and is not required here.

[0135] The embodiment of the present disclosure further explains the computing resource extension process. In the process of using computing resources by a user, if the resource node providing the service enters a bottleneck state or cannot provide the service in terms of part of the resource parameters or professional applications, the proxy gateway will assist the computing resource pool to provide the calculation offloading service, and the specific steps are as follows:

[0136] The resource monitoring module in the proxy gateway 12 detects that the computing resource under a certain address is insufficient, the computing offloading module determines it as a node to be offloaded, analyzes the service type in the node to be offloaded, and determines the service that needs to be offloaded as an offloading task. The offloading task is analyzed to obtain the allocable computing resource node and its address corresponding to the service type of different offloading tasks, and the allocable computing resource node and its offloading task type are informed to the node to be offloaded. The node to be offloaded encapsulates an offloading notification packet carrying an SRv6 extension header, the ipv6 address of the computing resource node is encapsulated in the Segment list of the extension header of the offloading notification packet as the forwarding path of the packet, and the offloading task type, the computing task data interface and the computing result receiving interface are carried in the TLV of the offloading notification packet. The offloading notification packet is sent to each computing resource node by the computing offloading module according to the forwarding path, the computing resource node obtains the data of the offloading task from the node to be offloaded according to the computing task data interface informed by the offloading notification packet, and sends the response result to the computing result receiving interface after completing the computing offloading task. When the node to be offloaded receives all the response results, it informs the computing offloading module that the computing resource extension is completed.

[0137] The application scenario is explained as follows in the embodiments of the present disclosure.

[0138] When the resource monitoring module in the proxy gateway detects that the high-concurrency application and the image processing application of the computing resource node N1 are in an abnormal running state and are limited by performance bottlenecks, the node N1 is marked as a node to be offloaded. The resource monitoring module analyzes the service type in the virtual machine, and determines that the service type that needs to be offloaded is “high concurrency” and “image processing” as an offloading task. The computing offloading module in the proxy gateway analyzes the offloading task, obtains the resource node P1 corresponding to the “high concurrency” offloading task type and the resource node P2 corresponding to the “image processing” offloading task type, and informs the node to be offloaded N1 of the resource nodes and their computing task types.

[0139] Referring to FIG. 10, it is a second schematic diagram of the offloading notification packet in the embodiments of the present disclosure. The node to be offloaded N1 knows that the End SID of the computing resource node P1 and P2 is “2001:DB8:4::4” and “2001:DB8:6::6” respectively, and determines that the Segment List is:

[0140] At the same time, add the TLV Option, the "high concurrency" offloading task type code value is "02", add "A" as the code segmentation identifier, "8001" is the computing task data interface, "8002" is the computing result receiving interface, the "image processing" offloading task type code value is "03", add "A" as the code segmentation identifier, "8088" is the computing task data interface, and "8080" is the computing result receiving interface. Node N1 encapsulates the SRv6 extension header containing the above data to obtain an offloading notification data packet and sends the offloading notification data packet to the computing offloading module. The offloading notification data packet is sent to the computing power resource nodes P1 and P2 in sequence according to a forwarding path by the computing offloading module. Node P1 parses the TLV content in the offloading notification data packet to obtain the computing task data interface "8001", obtains the corresponding offloading task data from node N1, sends the response result to the computing result receiving interface "8002" after completing the computing offloading task, and node N1 closes the "8001" and "8002" interfaces. Then, node P1 forwards the data packet to node P2. Node P2 parses the TLV content in the computing offloading notification data packet to obtain the computing task data interface "8088", obtains the offloading task data from node N1, sends the response result to the computing result receiving interface "8080" after completing the computing offloading task, and node N1 closes the "8088" and "8080" interfaces. The waiting offloading node N1 receives all the response results and informs the computing offloading module that the computing power resource extension is completed.

[0141] By using the technical means of the embodiments of the present disclosure, if the purchased computing power resources of a user do not meet the actual computing power demand in the process of using the computing power resources, resource extension can be performed according to the use needs of the user, professional extension strategies are provided for the normal operation of the business application of the user, and there is no need to repurchase other computing power resources and perform data and application migration. Specifically, the computing power resource extension is based on SRv6, the state of node application running can be analyzed in real time in the process of using the computing power resources, the SRv6 computing offloading notification data packet is encapsulated for offloading task notification for the detachable offloading business type, other nodes are enabled to complete the current task in cooperation with the waiting offloading node, the communication between multiple nodes is completed through a single data packet forwarding path, it is convenient and efficient, task allocation is performed for different offloading task types to ensure the effective completion of the offloading task, and the situation that the computing power resource bottleneck and reasons are not visible to the user and the like are avoided to cause the effective resource extension.

[0142] Referring to FIG. 11, it is a first flow diagram of a computing power resource management method provided by the embodiments of the present disclosure. The embodiments of the present disclosure further provide a computing power resource management method, which is executed by the computing power resource management system 10 in the above embodiments. The computing power resource management system includes a resource demand platform, a proxy gateway, and a computing power resource pool. The method includes steps S11 to S13:

[0143] S11, the resource demand platform acquires the demand description text input by a user;

[0144] S12, the resource demand platform parses the business type and the computing resource parameter of the user demand according to the demand description text;

[0145] S13, the proxy gateway acquires corresponding computing resource nodes in the computing resource pool that the proxy gateway proxies according to the business type and the computing resource parameter, configures the computing resource nodes, and generates the access address of the computing resource nodes.

[0146] As a preferred embodiment, the demand description text is a business type description text or a resource parameter description text. Then, step S12, i.e., the parsing of the business type and the computing resource parameter of the user demand according to the demand description text, includes steps S121 to S124:

[0147] S121, when the demand description text is a business type description text, the business type description text is parsed to obtain the business type;

[0148] S122, a preset business knowledge base is searched to acquire the computing resource parameter corresponding to the business type; wherein, the preset business knowledge base records the corresponding relationship between the business type and the computing resource parameter;

[0149] S123, when the demand description text is a resource parameter description text, the resource parameter description text is parsed to obtain the computing resource parameter;

[0150] S124, the business knowledge base is searched to acquire the business type corresponding to the computing resource parameter.

[0151] As a preferred embodiment, step S13, i.e., the proxy gateway acquires corresponding computing resource nodes in the computing resource pool that the proxy gateway proxies according to the business type and the computing resource parameter, configures the computing resource nodes, and generates the access address of the computing resource nodes, includes steps S131 to S133:

[0152] S131, the corresponding computing resource pool is determined according to the business type;

[0153] S132, the corresponding computing resource node in the computing resource pool is matched according to the computing resource parameter;

[0154] S133, the access address of the computing resource node is generated.

[0155] As a preferred implementation, the resource demand platform can also display the demand description text, the business type, the computing power resource parameter, and the access address of the computing power resource node.

[0156] By using the technical means of the embodiments of the present disclosure, the computing power demand is analyzed based on natural language, the business type description and the resource parameter description are supported, and the user can describe the required business type or resource parameter in natural language. Through the business knowledge base and in combination with the professional computing power resource application scene experience, bidirectional speculation between the resource parameter and the business type is performed. For a user who clearly knows the computing power resource parameter, a computing power resource consistent with the computing power resource parameter can be provided. For a user who can only provide a business type scene description, more scene related information can be provided to assist the user in selection, so that the user is recommended and provided with a suitable computing power resource, and more perfect computing power demand analysis and more efficient resource utilization are achieved.

[0157] Referring to FIG. 12, it is a second flowchart of a computing power resource management method in the embodiments of the present disclosure. The embodiments of the present disclosure are further implemented on the basis of the above-mentioned embodiments. The method further includes steps S21 to S23:

[0158] S21, the proxy gateway monitors the running state information of each computing power resource node; wherein the running state information is a normal running state or an abnormal running state;

[0159] S22, the proxy gateway acquires the computing power resource node in the abnormal running state as a to-be-unloaded node;

[0160] S23, the proxy gateway allocates the configured business of the to-be-unloaded node to other allocable computing power resource nodes.

[0161] Step S23, that is, the allocation of the configured business of the to-be-unloaded node to other allocable computing power resource nodes, includes steps S231 to S234:

[0162] S231, acquiring the computing power resource node in the abnormal running state as a to-be-unloaded node;

[0163] S232, analyzing the configured business of the to-be-unloaded node, denoted as an unloading task;

[0164] S233, determining other allocable computing power resource nodes according to the business type of the unloading task as allocable computing power resource nodes, and acquiring the addresses of the allocable computing power resource nodes;

[0165] S234, a deinstallation notification data packet is generated, and the deinstallation notification data packet is sent to each of the distributable computing resource nodes according to the address, so that the distributable computing resource nodes obtain the deinstallation task allocated to the distributable computing resource nodes through the deinstallation notification data packet and return a response result.

[0166] Preferably, the deinstallation notification data packet is composed of data fields corresponding to each of the distributable computing resource nodes, and each of the data fields includes an encoding bit, a first interface bit, and a second interface bit.

[0167] The encoding bit is composed of an encoding segmentation identifier and an encoding value of the deinstallation task, the first interface bit is used to provide a data interface for the distributable computing resource node to obtain the deinstallation task allocated to the distributable computing resource node, and the second interface bit is used to provide an interface for the distributable computing resource node to return a response result after obtaining the deinstallation task.

[0168] By using the technical means of the embodiments of the present disclosure, if the purchased computing resource of a user does not meet the actual computing resource demand in the process of using the computing resource by the user, resource expansion can be performed according to the use needs of the user, professional expansion strategies are provided for normal operation of the business application of the user, and there is no need to purchase other computing resources and perform data and application migration.

[0169] It should be noted that all process steps of the computing resource management method provided in the embodiments of the present disclosure are executed by the computing resource management system in the above embodiments, and the working principles and beneficial effects of the two are one-to-one correspondence, and thus will not be repeated.

[0170] The embodiments of the present disclosure also provide a computing resource management system, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the computing resource management method according to any one of the above embodiments when executing the computer program.

[0171] It should be noted that the computing resource management system provided in the embodiments of the present disclosure is used to execute all process steps of the computing resource management method in the above embodiments, and the working principles and beneficial effects of the two are one-to-one correspondence, and thus will not be repeated.

[0172] The embodiments of the present disclosure also provide a resource demand platform, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the steps executed by the resource demand platform in the computing resource management method according to any one of the above embodiments when executing the computer program.

[0173] The embodiment of the present disclosure further provides a proxy gateway, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, when the computer program is executed by the processor, steps performed by the proxy gateway in the computing resource management method according to any one of the above embodiments are implemented.

[0174] The embodiment of the present disclosure further provides a computing resource pool, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, when the computer program is executed by the processor, steps performed by the computing resource pool in the computing resource management method according to any one of the above embodiments are implemented.

[0175] The embodiment of the present disclosure further provides a computer readable storage medium, comprising a stored computer program, wherein when the computer program is running, the computer readable storage medium controls a device where the computer readable storage medium is located to perform the computing resource management method according to any one of the above embodiments.

[0176] The embodiment of the present disclosure further provides a computer program product, comprising a computer program or computer instructions, when the computer program or the computer instructions are executed by a processor, the computing resource management method according to any one of the above embodiments is implemented.

[0177] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiments can be included. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM) or a random access memory (RAM) and the like.

[0178] The above is the preferred embodiment of the present disclosure, and it should be pointed out that for those skilled in the art, without departing from the principle of the present disclosure, a number of improvements and refinements can be made, and these improvements and refinements are also considered to be within the protection scope of the present disclosure.

Claims

1. A hash power resource management system, wherein, The resource demand platform, the proxy gateway and the computing resource pool are included. The resource demand platform is configured to acquire a demand description text input by a user, and parse a business type and a computing resource parameter of a user demand according to the demand description text. The proxy gateway is configured to acquire corresponding computing resource nodes in the proxy computing resource pool according to the business type and the computing resource parameter, configure the computing resource nodes, and generate access addresses of the computing resource nodes.

2. The computing resource management system of claim 1, wherein, The resource demand platform includes a resource demand platform interface, a business type recommendation module and a parameter analysis module. The resource demand platform interface is configured to acquire a demand description text input by a user, wherein the demand description text is a business type description text or a resource parameter description text. When the demand description text is the business type description text, the business type recommendation module is configured to parse the business type description text to obtain a business type, and search a preset business knowledge base to acquire a computing resource parameter corresponding to the business type, wherein the preset business knowledge base records a corresponding relationship between a business type and a computing resource parameter. When the demand description text is the resource parameter description text, the parameter analysis module is configured to parse the resource parameter description text to obtain a computing resource parameter, and send the computing resource parameter to the business type recommendation module. The business type recommendation module is further configured to search the business knowledge base to acquire a business type corresponding to the computing resource parameter.

3. The computing resource management system of claim 2, wherein, The resource demand platform further includes an address coding module, and the proxy gateway includes a computing resource configuration module. The address coding module is configured to determine an address of a corresponding proxy gateway according to the business type to establish a communication link between the resource demand platform and the proxy gateway. The business type recommendation module is further configured to send the computing resource parameter to the computing resource configuration module. The computing resource configuration module is configured to match corresponding computing resource nodes in the proxy computing resource pool according to the computing resource parameter, and generate access addresses of the computing resource nodes.

4. The computing resource management system of claim 3, wherein, The resource demand platform interface is further configured to display the demand description text, the business type, the computing resource parameter and the access addresses of the computing resource nodes.

5. The computing resource management system of any one of claims 1 to 4, wherein, The proxy gateway further includes a resource monitoring module and a computing offloading module. The resource monitoring module is configured to monitor running state information of each computing resource node, wherein the running state information is a normal running state or an abnormal running state. The computing offloading module is configured to acquire an abnormal running state computing resource node as a to-be-offloaded node, and assign configured businesses of the to-be-offloaded node to other assignable computing resource nodes.

6. The computing resource management system of claim 5, wherein, The computing offloading module is specifically configured to: acquire an abnormal running state computing resource node as a to-be-offloaded node; analyze configured businesses of the to-be-offloaded node, and record the configured businesses as offloading tasks; determine other assignable computing resource nodes as assignable computing resource nodes according to business types of the offloading tasks, and acquire addresses of the assignable computing resource nodes; and assign the offloading tasks to the assignable computing resource nodes according to the addresses of the assignable computing resource nodes. Generate an offloading notification data packet, and send the offloading notification data packet to each of the allocable computing resource nodes according to the address, so that the allocable computing resource nodes obtain the offloading tasks allocated to them through the offloading notification data packet and return a response result.

7. The computing resource management system of claim 6, wherein, The offloading notification data packet is composed of data fields corresponding to each of the allocable computing resource nodes, and each of the data fields includes an encoding bit, a first interface bit, and a second interface bit. The encoding bit is composed of an encoding segmentation identifier and an encoding value of the offloading task, the first interface bit is used to provide a data interface for the allocable computing resource node to obtain the offloading task allocated to it, and the second interface bit is used to provide an interface for the allocable computing resource node to return a response result after obtaining the offloading task.

8. A computing resource management method, executed by a computing resource management system including a resource demand platform, a proxy gateway, and a computing resource pool, comprising: The resource demand platform obtains a demand description text input by a user. The resource demand platform parses the demand description text to obtain a business type and a computing resource parameter of a user demand. The proxy gateway configures corresponding computing resource nodes in the computing resource pool that the proxy gateway proxies according to the business type and the computing resource parameter, and generates an access address of the computing resource nodes.

9. The computing resource management method of claim 8, wherein, The demand description text is a business type description text or a resource parameter description text. The resource demand platform parses the demand description text to obtain a business type and a computing resource parameter of a user demand, including: When the demand description text is a business type description text, parse the business type description text to obtain a business type. Find a preset business knowledge base to obtain a computing resource parameter corresponding to the business type; wherein the preset business knowledge base records a correspondence between a business type and a computing resource parameter. When the demand description text is a resource parameter description text, parse the resource parameter description text to obtain a computing resource parameter. Find the business knowledge base to obtain a business type corresponding to the computing resource parameter.

10. The computing resource management method of claim 8, wherein, The proxy gateway configures corresponding computing resource nodes in the computing resource pool that the proxy gateway proxies according to the business type and the computing resource parameter, and generates an access address of the computing resource nodes, including: Determine a corresponding computing resource pool according to the business type. Match a corresponding computing resource node in the computing resource pool according to the computing resource parameter. Generate an access address of the computing resource node.

11. The computing resource management method of claim 8, wherein, The method further comprises: Display the demand description text, the business type, the computing resource parameter, and the access address of the computing resource node.

12. The computing resource management method of any one of claims 8 to 11, wherein, The method further comprises: The proxy gateway monitors running state information of each of the computing resource nodes; wherein the running state information is a normal running state or an abnormal running state. The proxy gateway obtains a computing resource node in an abnormal running state as a to-be-offloaded node. The proxy gateway allocates the configured service of the to-be-unloaded node to other allocable computing resource nodes.

13. The computing resource management method of claim 12, wherein, The method further comprises: The proxy gateway acquires a computing resource node in an abnormal operating state as a to-be-unloaded node; Analyzed is the configured service of the to-be-unloaded node, denoted as an unloading task; According to the service type of the unloading task, other allocable computing resource nodes are determined as allocable computing resource nodes, and the addresses of the allocable computing resource nodes are acquired; An unloading notification data packet is generated, and the unloading notification data packet is sent to each allocable computing resource node according to the addresses, so that the allocable computing resource nodes acquire the unloading tasks allocated thereto through the unloading notification data packet and return response results.

14. A computing resource management system comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein, The processor executes the computer program to implement the computing resource management method in any one of claims 8 to 13.

15. A computer readable storage medium, wherein, The computer readable storage medium comprises a stored computer program, wherein the computer program controls the device where the computer readable storage medium is located to execute the computing resource management method in any one of claims 8 to 13 when the computer program is running.

16. A computer program product, wherein, The computer program product comprises a computer program or computer instructions, which, when executed by a processor, implement the computing resource management method in any one of claims 8 to 13.

Images