A network device computing power information collection and management method and system

Abstract

The present disclosure belongs to the technical field of communication, and provides a network device computing power information collection and management method and system. The method comprises the following steps: a computing power information collection module cooperates with a computing power collection agent to acquire and process computing power information from a local computing power cluster, and delivers the processed computing power information to a computing power management module; a computing power routing module receives updated computing power routing information announced by a remote device, converts the computing power routing information into computing power information, and delivers the computing power information to the computing power management module; the computing power management module receives the computing power information and stores the computing power information in a computing power information table; the computing power information table is monitored for computing power information changes, and local computing power change information is delivered to the computing power routing module; the computing power routing module dynamically generates or updates a computing power routing table based on the computing power change information, and obtains a current local computing power routing table; and the current local computing power routing table is used to announce local computing power routing to a remote device. The local computing power information and the computing power information announced by the remote device can be managed simultaneously, and the management and use efficiency of the computing power information is improved.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the technical field of communication, and particularly relates to a network device computing power information collection and management method and system. BACKGROUND

[0002] With the continuous progress of cloud computing and big data technology, the effective management and scheduling of computing power information has become increasingly critical. However, current network devices generally lack the ability to manage computing power information, which not only limits the full utilization of computing power resources within the network, but also reduces overall computing efficiency.

[0003] In a computing power network, the perception and processing of computing power information are the basis for realizing the optimal allocation of computing power resources and the calculation of traffic forwarding paths. This involves comprehensive management of local and remote computing power information. Current network devices mainly focus on data packet forwarding and control, and lack in computing power information management. Therefore, network devices urgently need a method and system for realizing network device computing power information collection and management. SUMMARY

[0004] To solve the above problems, the present disclosure provides a network device computing power information collection and management method and system, which can simultaneously manage local computing power information and remote announced computing power information through the coordinated operation of various modules, thereby improving the management and use efficiency of computing power information.

[0005] In a first aspect, a network device computing power information collection and management method is provided, comprising:

[0006] The computing power information collection module acquires and processes computing power information from a local computing power cluster according to a local computing power collection task issued by the computing power management module, in accordance with the computing power information type and format defined by the computing power management module, and transfers the processed computing power information to the computing power management module, wherein the computing power information collection module is deployed on the network device, and the computing power collection agent is deployed on the computing power nodes of the computing power cluster;

[0007] The computing power routing module receives updated computing power routing information announced by a remote device, converts the updated computing power routing information into computing power information, and transfers the converted computing power information to the computing power management module;

[0008] The computing power management module receives the processed computing power information transferred by the computing power information collection module and the converted computing power information transferred by the computing power routing module, and stores the processed computing power information and the converted computing power information in a computing power information table;

[0009] The computing power management module monitors computing power changes based on the local computing power information in the computing power information table, and transfers the local computing power change information to the computing power routing module;

[0010] The computing power routing module dynamically generates or updates the computing power routing table based on the local computing power change information to obtain a current local computing power routing table, and announces the local computing power routing to the remote device based on the current local computing power routing table.

[0011] Further, the updated computing power routing information is converted into computing power information, including:

[0012] The computing power routing information is converted into the computing power information by using a field mapping relationship between the computing power routing table and the computing power information table, wherein the field mapping relationship is that a Destination field in the computing power routing table is mapped to an S_IP field in the computing power information table, an APP field in the computing power routing table is mapped to an S_NAME field in the computing power information table, bgp of a TYPE field in the computing power routing table is mapped to remote in a Site field in the computing power information table, and other fields in the computing power routing table are mapped to other fields in the computing power information table in one-to-one mapping, and the other fields include CPU and / or GPU.

[0013] Further, the computing power routing module dynamically generates or updates the computing power routing table based on the local computing power change information to obtain a current local computing power routing table, including:

[0014] The computing power routing module converts the computing power change information into the computing power routing information by using a field mapping relationship between the computing power routing table and the computing power information table, and dynamically generates or updates the computing power routing table by using the converted computing power routing information to obtain a current local computing power routing table, wherein the field mapping relationship is that an S_IP field in the computing power information table is mapped to a Destination field in the computing power routing table, an S_NAME field in the computing power information table is mapped to an APP field in the computing power routing table, local in a Site field in the computing power information table is mapped to direct of a TYPE field in the computing power routing table, and other fields in the computing power information table are mapped to other fields in the computing power routing table in one-to-one mapping, and the other fields include CPU and / or GPU.

[0015] Further, the defined computing power information type and format include a unique identification field with a data type of String, a service type field with a data type of String, and a resource type field with a data type of String.

[0016] Further, before the computing power management module transmits the local computing power change information to the computing power routing module, the method further includes:

[0017] The computing power management module calculates a current local computing power change value by using a formula A' = T | (C-C') |, wherein A' is the current local computing power change value, T is a change rate, C' is a changed total amount of computing power, C is a current local total amount of computing power, and different total amounts of computing power correspond to different change rates.

[0018] The computing power management module determines whether the current local computing power change value exceeds the defined computing power change threshold value;

[0019] The computing power management module transmits the computing power change information to the computing power routing module, including:

[0020] The computing power management module transmits the local computing power change information to the computing power routing module when the current local computing power change value exceeds the defined computing power change threshold value.

[0021] Further, the processed computing power information is transmitted to the computing power management module, including:

[0022] The processed computing power information is transmitted to the computing power management module through the gRPC protocol, the SNMP protocol or the Prometheus toolkit.

[0023] Further, the computing power routing module receives the updated computing power routing information announced by the remote device, including:

[0024] The computing power routing module receives the updated computing power routing information announced by the remote device through linkage with the BGP routing protocol; or

[0025] The computing power routing module receives the updated computing power routing information announced by the remote device through linkage with the SDN controller.

[0026] In a second aspect, a network device computing power information collection and management system is provided, including: a computing power information collection module, a computing power routing module, a computing power management module and a computing power information table; wherein:

[0027] The computing power information collection module is configured to acquire and process computing power information from the local computing power cluster according to the local computing power collection task issued by the computing power management module in accordance with the computing power information type and format defined by the computing power management module, and transmit the processed computing power information to the computing power management module, wherein the computing power information collection module is deployed on the network device, and the computing power collection agent is deployed on the computing power node of the computing power cluster;

[0028] The computing power routing module is configured to receive the updated computing power routing information announced by the remote device, convert the updated computing power routing information into computing power information, transmit the converted computing power information to the computing power management module, and locally generate or update the computing power routing table based on the computing power change information to obtain the current local computing power routing table; and announce the local computing power routing to the remote device based on the current local computing power routing table.

[0029] The computing power management module is configured to receive the processed computing power information transmitted by the computing power information collection module and the converted computing power information transmitted by the computing power routing module, and store the processed computing power information and the converted computing power information in a computing power information table; based on the local computing power information in the computing power information table, monitor the computing power change, and transmit the local computing power change information to the computing power routing module.

[0030] The computing power information table is configured to store the processed computing power information and the converted computing power information.

[0031] Further, the computing power routing module is specifically configured to convert the computing power routing information into the computing power information by using a field mapping relationship between the computing power routing table and the computing power information table; wherein the field mapping relationship is that a Destination field in the computing power routing table is mapped to an S_IP field in the computing power information table; an APP field in the computing power routing table is mapped to an S_NAME field in the computing power information table; a bgp of a TYPE field in the computing power routing table is mapped to a remote in a Site field in the computing power information table; and other fields in the computing power routing table are mapped to other fields in the computing power information table in a one-to-one manner, and the other fields include CPU and / or GPU.

[0032] Further, the computing power routing module is specifically configured to convert the computing power change information into the computing power routing information by using the field mapping relationship between the computing power routing table and the computing power information table, and generate or update the computing power routing table dynamically by using the converted computing power routing information to obtain a current local computing power routing table; wherein the field mapping relationship is that an S_IP field in the computing power information table is mapped to a Destination field in the computing power routing table; an S_NAME field in the computing power information table is mapped to an APP field in the computing power routing table; a local in a Site field in the computing power information table is mapped to a direct of a TYPE field in the computing power routing table; and other fields in the computing power information table are mapped to other fields in the computing power routing table in a one-to-one manner, and the other fields include CPU and / or GPU.

[0033] In a third aspect, an electronic device is provided, which includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus;

[0034] The memory is configured to store a computer program.

[0035] The processor is configured to execute the program stored in the memory to implement the method described above.

[0036] In a fourth aspect, a computer storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the method described above.

[0037] Compared with the prior art, the present disclosure has the following advantages:

[0038] Local computing power information and remote announced computing power information can be managed simultaneously, and the management and use efficiency of computing power information is improved through integration of all information resources. Basic information is provided for network device computing power information evaluation and user service discovery.

[0039] Other features and advantages of the present disclosure will be described in the following description, and some will become apparent from the description, or will be understood through implementation of the present disclosure. The purpose and other advantages of the present disclosure can be achieved and obtained through the structure indicated in the specification, claims and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0041] Figure 1 A network device computing power management method according to an embodiment of the present disclosure is shown in the schematic diagram;

[0042] Figure 2 A schematic diagram of the interaction between the computing power collection module and the computing power management module to realize local computing power collection according to an embodiment of the present disclosure is shown;

[0043] Figure 3 A schematic diagram of the interaction between the computing power routing module and the computing power management module to realize remote computing power collection according to an embodiment of the present disclosure is shown;

[0044] Figure 4 A process schematic diagram of the computing power information change exceeding the threshold value triggering the computing power information transmission in the gRPC channel according to an embodiment of the present disclosure is shown;

[0045] Figure 5 A networking topology diagram according to an embodiment of the present disclosure is shown;

[0046] Figure 6 A network device computing power management system according to an embodiment of the present disclosure is shown in the schematic diagram. DETAILED DESCRIPTION

[0047] In computing power networks, network devices need to be aware of both local and remote computing power information. How to centrally and efficiently manage this information is a current challenge. This invention provides a method for collecting and managing computing power information. By linking with routing protocols, it identifies computing power information advertised by these protocols, thereby enabling network devices to manage remote computing power information. This provides foundational information for network devices to evaluate computing power information and discover user services. This method can simultaneously manage both local and remotely advertised computing power information, improving the efficiency of computing power information management and utilization through the integration of all information resources.

[0048] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0049] like Figure 1 The diagram shown is a schematic representation of a network device computing power management method according to an embodiment of this disclosure. The network device computing power information collection and management method provided in this embodiment includes the following steps:

[0050] Step 101: The computing power information acquisition module collects local computing power information and transmits it to the computing power management module.

[0051] Specifically, the computing power information acquisition module, according to the computing power information type and format defined by the computing power management module, and based on the local computing power acquisition task issued by the computing power management module, collaborates with the computing power acquisition agent to obtain and process computing power information from the local computing power cluster, and then transmits the processed computing power information to the computing power management module.

[0052] The computing power information collection module is deployed on network devices, while the computing power collection agent is deployed on the computing power nodes of the computing power cluster.

[0053] To achieve unified management of computing power information, the computing power management module can define the format of computing power information requests as shown in Table 1 below.

[0054]

[0055] The computing power acquisition module performs local computing power acquisition according to the tasks issued by the computing power management module. The computing power acquisition module consists of computing power acquisition modules deployed on network devices and computing power acquisition agents deployed on computing power cluster nodes. The modules use gRPC (Google Remote Procedure Call) channels to transmit information. Figure 2Figure 1 is a schematic diagram of information interaction between modules, comprising the following steps:

[0056] 1. The computing power management module defines the type and format of the collected computing power information.

[0057] 2. The computing power collection module located at the network device sends a computing power collection request to the computing power collection agent located at the computing power cluster node using the GRPC method.

[0058] 3. The computing power collection agent located at the computing power cluster node responds to the computing power collection request.

[0059] 4. The computing power collection agent located at the computing power cluster node periodically sends the collected computing power information to the computing power collection module.

[0060] 5. The computing power collection module located at the network device confirms the reporting of computing power information.

[0061] 6. The computing power collection module located at the network device processes the new computing power information and transmits the computing power information to the computing power management module.

[0062] The computing power information collection gRPC data format can be as follows:

[0063] {

[0064] "cpu": {

[0065] "model": "Intel Core i7-11700K",

[0066] "cores": 8,

[0067] "threads": 16,

[0068] "base_clock": 3.6,

[0069] "max_boost_clock": 5.0,

[0070] "cache": 16

[0071] },

[0072] "gpu": {

[0073] "model": "NVIDIA GeForce RTX 3070",

[0074] "vram": 8,

[0075] "core_clock": 1725,

[0076] "boost_clock": 1815,

[0077] "cuda_cores": 5888

[0078] }

[0079] }

[0080] The above local computing power collection process uses gRPC protocol, which provides a transmission channel for the collection of computing power information, but does not directly intervene in the computing power collection process. gRPC can also be used to extend related protocols and fields to implement the collection of computing power information and other resources.

[0081] SNMP (Simple Network Management Protocol) can also be used to collect computing power information. Prometheus toolkit can also be used to collect computing power information, which is an open source system monitoring and alerting toolkit that can also be used for computing power information collection, especially in containerization and microservices architecture.

[0082] Step 102: The computing power routing module delivers computing power information to the computing power management module based on the updated computing power routing information announced by the remote device.

[0083] Specifically, the computing power routing module receives the updated computing power routing information announced by the remote device, converts the updated computing power routing information into computing power information, and delivers the converted computing power information to the computing power management module.

[0084] In this step 102, the computing power management module and the computing power routing module interact to realize the conversion of computing power information and computing power routing.

[0085] The computing power information table design can be as shown in Table 2, and the computing power routing table design can be as shown in Table 3:

[0086]

[0087] Specifically, when the network device receives the computing power routing update announced by the remote device, it converts the computing power routing into computing power information and delivers it to the computing power management module using the gRPC channel. This process is shown in Figure 3 , including:

[0088] 1. The computing power routing module receives the computing power routing update;

[0089] 2. The computing power routing module converts the computing power routing information into computing power information for delivery;

[0090] 3. The computing power management module stores the computing power information in the computing power information table;

[0091] 4. The computing power management module confirms the computing power transmission operation.

[0092] The mapping relationship between the fields of the computing power routing table and the computing power information table is as follows:

[0093] The Destination field in the computing power routing table is mapped to the S_IP field in the computing power information table.

[0094] The APP field in the computing power routing table is mapped to the S_NAME field in the computing power information table.

[0095] The TYPE field in the computing power routing table is mapped to the remote in the computing power information table.

[0096] The CPU, GPU, and other fields in the computing power routing table are one-to-one mapped to the fields in the computing power information table.

[0097] The above process of obtaining computing power information of the remote device is implemented using the method of BGP routing protocol linkage, and the local computing power is announced to the remote and the remote information is announced to the local through the BGP routing protocol linkage. The method of SDN controller linkage can also be used to achieve this. This method uses the flexibility of software-defined networks to achieve the announcement of computing power information.

[0098] In the above steps 101 and 102, these alternatives provide flexibility and selectivity, so that the computing power information management system can select the most suitable collection and announcement method according to the specific application scenario and network environment, thereby optimizing the management and use efficiency of computing power resources.

[0099] It should be noted that the above steps 101 and 102 have no execution relationship, but are executed in parallel.

[0100] Step 103: The computing power management module receives the processed computing power information and the converted computing power information, and stores them in the computing power information table.

[0101] Specifically, the computing power management module receives the processed computing power information transmitted by the computing power information collection module and the converted computing power information transmitted by the computing power routing module, and stores the processed computing power information and the converted computing power information in the computing power information table. The computing power management module is deployed on a network device.

[0102] Step 104: The computing power management module monitors the computing power changes based on the local computing power information in the computing power information table, and transmits the local computing power change information to the computing power routing module.

[0103] The computing power management module of the network device transmits the local computing power information to the computing power routing module to reduce the shock of computing power routing and improve transmission efficiency. Before the computing power management module transmits the computing power change information to the computing power routing module, the method further comprises:

[0104] The computing power management module calculates the current local computing power change value based on the monitored computing power change by using the formula A' = T|(C-C')|; wherein A' is the current local computing power change value, T is the change rate, C' is the changed total computing power, and C is the current local total computing power. Different total computing power ranges correspond to different change rates.

[0105] The computing power management module determines whether the current local computing power change value exceeds the defined computing power change threshold.

[0106] At this time, the step 104 specifically comprises: the computing power management module transmits the computing power change information to the computing power routing module when the current local computing power change value exceeds the defined computing power change threshold.

[0107] Here, the computing power management module defines the computing power change threshold for different resource conditions of computing power information. When the computing power information parameter change exceeds the defined computing power change threshold, the computing power management module transmits the computing power change information to the computing power routing module. If the computing power information parameter change does not exceed the computing power change threshold, the computing power parameter change is not transmitted.

[0108] It should be noted that the computing power management module transmits the local computing power information to the computing power routing module, and the computing power routing module transmits the computing power obtained by the bgp neighbor to the computing power management module. Here, to avoid the flooding effect, the transmission of computing power is unidirectional, and the computing power information transmitted from the remote device is no longer transmitted to the remote device, but the local computing power information can be transmitted to the remote device.

[0109] The process of transmitting the computing power information in the gRPC channel triggered by the change of the local computing power information exceeding the computing power change threshold is as shown in Figure 4 .

[0110] 1. The computing power management module defines the computing power change threshold.

[0111] 2. The computing power management module monitors the change of the local computing power information table.

[0112] 3. The local computing power information is transmitted when the computing power information change exceeds the computing power change threshold.

[0113] 4. The computing power routing module updates the computing power routing table based on the transmitted computing power information by using the mapping relationship between the computing power information table and the computing power routing table field.

[0114] 5. The computing power routing module confirms the computing power information transmission.

[0115] Mapping relationship between the field of the computing power information table and the field of the computing power routing table:

[0116] The S_IP field in the computing power information table is mapped to the Destination field in the computing power routing table.

[0117] The S_NAME field in the computing power information table is mapped to the APP field in the computing power routing table.

[0118] The local in the Site field in the computing power information table is mapped to the direct in the TYPE field in the computing power routing table.

[0119] The CPU, GPU and other fields in the computing power information table are one-to-one mapped to the fields in the computing power routing table.

[0120] Step 105: The computing power routing module dynamically generates or updates the computing power routing table based on the local computing power change information to obtain a current local computing power routing table, and announces the local computing power routing based on the current local computing power routing table.

[0121] In the computing power information management system of the application, various alternative solutions are provided for the collection of local computing power information and the announcement of remote computing power information to adapt to different network environments and technical requirements.

[0122] The following describes the scheme of the embodiment of the disclosure by taking the networking topology diagram shown in FIG. 1 as an example. Figure 5

[0123] Figure 5 The scheme includes the following three steps:

[0124] 1. The computing power management module located at the network device defines the computing power collection parameters and the computing power change threshold for the announcement of the computing power.

[0125] 2. The computing power collection module located at the network device 1 is responsible for collecting the computing power information of the computing power node 1 and the computing power node 2 under the node, and the computing power collection module located at the network device 2 is responsible for collecting the computing power information of the computing power node 3 under the node.

[0126] 3. The computing power routing modules of the network devices announce the computing power routing of each other.

[0127] The computing power information table of the network device 1 is shown in Table 4 as follows:

[0128]

[0129] The computing power routing table of the network device 1 is shown in Table 5 as follows:

[0130]

[0131] The computing power information table of the network device 2 is shown in Table 6 as follows: ​

[0132]

[0133] The network device 2 computing power routing table is as shown in Table 7:

[0134] Based on the above method, the network device computing power information collection and management system corresponding to the above method is also provided, Figure 6 The network device computing power information collection and management system according to the embodiment of the present disclosure is shown in the schematic diagram, which comprises a computing power collection module, a computing power routing module, a computing power management module and a computing power information table, wherein:

[0135] The computing power information collection module is configured to collect and process computing power information from the local computing power cluster according to the local computing power collection task issued by the computing power management module and in cooperation with the computing power collection agent according to the computing power information type and format defined by the computing power management module, and transmit the processed computing power information to the computing power management module, wherein the computing power information collection module is deployed on the network device, and the computing power collection agent is deployed on the computing power node of the computing power cluster;

[0136] The computing power routing module is configured to receive the updated computing power routing information announced by the remote device, convert the updated computing power routing information into computing power information, transmit the converted computing power information to the computing power management module, dynamically generate or update the computing power routing table based on the local computing power change information to obtain the current local computing power routing table, and announce the local computing power routing to the remote device based on the current local computing power routing table.

[0137] The computing power management module is configured to receive the processed computing power information transmitted by the computing power information collection module and the converted computing power information transmitted by the computing power routing module, store the processed computing power information and the converted computing power information into the computing power information table, maintain the computing power information table, monitor the computing power change based on the local computing power information in the computing power information table, and transmit the local computing power change information to the computing power routing module.

[0138] The computing power information table is configured to store the processed computing power information and the converted computing power information for subsequent retrieval and use.

[0139] Specifically, the computing power management module provides a northbound interface for a controller or other third-party platform to call, so as to define global parameters on the controller and issue them to the network device computing power management module.

[0140] Through the cooperative operation of the above modules, the network device can effectively manage and use the local and remote computing power resources, thereby optimizing the user service discovery process and improving the overall network performance.

[0141] Further, the computing power routing module is specifically configured to convert the computing power routing information into the computing power information by using a field mapping relationship between the computing power routing table and the computing power information table, wherein the field mapping relationship is that a Destination field in the computing power routing table is mapped to an S_IP field in the computing power information table, an APP field in the computing power routing table is mapped to an S_NAME field in the computing power information table, bgp of a TYPE field in the computing power routing table is mapped to remote in a Site field in the computing power information table, and other fields in the computing power routing table are one-to-one mapped to other fields in the computing power information table, and the other fields include CPU and / or GPU.

[0142] Further, the computing power routing module is specifically configured to convert the computing power routing information into the computing power information by using a field mapping relationship between the computing power routing table and the computing power information table, wherein the field mapping relationship is that a Destination field in the computing power routing table is mapped to an S_IP field in the computing power information table, an APP field in the computing power routing table is mapped to an S_NAME field in the computing power information table, bgp of a TYPE field in the computing power routing table is mapped to remote in a Site field in the computing power information table, and other fields in the computing power routing table are one-to-one mapped to other fields in the computing power information table, and the other fields include CPU and / or GPU.

[0143] Based on the same inventive concept as disclosed above, the disclosure also provides an electronic device. The electronic device of the embodiment of the disclosure includes at least one processor and at least one memory electrically connected, wherein the memory is electrically connected with the processor, and the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method as described above.

[0144] It should be noted that the electrical connection between the above-mentioned various units does not necessarily mean the connection between the lines, and the indirect connection mode can also be applicable to the embodiment of the disclosure as long as the purpose of the disclosure is achieved.

[0145] Based on the same inventive concept, the disclosure also provides a computer storage medium, wherein the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method as described above.

[0146] Although the present disclosure is explained in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalent features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for collecting and managing computing power information of a network device, characterized in that, The method comprises the following steps: The computing power information collection module collects computing power information according to a computing power information type and a format defined by the computing power management module and a local computing power collection task issued by the computing power management module, and obtains and processes the computing power information from a local computing power cluster in cooperation with a computing power collection agent, and delivers the processed computing power information to the computing power management module, wherein the computing power information collection module is deployed on a network device, and the computing power collection agent is deployed on a computing power node of the computing power cluster; The computing power routing module receives updated computing power routing information announced by a remote device, converts the updated computing power routing information into computing power information, and delivers the converted computing power information to the computing power management module; The computing power management module receives the processed computing power information delivered by the computing power information collection module and the converted computing power information delivered by the computing power routing module, and stores the processed computing power information and the converted computing power information in a computing power information table; The computing power management module monitors computing power changes based on local computing power information in the computing power information table, and delivers local computing power change information to the computing power routing module; The computing power routing module dynamically generates or updates a computing power routing table based on the local computing power change information, obtains a current local computing power routing table, and announces local computing power routing to the remote device based on the current local computing power routing table; The method for converting the updated computing power routing information into computing power information comprises the following steps: The computing power routing table and the computing power information table are mapped by field, and the computing power routing information is converted into computing power information, wherein the Destination field in the computing power routing table is mapped into the S_IP field in the computing power information table, the APP field in the computing power routing table is mapped into the S_NAME field in the computing power information table, the TYPE field bgp in the computing power routing table is mapped into remote in the Site field in the computing power information table, and other fields in the computing power routing table are one-to-one mapped into other fields in the computing power information table, and the other fields include CPU and / or GPU.

2. The method of claim 1, wherein, The computing power routing module dynamically generates or updates the computing power routing table based on the local computing power change information, and obtains the current local computing power routing table, which comprises the following steps: The computing power routing module converts the computing power change information into computing power routing information by using the field mapping relationship between the computing power routing table and the computing power information table, and dynamically generates or updates the computing power routing table by using the converted computing power routing information, and obtains the current local computing power routing table.

3. The method according to claim 1 or 2, characterized in that, The defined computing power information type and format comprise a unique identification field, the data type of which is String, a service type field, the data type of which is String, and a resource type field, the data type of which is String.

4. The method according to claim 1 or 2, characterized in that, Before the computing power management module delivers the local computing power change information to the computing power routing module, the method further comprises the following steps: The computing power management module calculates a current local computing power change value by using a formula A' = T | (C-C') |, wherein A' is the current local computing power change value, T is a change rate, C' is a changed total amount of computing power, and C is a current local total amount of computing power, different total amounts of computing power correspond to different change rates; The computing power management module determines whether the current local computing power change value exceeds a defined computing power change threshold value; and The computing power management module delivers computing power change information to the computing power routing module, including: The computing power management module delivers local computing power change information to the computing power routing module when the current local computing power change value exceeds a defined computing power change threshold.

5. The method according to claim 1 or 2, characterized in that, The processed computing power information is delivered to the computing power management module, including: The processed computing power information is delivered to the computing power management module through a gRPC protocol, an SNMP protocol, or a Prometheus toolkit.

6. The method of claim 1 or 2, wherein, The computing power routing module receives updated computing power routing information announced by a remote device, including: The computing power routing module receives updated computing power routing information announced by a remote device through linkage with a BGP routing protocol; or The computing power routing module receives updated computing power routing information announced by a remote device through linkage with an SDN controller.

7. A network device computing power information collection and management system, characterized in that, including: The computing power information collection module, the computing power routing module, the computing power management module, and the computing power information table; wherein: The computing power information collection module is configured to acquire and process computing power information from a local computing power cluster according to a local computing power collection task issued by the computing power management module in a computing power information type and format defined by the computing power management module, in cooperation with a computing power collection agent, and deliver the processed computing power information to the computing power management module, wherein the computing power information collection module is deployed on a network device, and the computing power collection agent is deployed on a computing power node of the computing power cluster; The computing power routing module is configured to receive updated computing power routing information announced by a remote device, convert the updated computing power routing information into computing power information, deliver the converted computing power information to the computing power management module, and dynamically generate or update a computing power routing table based on local computing power change information to obtain a current local computing power routing table, and announce local computing power routing to the remote device based on the current local computing power routing table; The computing power management module is configured to receive processed computing power information delivered by the computing power information collection module and converted computing power information delivered by the computing power routing module, store the processed computing power information and the converted computing power information in the computing power information table, monitor computing power changes based on local computing power information in the computing power information table, and deliver local computing power change information to the computing power routing module; The computing power information table is configured to store the processed computing power information and the converted computing power information; The computing power routing module is specifically configured to convert the computing power routing information into the computing power information by using a field mapping relationship between the computing power routing table and the computing power information table, wherein the field mapping relationship is that a Destination field in the computing power routing table is mapped to an S_IP field in the computing power information table, an APP field in the computing power routing table is mapped to an S_NAME field in the computing power information table, a TYPE field in the computing power routing table is mapped to remote in a Site field in the computing power information table, and other fields in the computing power routing table are mapped to other fields in the computing power information table in a one-to-one manner, and the other fields include CPU and / or GPU.

8. The network device computing power information collection and management system according to claim 7, characterized in that, The computing power routing module is specifically configured to convert the computing power change information into computing power routing information by using a field mapping relationship between the computing power routing table and the computing power information table, and dynamically generate or update the computing power routing table by using the converted computing power routing information to obtain the current local computing power routing table.

9. An electronic device, comprising: The device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus. The memory is used for storing a computer program. The processor is used for executing the program stored on the memory to realize the method of any one of claims 1-6.

10. A computer storage medium, characterized in that, The computer storage medium stores a computer program, and the computer program is executed by the processor to realize the method of any one of claims 1-6.

Images