A computing power network of an intelligent computing center cloud platform based on a predefined rule routing method and device

Abstract

The application provides a computing power network routing method and device based on predefined rules for an intelligent computing center cloud platform, and relates to the technical fields of intelligent computing centers, intelligent computing centers and computing power infrastructure, which comprises the following steps: S1: when a gateway receives an access target computing power application computing power running data packet, the gateway acquires a destination address carried in the computing power running data packet, and the destination address is a network address of the target computing power application; S2: the gateway queries a computing power network routing table based on the network address of the target computing power application, and obtains a plurality of candidate next hop routing addresses corresponding to the network address of the target computing power application and a routing priority of each candidate next hop routing address; S3: the gateway selects an optimal next hop routing address from the plurality of candidate next hop routing addresses corresponding to the network address of the target computing power application based on predefined rules; and S4: the gateway sends the computing power running data packet to the optimal next hop routing address.

Description

Technical Field

[0001] This invention relates to the technical fields of intelligent computing centers, smart computing centers, smart computing clouds and computing power infrastructure, and specifically to a routing method and apparatus for a computing power network of an intelligent computing center cloud platform based on predefined rules. Background Technology

[0002] With the rapid development of artificial intelligence technology, "intelligent computing centers" and "smart computing centers" have emerged.

[0003] An "intelligent computing center" refers to a facility that provides the necessary computing power, data, and algorithms for artificial intelligence applications (such as the development, training, and inference of deep learning models) by utilizing large-scale heterogeneous computing resources, including general-purpose and intelligent computing power. Intelligent computing centers encompass facilities, hardware, and software, and can provide full-stack capabilities from underlying computing power to top-level application enablement.

[0004] "Intelligent computing center" includes, but is not limited to, "intelligent computing center".

[0005] "Intelligent computing center" or artificial intelligence computing center is a type of computing infrastructure that provides computing power services, data services, and algorithm services required for artificial intelligence applications, based on artificial intelligence theory and adopting artificial intelligence computing architecture.

[0006] "Computing power" is the core of "intelligent computing center" and "smart computing center". It is the ability of computer equipment or computing / data center to process information. It is the ability of computer hardware and software to work together to perform a certain computing requirement. It is the computing power to achieve the target output by processing information data. It is a new type of productivity that integrates information computing power, network carrying capacity and data storage capacity. It mainly provides services to society through computing power infrastructure.

[0007] In existing technologies, intelligent computing centers can provide a large number of computing power service nodes (such as computing containers or virtual machines). These computing power service nodes form a computing power network and provide computing power services to external users through a gateway. Computing power applications can be deployed on the computing power service nodes, and external users can access the computing power applications deployed on the computing power service nodes through the gateway. When the gateway receives computing power execution data packets sent by external users to computing power applications, it needs to route the user's computing power execution data packets to the computing power service node where the computing power application resides. With the increase in the number of tenants in intelligent computing centers and the increase in tenant demand, the number of computing power service nodes in the computing power network has also increased dramatically, and the structure of the computing power network has become increasingly complex. How to route and forward computing power execution data packets in a complex computing power network to optimize the network path for external services and improve the transmission efficiency of computing power execution data packets is a gap in the computing power field, and no solution has been proposed yet. Summary of the Invention

[0008] This invention provides a routing method and apparatus for a computing power network of an intelligent computing center cloud platform based on predefined rules, which is used to solve the problem of how to route and forward computing power operation data packets in a complex computing power network, so as to improve the transmission efficiency of computing power operation data packets.

[0009] To solve the above-mentioned technical problems, the present invention is implemented as follows:

[0010] In a first aspect, the present invention provides a routing method for a computing power network of an intelligent computing center cloud platform based on predefined rules, wherein the computing power network includes a gateway and multiple computing power service nodes connected to the gateway, and the method includes:

[0011] Step S1: When the gateway receives a computing power operation data packet that accesses the target computing power application, it obtains the destination address carried in the computing power operation data packet, where the destination address is the network address of the target computing power application.

[0012] Step S2: The gateway queries the computing power network routing table based on the network address of the target computing power application to obtain multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application, and the routing priority of each candidate next-hop routing address; wherein, the computing power network routing table records the routing information of multiple computing power applications, wherein the routing information of each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address; the network address of the computing power application is uniformly allocated by the intelligent computing center cloud platform, and each candidate next-hop routing address corresponds to one computing power service node, and the candidate next-hop routing address is the network address of the corresponding computing power service node;

[0013] Step S3: The gateway selects the best next-hop routing address from multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application based on predefined rules. The predefined rules include: prioritizing the candidate next-hop routing address with the highest routing priority.

[0014] Step S4: The gateway sends the computing power operation data packet to the optimal next-hop routing address.

[0015] Optionally, the multiple computing service nodes corresponding to the multiple candidate next-hop routing addresses of the computing power application include a first computing service node and a second computing service node. The first computing service node is the computing service node that deploys the computing power application, and the second computing service node is the computing service node that does not deploy the computing power application and can be routed to the first computing service node. The routing priority of the first computing service node is higher than the routing priority of the second computing service node.

[0016] Optionally, step S3 includes:

[0017] Step S31: The gateway determines the optimal next-hop routing address of the computing power operation data packets for a first proportion as the candidate next-hop routing address of the first computing power service node corresponding to the target computing power application, and determines the optimal next-hop routing address of the computing power operation data packets for a second proportion as the candidate next-hop routing address of the second computing power service node corresponding to the target computing power application, wherein the first proportion is greater than the second proportion; the predefined rule further includes: allocating computing power operation data packets to the first computing power service node and the second computing power service node among the multiple computing power service nodes corresponding to the computing power application according to a preset proportion.

[0018] Optionally, the method further includes:

[0019] Step S5: The gateway monitors the computing power service status of the computing power application, and the computing power service status includes information on the computing power service nodes that deploy the computing power application;

[0020] Step S6: If a change is detected in the computing power service node where the computing power application is deployed, obtain the network address of the newly deployed computing power service node, and update the routing information of the computing power application in the computing power network routing table based on the network address of the newly deployed computing power service node.

[0021] Optionally, step S6 includes:

[0022] Step S61: If the computing power application corresponds to multiple second computing power service nodes, determine the routing priority of each second computing power service node, wherein the routing priority of each second computing power service node is the same, or the routing priority of each second computing power service node is determined according to the running status information of the second computing power service node, wherein the running status information includes load, the higher the load, the lower the routing priority, and the lower the load, the higher the routing priority.

[0023] Optionally, the predefined rule further includes: if multiple computing power operation data packets carry the same source address, select the same candidate next-hop routing address for the multiple computing power operation data packets.

[0024] Optionally, the routing priority is represented by routing distance information. Among the multiple computing service nodes corresponding to each computing power application, the routing distance information of the computing service node that deploys the computing power application is less than the routing distance information of the computing service node that does not deploy the computing power application. The smaller the routing distance information, the higher the routing priority, and the larger the routing distance information, the lower the routing priority.

[0025] Secondly, the present invention provides a routing device for a computing power network of an intelligent computing center cloud platform based on predefined rules. The computing power network includes a gateway and multiple computing power service nodes connected to the gateway. The device includes:

[0026] The first acquisition module is used to, upon receiving a computing power operation data packet for accessing a target computing power application, acquire the destination address carried in the computing power operation data packet, wherein the destination address is the network address of the target computing power application.

[0027] The routing query module is used to query the computing power network routing table based on the network address of the target computing power application to obtain multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application, and the routing priority of each candidate next-hop routing address. The computing power network routing table records routing information for multiple computing power applications. The routing information for each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address. The network address of the computing power application is uniformly allocated by the intelligent computing center cloud platform. Each candidate next-hop routing address corresponds to one computing power service node, and the candidate next-hop routing address is the network address of the corresponding computing power service node.

[0028] The routing module is used to select the best next-hop routing address from multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application based on predefined rules, wherein the predefined rules include: prioritizing the candidate next-hop routing address with the highest routing priority;

[0029] The forwarding module is used to send the computing power operation data packets to the optimal next-hop routing address.

[0030] Optionally, the multiple computing service nodes corresponding to the multiple candidate next-hop routing addresses of the computing power application include a first computing service node and a second computing service node. The first computing service node is the computing service node that deploys the computing power application, and the second computing service node is the computing service node that does not deploy the computing power application and can be routed to the first computing service node. The routing priority of the first computing service node is higher than the routing priority of the second computing service node.

[0031] Thirdly, the present invention provides a gateway device, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, it implements the steps of the routing method based on predefined rules for the computing power network of the intelligent computing center cloud platform as described in the first aspect above.

[0032] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the steps of the routing method for the computing power network of the intelligent computing center cloud platform based on predefined rules as described in the first aspect above.

[0033] Fifthly, the present invention provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the routing method based on predefined rules for the computing power network of the intelligent computing center cloud platform as described in the first aspect above.

[0034] This invention has the following beneficial effects: A computing power network routing table is configured on the gateway in the computing power network. This routing table records routing information for multiple computing power applications. The routing information for each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address. Each candidate next-hop routing address corresponds to one computing power service node. The routing priority of the computing power service node deploying the computing power application is higher than the routing priority of the computing power service node not deploying the computing power application. When the gateway receives a computing power operation data packet sent by an external user to the computing power application, it can query... The system queries the computing power network routing table to obtain multiple candidate next-hop routing addresses corresponding to the network address of the computing power application, as well as the routing priority of each candidate next-hop routing address. Based on predefined rules, it selects the best next-hop routing address from the multiple candidate next-hop routing addresses, thereby forwarding the computing power operation data packets sent by external users to the selected best next-hop routing address. This optimizes the network path for external services, ensures data transmission efficiency, and helps the intelligent computing center cloud platform support the deployment and operation of large-scale services. In turn, it helps the intelligent computing center cloud platform improve service performance and availability, and supports the large-scale expansion of the intelligent computing center cloud platform. Attached Figure Description

[0035] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0036] Figure 1 This is a schematic diagram of the computing power network structure of the intelligent computing center cloud platform of the present invention;

[0037] Figure 2 This is a flowchart illustrating the routing method of the computing power network of the intelligent computing center cloud platform based on predefined rules according to the present invention.

[0038] Figure 3 This is a schematic diagram of a computing network routing table according to an embodiment of the present invention;

[0039] Figure 4 This is a flowchart illustrating the routing method of the computing power network of the intelligent computing center cloud platform based on predefined rules according to the present invention.

[0040] Figure 5 This is a schematic diagram illustrating the update process of the computing power network routing table in an embodiment of the present invention.

[0041] Figure 6 This is a schematic diagram of a computing network routing table according to an embodiment of the present invention;

[0042] Figure 7 This is a schematic diagram of the routing device for the computing power network of the intelligent computing center cloud platform of the present invention, which is based on predefined rules.

[0043] Figure 8 This is a schematic diagram of the gateway device of the present invention. Detailed Implementation

[0044] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0045] The technical terms involved in this invention will be briefly explained below.

[0046] The “computing power” mentioned in this invention refers to: the ability of computer equipment or computing / data center to process information; the ability of computer hardware and software to work together to perform a certain computing requirement; the computing power to achieve the target result output by processing information data; and a new type of productivity that integrates information computing power, network carrying capacity, and data storage capacity, mainly providing services to society through computing power infrastructure.

[0047] The "computational power" (CP) described in this invention refers to the ability of a data center server to process data and output results. It is a comprehensive indicator of a data center's computing power, encompassing general computing power, supercomputing power, and intelligent computing power. The commonly used unit of measurement is floating-point operations per second (FLOPS, 1 EFLOPS = 10^18 FLOPS), with higher values ​​indicating stronger overall computing power. It is estimated that 1 EFLOPS is approximately equivalent to the computing power output of 5 Tianhe-2A supercomputers, 500,000 mainstream server CPUs, or 2 million mainstream laptops. The calculation formula is: CP = CP 通用 +CP 智能 +CP 超级 .

[0048] The "Network Power" (NP) mentioned in this invention refers to the performance of data transmission capability of computing facilities, which includes comprehensive capabilities such as network architecture, network bandwidth, transmission latency, intelligent management and scheduling, and involves network transmission within and between data centers. It is a comprehensive indicator for measuring network transmission scheduling capability.

[0049] The "Storage Power" (SP) described in this invention refers to the comprehensive capabilities of a data center in four aspects: data storage capacity, performance, security and reliability, and green and low-carbon operation. It is a comprehensive indicator for measuring the data storage capacity of a data center, including external storage devices such as storage arrays and internal storage devices within servers. The commonly used unit of measurement for storage capacity is exabytes (EB, 1EB = 2^60 bytes), the commonly used unit of measurement for performance is the number of read / write operations per second (IOPS / TB), and the disaster recovery ratio is an important indicator of security and reliability.

[0050] The "computing infrastructure" mentioned in this invention refers to a new type of information infrastructure that integrates information computing power, network carrying capacity, and data storage capacity, enabling centralized computing, storage, transmission, and application of information.

[0051] The "new information infrastructure" mentioned in this invention refers to network infrastructure such as 5G networks, fiber optic broadband networks, backbone networks, international communication networks, and satellite internet; computing infrastructure such as data centers, general computing centers, intelligent computing centers, and supercomputing centers; and new technology facilities such as artificial intelligence, blockchain, and quantum computing.

[0052] The “computing power” mentioned in this invention includes: general computing power, intelligent computing power, and supercomputing power.

[0053] The "general computing power" mentioned in this invention refers to the computing power provided by servers based on CPU (Central Processing Unit) chips, which is used to support basic general computing such as cloud computing and edge computing.

[0054] The "intelligent computing power" mentioned in this invention refers to: a computing platform deployed on a large scale based on dedicated chips such as GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), and ASIC (Application Specific Integrated Circuit) for various artificial intelligence innovative applications, such as natural language processing and machine vision.

[0055] The “supercomputing power” mentioned in this invention refers to the computing power provided by high-performance computing clusters such as supercomputers. It utilizes the centralized computing resources of multiple computer systems working in parallel and uses a dedicated operating system to handle extremely complex or data-intensive problems. It is mainly used for computing in cutting-edge scientific fields, such as planetary simulation, drug molecule design, and gene analysis.

[0056] The "intelligent computing center" described in this invention refers to a facility that, through the use of large-scale heterogeneous computing resources, including general-purpose computing power (CPU) and intelligent computing power (GPU, FPGA, ASIC, etc.), primarily provides the necessary computing power, data, and algorithms for artificial intelligence applications (such as the development, training, and inference of deep learning models). The intelligent computing center encompasses facilities, hardware, and software, and can provide full-stack capabilities from underlying computing power to top-level application enablement.

[0057] The "intelligent computing center cloud platform" or "intelligent computing cloud" mentioned in this invention refers to a cloud computing platform that integrates hardware and software resources of an intelligent computing center.

[0058] The "intelligent computing center" mentioned in this invention includes, but is not limited to, "smart computing center".

[0059] The "intelligent computing center" mentioned in this invention, also known as an artificial intelligence computing center, is a type of computing infrastructure that provides computing power services, data services, and algorithm services required for artificial intelligence applications, based on artificial intelligence theory and adopting an artificial intelligence computing architecture.

[0060] The "computing center" mentioned in this invention refers to a facility that is mainly composed of infrastructure such as wind, thermal, hydro, and electricity, and IT hardware and software equipment, and has computing power, carrying capacity, and storage capacity, including general data centers, intelligent computing centers, supercomputing centers, etc.

[0061] The "supercomputing center" mentioned in this invention refers to a supercomputing data center, which is a data center based on supercomputers or large-scale computing clusters. It can provide large-scale computing, storage and network services and is widely used in aerospace, defense, oil exploration, climate modeling and genome sequencing and other application scenarios.

[0062] The “computing resources” mentioned in this invention refer to the technologies and facilities required for the development of the digital society that have the ability to compute, transmit, store and apply information, including but not limited to computing resources such as CPUs and GPUs, network resources such as switches and routers, storage resources such as storage arrays and distributed storage, security resources such as firewalls and intrusion detection systems, and supporting and guaranteeing resources such as wind, fire, water and electricity.

[0063] The "computing power service node" mentioned in this invention refers to a container, virtual machine, or server that can provide computing power services.

[0064] The "computing power network" described in this invention refers to a network consisting of multiple computing power service nodes and gateways for connecting the computing power service nodes.

[0065] The “computing power application” mentioned in this invention refers to an application (app) or algorithm or model that runs on a computing power service node and can provide computing power services to users.

[0066] The "gateway" mentioned in this invention, also known as a "computing power gateway," refers to the core forwarding device in a computing power network when transmitting computing power operation data. The gateway determines the forwarding path of the computing power operation data.

[0067] First and foremost, it is crucial to emphasize that the difficulty of computing power-based data routing (hereinafter referred to as "computing power routing") in this invention far exceeds that of traditional network data routing (hereinafter referred to as "network routing"). Computing power routing represents a fundamental paradigm shift and a revolutionary change. The following is a detailed technical analysis of why the difficulty of computing power routing in this invention is far greater than that of traditional network routing:

[0068] 1. Dimensions of routing decisions: from "single objective" to "multi-objective trade-offs".

[0069] Traditional network routing primarily optimizes network metrics such as latency, bandwidth, hop count, and packet loss rate, which are relatively easy to quantify and measure.

[0070] The computing power routing of this invention requires simultaneous optimization of the following two types of heterogeneous resources:

[0071] (1) Network resources, including latency, bandwidth, etc.

[0072] (2) Computing resources, including CPU / GPU utilization, memory size, storage I / O, specific hardware accelerators, etc.

[0073] This is a multi-objective optimization problem, and often these objectives conflict with each other (for example, the node with the strongest computing power may have high network latency), making the trade-offs and decisions extremely complex.

[0074] 2. The dynamic nature of routing states: from "relatively stable" to "constantly changing".

[0075] Traditional network routing: Although network topology and link status may change, the frequency of change is relatively low (measured in seconds or minutes), and routing protocols have convergence time.

[0076] The computing power routing of this invention addresses the highly dynamic state of computing nodes. A GPU node's computing power can change from idle to fully loaded within milliseconds. The start and end of tasks instantly alter a node's load. This necessitates that the computing power routing system possess near real-time perception and decision-making capabilities, with update and convergence speeds far exceeding those required for network routing.

[0077] 3. Global nature of routing information: from "local information" to "global state".

[0078] Traditional network routing typically employs distributed algorithms (such as link-state routing), where each router only knows the overall network topology but does not need to know the details of every data flow in the network.

[0079] The computing power routing of this invention involves a scheduler (controller) that theoretically needs to know the real-time computing power status of all nodes in the entire network and the real-time requirements of all tasks. This has created a strong demand for centralized, global view controllers (such as software-defined networking (SDN) controllers), but it also brings the risks of scalability issues and single points of failure.

[0080] 4. Heterogeneity and Abstraction: From "Standard Unit" to "Various Differences".

[0081] Traditional network routing deals with standardized data packets (IP packets), which have a uniform format and predictable behavior.

[0082] The computing power routing of this invention includes the following three aspects:

[0083] (1) Heterogeneous computing power, with computing nodes varying greatly, from x86 CPUs to ARM CPUs, to various types of GPUs, neural network processors (NPUs), and field programmable gate arrays (FPGAs). Their architectures, instruction sets, and performance characteristics are completely different.

[0084] (2) The tasks are heterogeneous, and the computing tasks are also different. Some require high parallelism (suitable for GPU), some require high single-core performance (suitable for CPU), some are memory intensive, and some are I / O intensive.

[0085] (3) Abstraction is difficult. How to use a unified "metric" to quantify different types of computing power on different nodes is a huge challenge. It is not possible to simply use "FLOPS" to measure all computing power as easily as bandwidth (Gbps).

[0086] 5. System Coupling and Fault Handling: From “Decoupling” to “Tight Coupling”.

[0087] Traditional network routing: The network is the transport layer, decoupled from the application layer. Network failures typically only affect communication, and applications can design retry mechanisms.

[0088] The computing power routing of this invention tightly couples the network and computing. Once a computing task is routed to a node, if that node fails mid-computation or the network suddenly interrupts, the entire task will fail, potentially requiring the computation to start from scratch rather than simply retransmitting a few data packets. This significantly increases the system's vulnerability and the complexity of fault handling.

[0089] These fundamental differences make computing power routing a core challenge and key technology in future computing power networks, cloud-edge-device collaboration, and AI computing networks. Essentially, it is a complex scheduling problem involving the joint optimization of multi-dimensional resources in a dynamic, heterogeneous, and distributed environment. Its difficulty is naturally far greater than traditional network routing, which only focuses on data packet paths. The extreme difficulty in implementing computing power routing stems from the fact that it does not simply address the problem of "transporting network data packets (network data routing)," but rather the complexity of deep integration of computing power and network, dynamic balancing of multiple factors (heterogeneous indicators such as network bandwidth, computing load, and task priority), and the lack of cross-domain (cloud, edge, and device) collaboration and standardized computing power measurement. This invention addresses this pressing issue in the field of computing power.

[0090] Please refer to Figure 1 This invention provides a computing power network for an intelligent computing center cloud platform, comprising: a gateway and multiple computing power service nodes connected to the gateway. In this invention, computing power applications can be deployed on the computing power service nodes, and these applications can be migrated from one computing power service node to another as needed, for example, using... Figure 1For example, computing application 1 is migrated from computing service node 1 to computing service node 3. It should be noted that multiple computing applications can be deployed in the computing network of this invention, and one or more computing applications can be deployed on a single computing service node. In this invention, the intelligent computing center cloud platform can configure and allocate a unique network address for each computing application. This network address is different from the network address of the computing service node where the computing application is deployed. Even if the computing application is migrated from one computing service node to another, its network address will not change. The network address of the computing application can also be called an external network address, which can be accessed by users outside the intelligent computing center cloud platform. In this invention, the computing service node itself also has a network address, such as... Figure 1 In this invention, the network address of computing service node 1 is IP1, the network address of computing service node 2 is IP2, and the network address of computing service node 3 is IP3. These network addresses can be either external or internal. In this invention, an external network address, also known as a public network address, refers to a network address uniformly allocated by the intelligent computing center cloud platform, which can be published to the outside world for external users to access. An internal network address, also known as a private network address, refers to a network address used internally by the intelligent computing center cloud platform; it is only used within the intelligent computing center cloud platform and is not published to the outside world, thus inaccessible to external users.

[0091] Please refer to Figure 2 This invention provides a routing method for a computing power network of an intelligent computing center cloud platform based on predefined rules. The computing power network is the aforementioned computing power network, including a gateway and multiple computing power service nodes connected to the gateway. The method includes:

[0092] Step S1: When the gateway receives a computing power operation data packet that accesses the target computing power application, it obtains the destination address carried in the computing power operation data packet, where the destination address is the network address of the target computing power application.

[0093] The computing power operation data packet refers to the computing power operation data packet sent by an external user to the target computing power application. The computing power operation data packet carries a source address and a destination address. The source address is the user's network address, and the destination address is the target computing power application's network address.

[0094] Step S2: The gateway queries the computing power network routing table based on the network address of the target computing power application to obtain multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application, as well as the routing priority of each candidate next-hop routing address;

[0095] The computing power network routing table records routing information for multiple computing power applications. For example, please refer to... Figure 3The computing power network routing table records routing information for two computing power applications, assuming they are computing power application 1 and computing power application 2. The network address of computing power application 1 is EIP1, and the network address of computing power application 2 is EIP2. It should be noted that as the number of tenants on the intelligent computing center cloud platform increases, the number of computing power applications deployed in the computing power network becomes very large; this embodiment only uses two computing power applications as an example. The routing information for each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address. Figure 3 Taking computing power application 1 as an example, the routing information includes: the network address of computing power application 1, EIP1, three candidate next-hop routing addresses, IP1, IP2 and IP3, and the routing priority of each candidate next-hop address.

[0096] The network address of the computing power application is uniformly allocated by the intelligent computing center cloud platform and is unique. This network address can also be called an external network address or a public network address, which can be published to the outside world for external users to access.

[0097] In this invention, a candidate next-hop routing address refers to a routing address that the gateway can forward after receiving a computing power operation data packet whose destination address is the network address of the computing power application. Each computing power application's routing information may contain multiple candidate next-hop routing addresses for the gateway to select from. Each candidate next-hop routing address corresponds to one computing power service node, and the candidate next-hop routing address is the network address of the corresponding computing power service node. The multiple computing power service nodes corresponding to the multiple candidate next-hop routing addresses of the computing power application include a first computing power service node and a second computing power service node. The first computing power service node is the computing power service node that deploys the computing power application, and the second computing power service node is the computing power service node that does not deploy the computing power application and can be routed to the first computing power service node. The routing priority of the first computing power service node is higher than that of the second computing power service node.

[0098] Combination Figure 1 To illustrate, assume that the computing service node corresponding to candidate next-hop routing address IP1 is computing service node 1, the computing service node corresponding to candidate next-hop routing address IP2 is computing service node 2, and the computing service node corresponding to candidate next-hop routing address IP3 is computing service node 3. If computing application 1 is deployed on computing service node 1, then computing service node 1 is the first computing service node mentioned above. Both computing service nodes 2 and 3 can be routed to computing service node 1, which is the second computing service node mentioned above.

[0099] In other words, the computing power operation data packets sent by external users to computing power application 1 can be directly routed to computing power service node 1 through the gateway to access computing power application 1 deployed on computing power service node 1, or they can be routed to computing power service node 2 or computing power service node 3 through the gateway first, and then the computing power service node 2 or computing power service node 3 can route them to computing power service node 1 to access computing power application 1 deployed on computing power service node 1.

[0100] In this invention, routing priority is used to indicate which next-hop routing address to prioritize forwarding to. For example... Figure 3 As shown, routing priority can be represented by a preset value, for example... Figure 3 In this context, a smaller numerical value indicates a higher priority. Priority can also be expressed in words, such as first priority, second priority, etc.

[0101] It is understandable that directly routing the computing power running data packets to the computing power service node that deploys the computing power application can enable access to the computing power application as quickly as possible. Therefore, in this invention, the computing power service node that deploys the computing power application (the first computing power service node) is given a higher priority than the computing power service node that does not deploy the computing power application (the second computing power service node).

[0102] Step S3: The gateway selects the best next-hop routing address from multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application based on predefined rules. The predefined rules include: prioritizing the candidate next-hop routing address with the highest routing priority.

[0103] Step S4: The gateway sends the computing power operation data packet to the optimal next-hop routing address.

[0104] In this invention, the gateway in the computing power network is configured with a computing power network routing table, which records routing information for multiple computing power applications. The routing information for each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address. Each candidate next-hop routing address corresponds to one computing power service node. The routing priority of the computing power service node deploying the computing power application is higher than the routing priority of the computing power service node not deploying the computing power application. When the gateway receives a computing power execution data packet sent by an external user to a computing power application, it can query the computing power... The network routing table obtains multiple candidate next-hop routing addresses corresponding to the network address of the computing power application, as well as the routing priority of each candidate next-hop routing address. Based on predefined rules, it selects the best next-hop routing address from multiple candidate next-hop routing addresses, thereby forwarding the computing power operation data packets sent by external users to the selected best next-hop routing address. This optimizes the network path for external services, ensures data transmission efficiency, and helps the intelligent computing center cloud platform support the deployment and operation of large-scale services. In turn, it helps the intelligent computing center cloud platform improve service performance and availability, and supports the large-scale expansion of the intelligent computing center cloud platform.

[0105] In some embodiments, optionally, step S3 includes:

[0106] Step S31: The gateway determines the optimal next-hop routing address of the computing power operation data packets for a first proportion as the candidate next-hop routing address of the first computing power service node corresponding to the target computing power application, and determines the optimal next-hop routing address of the computing power operation data packets for a second proportion as the candidate next-hop routing address of the second computing power service node corresponding to the target computing power application, wherein the first proportion is greater than the second proportion; the predefined rule further includes: allocating computing power operation data packets to the first computing power service node and the second computing power service node among the multiple computing power service nodes corresponding to the computing power application according to a preset proportion.

[0107] Still with Figure 1 and Figure 3 For example, when computing power application 1 is deployed on computing power service node 1, 80% (first proportion) of the computing power operation data packets sent to computing power application 1 can be forwarded to computing power service node 1, 10% (second proportion) of the computing power operation data packets sent to computing power application 1 can be forwarded to computing power service node 2, which will then route them to computing power service node 1, and 10% (second proportion) of the computing power operation data packets sent to computing power application 1 can be forwarded to computing power service node 3, which will then route them to computing power service node 1.

[0108] The aforementioned first and second ratios can be set as needed. For example, they can be flexibly adjusted based on the monitoring of the operating status information (such as load) of the computing power service nodes corresponding to the target computing power application. For instance, assuming the initial first ratio for computing power service node 1 is set to 80%, the second ratio for computing power service node 2 to 10%, and the second ratio for computing power service node 3 to 10%, when the load of computing power service node 1 exceeds a preset threshold, the first ratio for computing power service node 1 can be adjusted to 70%, the second ratio for computing power service node 2 to 15%, and the second ratio for computing power service node 3 to 15%.

[0109] In this invention, the second ratio corresponding to different second computing power service nodes can be the same or different. For example, the second ratio corresponding to computing power service node 2 is 15%, and the second ratio corresponding to computing power service node 3 is 5%. The specific value of the second ratio can also be flexibly adjusted according to the operating status information (such as load) of the second computing power service node.

[0110] In this invention, optional provisions may be made, please refer to [the relevant documentation]. Figure 4 The method further includes:

[0111] Step S5: The gateway monitors the computing power service status of the computing power application, and the computing power service status includes information on the computing power service nodes that deploy the computing power application;

[0112] Step S6: If a change is detected in the computing power service node where the computing power application is deployed, obtain the network address of the newly deployed computing power service node, and update the routing information of the computing power application in the computing power network routing table based on the network address of the newly deployed computing power service node.

[0113] Please refer to Figure 5 Assuming that computing application 1 (app) is initially deployed on computing service node 1, the computing network routing table configured on the gateway can be found in [reference needed]. Figure 5 In the routing table state 1, the routing priority of computing power service node 1 is higher than that of computing power service nodes 2 and 3. When it is detected that computing power application 1 has migrated from computing power service node 1 to computing power service node 3, the gateway updates the routing information of computing power application 1 in the computing power network routing table, configuring the routing priority of computing power service node 3 to be higher than that of computing power service node 1 and computing power service node 2. (See also...) Figure 5 The routing table in state 2.

[0114] It should be noted that when a new computing service node is deployed in the computing power network, and this new node can be routed to the same computing service node as a specific computing application, the computing power network routing table can be updated. The network address of the newly deployed computing service node can be added as the next-hop routing address in the routing information of that computing application, thus dynamically adapting to changes in the computing power network. Conversely, when a computing service node is removed from the computing power network—for example, when it stops operating—the next-hop routing addresses corresponding to that node can be queried from the computing power network routing table, and the node's next-hop routing address can be deleted from the corresponding routing information.

[0115] In some embodiments, optionally, step S6 includes:

[0116] Step S61: If the computing power application corresponds to multiple second computing power service nodes, determine the routing priority of each second computing power service node, wherein the routing priority of each second computing power service node is the same, or the routing priority of each second computing power service node is determined according to the running status information of the second computing power service node, wherein the running status information includes load, the higher the load, the lower the routing priority, and the lower the load, the higher the routing priority.

[0117] The gateway can periodically monitor the operating status information of the second computing power service node and dynamically update the routing priority of the second computing power service node in the computing power network routing table based on the operating status information of the second computing power service node.

[0118] In some embodiments, optionally, the predefined rule further includes: if the source address carried in multiple computing power operation data packets is the same, the same candidate next-hop routing address is selected for multiple computing power operation data packets, thereby ensuring that computing power operation data packets sent by the same user to the target computing power application are forwarded to the target computing power application through the same candidate next-hop routing address, which is beneficial to the data processing of the target computing power application after receiving them.

[0119] In some embodiments, alternatively, please refer to Figure 6 The routing priority is represented by routing distance information. Among the multiple computing service nodes corresponding to each computing application, the routing distance information of the computing service node deploying the computing application is less than the routing distance information of the computing service node not deploying the computing application. A smaller routing distance indicates a higher routing priority, and a larger routing distance indicates a lower routing priority. Of course, this invention does not exclude the use of other types of information to represent routing priority.

[0120] Please refer to Figure 7The present invention also provides a routing device 10 for a computing power network of an intelligent computing center cloud platform based on predefined rules. The computing power network includes a gateway and multiple computing power service nodes connected to the gateway. The routing device 10 includes:

[0121] The first acquisition module 11 is used to acquire, upon receiving a computing power operation data packet for accessing a target computing power application, the destination address carried in the computing power operation data packet, wherein the destination address is the network address of the target computing power application.

[0122] The routing query module 12 is used to query the computing power network routing table based on the network address of the target computing power application to obtain multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application, and the routing priority of each candidate next-hop routing address. The computing power network routing table records routing information for multiple computing power applications. The routing information for each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address. The network address of the computing power application is uniformly allocated by the intelligent computing center cloud platform. Each candidate next-hop routing address corresponds to one computing power service node. The candidate next-hop routing address is the network address of the corresponding computing power service node. The multiple computing power service nodes corresponding to the multiple candidate next-hop routing addresses of the computing power application include a first computing power service node and a second computing power service node. The first computing power service node is the computing power service node that deploys the computing power application, and the second computing power service node is the computing power service node that does not deploy the computing power application and can be routed to the first computing power service node. The routing priority of the first computing power service node is higher than the routing priority of the second computing power service node.

[0123] The routing module 13 is used to select the best next-hop routing address from multiple candidate next-hop routing addresses corresponding to the network address of the target computing power application based on predefined rules, wherein the predefined rules include: prioritizing the candidate next-hop routing address with the highest routing priority;

[0124] Forwarding module 14 is used to send the computing power operation data packet to the optimal next-hop routing address.

[0125] In this invention, the gateway in the computing power network is configured with a computing power network routing table, which records routing information for multiple computing power applications. The routing information for each computing power application includes: the network address of the computing power application, multiple candidate next-hop routing addresses, and the routing priority corresponding to each candidate next-hop routing address. Each candidate next-hop routing address corresponds to one computing power service node. The routing priority of the computing power service node deploying the computing power application is higher than the routing priority of the computing power service node not deploying the computing power application. When the gateway receives a computing power execution data packet sent by an external user to a computing power application, it can query the computing power... The network routing table obtains multiple candidate next-hop routing addresses corresponding to the network address of the computing power application, as well as the routing priority of each candidate next-hop routing address. Based on predefined rules, it selects the best next-hop routing address from multiple candidate next-hop routing addresses, thereby forwarding the computing power operation data packets sent by external users to the selected best next-hop routing address. This optimizes the network path for external services, ensures data transmission efficiency, and helps the intelligent computing center cloud platform support the deployment and operation of large-scale services. In turn, it helps the intelligent computing center cloud platform improve service performance and availability, and supports the large-scale expansion of the intelligent computing center cloud platform.

[0126] In some embodiments, optionally, the routing module 13 includes:

[0127] The determination submodule is used to determine the optimal next-hop routing address of a first proportion of the computing power operation data packets as the candidate next-hop routing address corresponding to the first computing power service node corresponding to the target computing power application, and to determine the optimal next-hop routing address of a second proportion of the computing power operation data packets as the candidate next-hop routing address corresponding to the second computing power service node corresponding to the target computing power application, wherein the first proportion is greater than the second proportion; the predefined rule further includes: allocating computing power operation data packets to the first computing power service node and the second computing power service node among the multiple computing power service nodes corresponding to the computing power application according to a preset proportion.

[0128] In some embodiments, optionally, the routing device 10 includes:

[0129] The monitoring module is used to monitor the computing power service status of the computing power application, the computing power service status including information on the computing power service nodes where the computing power application is deployed;

[0130] The second acquisition module is used to acquire the network address of the newly deployed computing service node of the computing application if a change is detected in the computing service node where the computing application is deployed, and update the routing information of the computing application in the computing network routing table based on the network address of the newly deployed computing service node.

[0131] In some embodiments, optionally, the second acquisition module is used to determine the routing priority of each second computing power service node if the computing power application corresponds to multiple second computing power service nodes, wherein the routing priority of each second computing power service node is the same, or the routing priority of each second computing power service node is determined according to the running status information of the second computing power service node, wherein the running status information includes load, the higher the load, the lower the routing priority, and the lower the load, the higher the routing priority.

[0132] In some embodiments, optionally, the predefined rule further includes: if multiple computing power operation data packets carry the same source address, selecting the same candidate next-hop routing address for the multiple computing power operation data packets.

[0133] In some embodiments, optionally, the routing priority is represented by routing distance information, wherein, among the multiple computing service nodes corresponding to each computing power application, the routing distance information of the computing service node that deploys the computing power application is less than the routing distance information of the computing service node that does not deploy the computing power application, wherein a smaller routing distance information indicates a higher routing priority, and a larger routing distance information indicates a lower routing priority.

[0134] Please refer to Figure 8 The present invention also provides a gateway device 20, including a processor 21, a memory 22, and a computer program stored in the memory 22 and executable on the processor 21. When the computer program is executed by the processor 21, it implements the various processes of the above-described embodiment of the routing method for the computing power network of the intelligent computing center cloud platform based on predefined rules, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0135] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the aforementioned embodiment of the routing method for the computing power network of the intelligent computing center cloud platform based on predefined rules, and achieves the same technical effect. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0136] The present invention also provides a computer program product, including computer instructions that, when executed by a processor, implement the above-described... Figure 2 The various processes of the computing power network of the intelligent computing center cloud platform shown are based on the routing selection method of predefined rules and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0137] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0138] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0139] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of the present invention.

Claims

1. A method for routing of a computing power network of an intelligent computing center cloud platform based on predefined rules, characterized in that, The computing power network includes a gateway and a plurality of computing power service nodes connected with the gateway, and the method comprises: Step S1: When the gateway receives a computing power running data packet accessing a target computing power application, the gateway acquires a destination address carried in the computing power running data packet, the destination address being a network address of the target computing power application; Step S2: The gateway queries a computing power network routing table based on the network address of the target computing power application to obtain a plurality of candidate next-hop routing addresses corresponding to the network address of the target computing power application and routing priorities of each candidate next-hop routing address; wherein the computing power network routing table records routing information of a plurality of computing power applications, wherein the routing information of each computing power application comprises a network address of the computing power application, a plurality of candidate next-hop routing addresses, and a routing priority corresponding to each candidate next-hop routing address, the network address of the computing power application being uniformly assigned by the intelligent computing center cloud platform, each candidate next-hop routing address corresponding to a computing power service node, and the candidate next-hop routing address being a network address of the corresponding computing power service node; Step S3: The gateway selects an optimal next-hop routing address from the plurality of candidate next-hop routing addresses corresponding to the network address of the target computing power application based on a predefined rule, wherein the predefined rule comprises preferentially selecting a candidate next-hop routing address with the highest routing priority; Step S4: The gateway sends the computing power running data packet to the optimal next-hop routing address; Wherein the plurality of computing power service nodes corresponding to the plurality of candidate next-hop routing addresses of the computing power application comprise a first computing power service node and a second computing power service node, the first computing power service node being a computing power service node deploying the computing power application, the second computing power service node being a computing power service node not deploying the computing power application and being capable of routing to the first computing power service node, and the routing priority of the first computing power service node being higher than the routing priority of the second computing power service node; Wherein the step S3 comprises: Step S31: The gateway determines that the optimal next-hop routing address of a first proportion of the computing power running data packets is a candidate next-hop routing address corresponding to the first computing power service node corresponding to the target computing power application, and determines that the optimal next-hop routing address of a second proportion of the computing power running data packets is a candidate next-hop routing address corresponding to the second computing power service node corresponding to the target computing power application, wherein the first proportion is greater than the second proportion; the predefined rule further comprises: for the first computing power service node and the second computing power service node in the plurality of computing power service nodes corresponding to the computing power application, the computing power running data packets are distributed according to a preset proportion.

2. The method of claim 1, wherein, Further comprising: Step S5: The gateway monitors a computing power service state of the computing power application, and the computing power service state comprises information of a computing power service node deploying the computing power application; Step S6: If a change in the computing power service node deploying the computing power application is monitored, the network address of the newly deployed computing power service node of the computing power application is obtained, and the routing information of the computing power application in the computing power network routing table is updated based on the network address of the newly deployed computing power service node of the computing power application.

3. The method of claim 2, wherein, The step S6 includes: Step S61: If the computing power application corresponds to a plurality of second computing power service nodes, the gateway determines the routing priority of each second computing power service node, wherein the routing priority of each second computing power service node is the same, or the routing priority of each second computing power service node is determined according to the running state information of the second computing power service node, wherein the running state information includes load, and the greater the load, the lower the routing priority, and the smaller the load, the higher the routing priority.

4. The method according to any one of claims 1 to 3, characterized in that, The predefined rule further includes: if the source addresses carried in a plurality of the computing power running data packets are the same, the same candidate next hop routing address is selected for a plurality of the computing power running data packets.

5. The method according to any one of claims 1 to 3, wherein The routing priority is represented by routing distance information, wherein in a plurality of computing power service nodes corresponding to each computing power application, the routing distance information of the computing power service node deploying the computing power application is smaller than the routing distance information of the computing power service node not deploying the computing power application, wherein the smaller the routing distance information, the higher the routing priority, and the greater the routing distance information, the lower the routing priority.

6. A computing power network routing device based on predefined rules of an intelligent computing center cloud platform, characterized in that, The computing power network includes a gateway and a plurality of computing power service nodes connected to the gateway, and the device includes: A first obtaining module is configured to, in a case where a computing power running data packet accessing a target computing power application is received, obtain a destination address carried in the computing power running data packet, the destination address being a network address of the target computing power application; A routing query module is configured to query a computing power network routing table based on the network address of the target computing power application, to obtain a plurality of candidate next hop routing addresses corresponding to the network address of the target computing power application and routing priorities of each candidate next hop routing address; wherein the computing power network routing table records routing information of a plurality of computing power applications, wherein the routing information of each computing power application includes: a network address of the computing power application, a plurality of candidate next hop routing addresses, and routing priorities of each candidate next hop routing address, the network address of the computing power application being uniformly allocated by the intelligent computing center cloud platform, each candidate next hop routing address corresponding to a computing power service node, and the candidate next hop routing address being a network address of the corresponding computing power service node; A routing selection module is configured to select a best next hop routing address from the plurality of candidate next hop routing addresses corresponding to the network address of the target computing power application based on a predefined rule, wherein the predefined rule includes: preferentially selecting a candidate next hop routing address with the highest routing priority; A forwarding module is configured to send the computing power running data packet to the best next hop routing address. The plurality of computing power service nodes corresponding to the plurality of candidate next hop routing addresses of the computing power application include a first computing power service node and a second computing power service node, the first computing power service node is a computing power service node deploying the computing power application, the second computing power service node is a computing power service node not deploying the computing power application and capable of routing to the first computing power service node, and the routing priority of the first computing power service node is higher than that of the second computing power service node. The routing selection module includes: The determining sub-module is configured to determine the best next hop routing address of the first proportion of the computing power running data packets as the candidate next hop routing address corresponding to the first computing power service node corresponding to the target computing power application, and determine the best next hop routing address of the second proportion of the computing power running data packets as the candidate next hop routing address corresponding to the second computing power service node corresponding to the target computing power application, wherein the first proportion is greater than the second proportion; and the predefined rule further includes: distributing the computing power running data packets to the first computing power service node and the second computing power service node in the plurality of computing power service nodes corresponding to the computing power application according to a preset proportion.

7. A gateway device, characterized by The processor, the memory, and the program stored on the memory and executable on the processor, wherein the program is executed by the processor to implement the steps of the computing power network routing selection method based on the predefined rule of the intelligent computing center cloud platform according to any one of claims 1 to 5. The computer program is stored on the computer readable storage medium, and the computer program is executed by the processor to implement the steps of the computing power network routing selection method based on the predefined rule of the intelligent computing center cloud platform according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer program includes computer instructions, and the computer instructions are executed by the processor to implement the steps of the computing power network routing selection method based on the predefined rule of the intelligent computing center cloud platform according to any one of claims 1 to 5.

9. A computer program product, characterised in that, ​

Images