Computing power routing method, system, equipment, medium and program product

By defining the network view, including the mapping relationship between the ingress gateway, egress gateway, and computing nodes, the problem of joint optimization and scheduling of computing power and network in the existing technology is solved, and a more reasonable routing strategy and improved resource utilization efficiency are achieved.

CN118945101AActive Publication Date: 2024-11-12CHINA MOBILE COMM LTD RES INST +1
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411333608.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-24
Publication Date
2024-11-12
Estimated Expiration
2044-09-24

AI Technical Summary

Technical Problem

Existing centralized and hybrid computing power routing schemes fail to effectively consider the mapping relationship between egress gateways and computing power nodes, making it difficult to jointly optimize and schedule computing power and network.

Method used

By collecting information on the ingress gateway and computing nodes connected to the egress gateway, the network view is determined, including the mapping relationship between the ingress gateway, egress gateway, and computing nodes. Based on this, computing routing decisions are made. The controller not only maintains the mapping relationship from the ingress gateway to the egress gateway, but also collects complete computing node information for joint optimization scheduling.

Benefits of technology

It achieves joint optimization scheduling of computing power and network, generates more reasonable routing strategies, improves service availability, load balancing and response speed, enhances network fault tolerance and resource utilization efficiency, simplifies management and improves network performance and resource utilization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118945101A_ABST
    Figure CN118945101A_ABST
Patent Text Reader

Abstract

The invention provides a computing power routing method, a computing power routing system, computing power routing equipment, a medium and a program product. The method comprises the following steps: collecting entry gateway information and computing power node information of a computing power node connected with an exit gateway; according to the collected entry gateway information and computing power node information of the computing power node connected with the exit gateway, determining a computing network view, the computing network view comprising a mapping relationship among the entry gateway, the exit gateway and the computing power node information; wherein the computing network view is used for making a computing power routing decision. The network controller not only collects the complete mapping relation from the entrance gateway to the exit gateway to the computing power state information, but also determines the computing network view through the mapping relation, further introduces the computing power information into the network, comprehensively considers the network information and the computing power information during routing decision making, performs joint optimization, and improves the routing decision making efficiency. And a more reasonable routing strategy is generated, and the service request is routed to the optimal service node along the optimal path.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the field of data communication network technology, and in particular to a computing power routing method, system, device, medium and program product. Background Art

[0002] In today's network, different routing decision schemes have emerged according to the different ways of sensing computing power information, the sensing location, and the location of computing power routing decisions. In both centralized and hybrid schemes, computing power information is obtained through cloud management and sent to the network controller; that is, the cloud management system only maintains the relevant information of the computing power nodes; the network controller only maintains the mapping relationship from the entry gateway to the exit gateway; the cloud and network management systems are independent, and the information between the two cannot be interoperable, which makes it difficult to jointly optimize the scheduling of computing power and the network. Joint optimization scheduling requires comprehensive consideration of the network information from the entry gateway to the exit gateway, the network information from the exit gateway to the computing power node, and the computing power node information. However, the existing centralized and hybrid schemes do not consider the mapping relationship between the computing power node and the exit gateway. Neither the controller nor the cloud management system perceives the mapping relationship. When the controller or the entry gateway makes a decision, it is difficult for computing power and the network to be jointly optimized. Summary of the invention

[0003] The embodiments of the present application provide a computing power routing method, system, device, medium and program product to solve the problem that the existing computing power and network are difficult to jointly optimize and schedule.

[0004] In order to solve the above technical problems, this application is implemented as follows:

[0005] In a first aspect, an embodiment of the present application provides a computing power routing method, including:

[0006] Collect the inlet gateway information and computing node information of the computing power node connected to the egress gateway;

[0007] Determine a computing network view according to the collected inlet gateway information and computing node information of the computing node connected to the egress gateway, wherein the computing network view includes: a mapping relationship between the inlet gateway, the egress gateway, and the computing node information;

[0008] The computing network view is used to make computing power routing decisions.

[0009] Optionally, the collecting of inlet gateway information and computing node information of computing nodes connected to the egress gateway includes:

[0010] Receive the computing node inlet gateway information connected to the egress gateway, the computing node inlet gateway information at least including: the egress gateway, the computing node inlet gateway IP;

[0011] Receive computing power node information, the computing power node information includes: the computing power node entry gateway IP, computing power node status information.

[0012] Optionally, the computing power node entry gateway information further includes an anycast address IP, wherein the anycast address IP is used to identify the same type of services;

[0013] The computing network view also includes: a mapping relationship between the computing power node information and different services.

[0014] Optionally, the receiving computing power node ingress gateway information connected to the egress gateway includes:

[0015] Based on the extended distributed routing protocol, receiving the computing power node entry gateway information;

[0016] Among them, the extended distributed routing protocol is an extended border gateway protocol link state protocol, and the extended border gateway protocol link state protocol includes the path attributes of the extended border gateway protocol link state protocol. The path attributes are used to notify the topology information of the connected computing power nodes and at least carry the computing power node entry gateway IP.

[0017] Optionally, also include:

[0018] Based on the computing network view, determine the weight information of the computing power nodes, and the weight information is used to make computing power routing decisions;

[0019] The weight information of the computing power node is sent to the entry gateway or the scheduling domain, wherein the scheduling domain includes at least one entry gateway.

[0020] Optionally, the determining the weight information of the computing power nodes based on the computing network view further includes:

[0021] For each of the entry gateways, determine whether the number of computing nodes for each type of service is greater than or equal to 2;

[0022] If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the entry gateway;

[0023] If the number of computing nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing nodes;

[0024] or,

[0025] generating the scheduling domain;

[0026] For each of the scheduling domains, determine whether the number of computing nodes for each type of service is greater than or equal to 2;

[0027] If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the scheduling domain;

[0028] If the number of computing power nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing power nodes.

[0029] Optionally, after sending the weight information of the computing power node to the ingress gateway or the scheduling domain, the method further includes:

[0030] If the computing power node information or service deployment changes, determine whether it is necessary to regenerate the weight information of the computing power node and issue it, wherein the service deployment is associated with the computing power node information and the computing power node inlet gateway information connected to the egress gateway;

[0031] If the weight needs to be regenerated and issued, the computing network view is regenerated based on the changed computing node information or the service deployment, and the computing node inlet gateway information connected to the egress gateway;

[0032] Based on the regenerated computing network view, the weight information of the computing power node is regenerated and sent to the entry gateway or scheduling domain.

[0033] In a second aspect, an embodiment of the present application provides a method for computing power routing, including:

[0034] Receive weight information of a computing power node, where the weight information is determined based on a computing network view, where the computing network view includes: a mapping relationship between an ingress gateway, an egress gateway, and computing power node information;

[0035] Based on the weight information, a computing power routing decision is made.

[0036] In a third aspect, an embodiment of the present application provides a computing power routing system, including:

[0037] The data module is used to collect the inlet gateway information and computing node information of the computing node connected to the egress gateway;

[0038] A computing network module, used to determine a computing network view according to the collected inlet gateway information and computing node information of the computing nodes connected to the egress gateway, wherein the computing network view includes: a mapping relationship between the inlet gateway, the egress gateway and the computing node information;

[0039] The computing network view is used to make computing power routing decisions.

[0040] In a fourth aspect, an embodiment of the present application provides an electronic 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, the steps of the computing power routing method described in the first aspect above are implemented.

[0041] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the computing power routing method described in the first aspect above are implemented.

[0042] In a sixth aspect, a computer program product is provided, comprising computer instructions, which, when executed by a processor, implement the steps of the computing power routing method as described in the first aspect above.

[0043] In the present application, by collecting the inlet gateway information and the computing power node information of the computing power node connected to the egress gateway, and based on the collected inlet gateway information and the computing power node information of the computing power node connected to the egress gateway, the computing network view is determined, and the mapping relationship between the inlet gateway, the egress gateway and the computing power node information can be obtained. In the embodiment of the present application, the controller not only maintains the mapping relationship from the inlet gateway to the egress gateway, but also collects the inlet gateway information and the computing power node information of the computing power node connected to the egress gateway, and further determines the computing network view. The mapping relationship between the inlet gateway, the egress gateway and the computing power node information in the computing network view provides a basis for the joint optimization scheduling of computing power and the network. Unlike the prior art, the computing power and the network do not consider the mapping relationship between the computing power node and the egress gateway. In the present application, the network controller not only collects the complete mapping relationship from the inlet gateway to the egress gateway and then to the computing power status information, but also determines the computing network view through the mapping relationship, further introduces the computing power information into the network, comprehensively considers the network information and the computing power information in the routing decision, performs joint optimization, generates a more reasonable routing strategy, and routes the service request along the best path to the best service node. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0045] Figure 1 It is a schematic diagram of a distributed computing power information perception method in the prior art;

[0046] Figure 2 It is a schematic diagram of a centralized service topology generation method in the prior art;

[0047] Figure 3 It is a schematic diagram of another centralized service topology generation method in the prior art;

[0048] Figure 4 is a flow chart of a computing power routing method provided in an embodiment of the present application;

[0049] Figure 5 It is a schematic diagram of generating a computing network view provided by an embodiment of the present application;

[0050] Figure 6 It is a schematic diagram of an extended BGP-LS path attribute provided in an embodiment of the present application;

[0051] Figure 7 It is a schematic diagram of generating computing power node weights provided in an embodiment of the present application;

[0052] Figure 8 is a schematic diagram of another computing power routing method provided in an embodiment of the present application;

[0053] Fig. 9 It is a structural diagram of a computing power routing system provided in an embodiment of the present application;

[0054] Fig.10 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0056] At present, different routing strategies have emerged according to the way computing power information is perceived, the location of perception, and the location of computing power routing decisions.

[0057] Please refer to Figure 1 ,Distributed: the computing power information is announced among the network nodes through the routing protocol, and the routing decision is made at the entry node;

[0058] Please refer to Figure 2 and Figure 3 ,Centralized: The computing power information is sent to the network controller through the cloud management system. The network controller makes centralized decisions and sends routing strategies to the routers;

[0059] Hybrid: The computing power information is sent to the network controller through the cloud management system. The network controller sends the computing power information to the entry node, and the entry node decides the routing strategy.

[0060] However, in the prior art, the mapping relationship between the egress gateway and the computing power node is not considered when making routing decisions.

[0061] In addition, since traditional routing protocols trigger notifications when network status changes, if computing power information is transmitted in a distributed routing protocol, the two have different change cycles, which will cause the routing protocol to oscillate and not converge, affecting the existing routing system. Therefore, in the routing system, it is necessary to consider how to minimize the impact on the existing system while also completing distributed computing network joint computing. Therefore, in order to avoid the above problems, the computing power information of this application is not notified between network nodes through the routing protocol, but is reported to the controller, specifically.

[0062] Please refer to Figure 4 , the embodiment of the present application provides a computing power routing method, including:

[0063] Step 11: Collect the inlet gateway information and computing node information of the computing node connected to the egress gateway;

[0064] Step 12: Determine a computing network view based on the collected inlet gateway information and computing node information of the computing node connected to the egress gateway, wherein the computing network view includes: a mapping relationship between the inlet gateway, the egress gateway, and the computing node information;

[0065] The computing network view is used to make computing power routing decisions.

[0066] In the present application, by collecting the inlet gateway information and the computing power node information of the computing power node connected to the egress gateway, and based on the collected inlet gateway information and the computing power node information of the computing power node connected to the egress gateway, the computing network view is determined, and the mapping relationship between the inlet gateway, the egress gateway and the computing power node information can be obtained. In the embodiment of the present application, the controller not only maintains the mapping relationship from the inlet gateway to the egress gateway, but also collects the inlet gateway information and the computing power node information of the computing power node connected to the egress gateway, and further determines the computing network view. The mapping relationship between the inlet gateway, the egress gateway and the computing power node information in the computing network view provides a basis for the joint optimization scheduling of computing power and the network. Unlike the prior art, the computing power and the network do not consider the mapping relationship between the computing power node and the egress gateway. In the present application, the network controller not only collects the complete mapping relationship from the inlet gateway to the egress gateway and then to the computing power status information, but also determines the computing network view through the mapping relationship, further introduces the computing power information into the network, comprehensively considers the network information and the computing power information in the routing decision, performs joint optimization, generates a more reasonable routing strategy, and routes the service request along the best path to the best service node.

[0067] Furthermore, it is also necessary to consider how to send the above mapping relationship to the controller so that it can perform comprehensive network and path calculation.

[0068] Optionally, the collecting of inlet gateway information and computing node information of computing nodes connected to the egress gateway includes:

[0069] Receive the computing node inlet gateway information connected to the egress gateway, the computing node inlet gateway information at least including: the egress gateway, the computing node inlet gateway IP;

[0070] Receive computing power node information, the computing power node information includes: the computing power node entry gateway IP, computing power node status information.

[0071] Optionally, the inlet gateway information of the computing node connected to the egress gateway received by the controller comes from the egress gateway. That is, the egress gateway sends the inlet gateway information of the computing node connected to the egress gateway, including (egress gateway, computing node inlet gateway IP).

[0072] Optionally, the computing power node information received by the controller comes from the cloud management system, that is, the cloud management system sends the computing power node information to the controller, including the computing power node entry gateway IP and computing power status information.

[0073] Please refer to Figure 5 Optionally, the controller completes the mapping from the egress gateway to the computing power status information according to the "computing power node entry gateway IP" (egress gateway, computing power node entry gateway IP, anycast IP, computing power status information), and then combines the mapping relationship between the entry gateway and the egress gateway to form a computing network view containing a complete mapping relationship from the entry gateway to the egress gateway and then to the computing power status information. The computing network view includes the entry gateway, the egress gateway, the computing power node entry gateway IP and the computing power status information. It is worth noting that in this application Figure 5 For illustrative purposes only, the mapping relationship between some computing nodes and gateways is omitted.

[0074] Optionally, computing power routing needs to select the best node from multiple nodes providing the same service while considering the network path. Therefore, anycast address is introduced to identify the same service, and computing power routing maps the anycast address to a real IP address.

[0075] Optionally, the computing power node entry gateway information further includes an anycast address IP, wherein the anycast address IP is used to identify the same type of services;

[0076] The computing network view also includes: a mapping relationship between the computing power node information and different services.

[0077] Optionally, the mapping relationship between anycast and real IP can be obtained by establishing a BGP peer (Border Gateway Protocol peer, BGP peer) between the computing power node ingress gateway and the computing power routing egress gateway. By establishing a BGP peer relationship, devices can exchange routing information to achieve routing selection and traffic transmission between networks. When the egress gateway sends the computing power node ingress gateway information connected to the egress gateway, it will also send the anycast IP. Please refer to Figure 6 , the same type of services can also be identified through anycast address IP, and a computing network view that also includes the mapping relationship between computing power node information and different services can be further constructed. By identifying the same type of services through anycast address IP, user requests can be automatically routed to the nearest or optimal node providing the service, which enhances the availability, load balancing and response speed of the service, while improving the network's fault tolerance and resource utilization efficiency, helping to improve the overall user experience.

[0078] Optionally, in some embodiments, the mapping relationship is reported by extending the Border Gateway Protocol Link-State (BGP-LS) protocol. The BGP-LS protocol is mainly used to distribute network topology information through the Border Gateway Protocol (BGP), and can pass the egress gateway information to the controller to more comprehensively understand the connectivity and status of the entire network.

[0079] Optionally, the receiving computing power node ingress gateway information connected to the egress gateway includes:

[0080] Based on the extended distributed routing protocol, receiving the computing power node entry gateway information;

[0081] Among them, the extended distributed routing protocol is an extended border gateway protocol link state protocol, and the extended border gateway protocol link state protocol includes the path attributes of the extended border gateway protocol link state protocol. The path attributes are used to notify the topology information of the connected computing power nodes and at least carry the computing power node entry gateway IP.

[0082] Optionally, in some embodiments, please refer to Figure 6, the extended BGP-LS in the embodiment of the present application: defines a type of BGP network layer reachability information (NLRI) type, including node information, link information, and IP prefix; in addition, a new type is defined in the BGP-LS path attribute (BGP-LS path attribute), which is an optional non-transitive attribute. It encodes the attributes of the object (link, node, and prefix). Among them, the extended BGP-LS path attribute specifies the topology information used to announce the connected computing power nodes in the BGP-LS attributes, and defines a new type-length-value (Type-Length-Value, TLV) that carries the IP of the connected computing power node entrance gateway and the configured anycast IP address.

[0083] In an embodiment of the present application, by receiving computing power node entry gateway information based on an extended distributed routing protocol, it is possible to dynamically and efficiently update and disseminate computing resource status, so that the network can intelligently select the best path to allocate user requests, optimize resource utilization, and improve service quality.

[0084] Optionally, in an embodiment of the present application, the controller makes a computing power routing decision based on the computing network view, and sends the routing decision directly to each entry gateway. The controller makes a computing power routing decision based on the computing network view, and sends the decision directly to each entry gateway, which can achieve global optimization, rapid response, load balancing and fault recovery, effectively improve network performance and resource utilization, simplify management, improve security, ensure the implementation of unified policies and dynamic adjustment capabilities, and is suitable for complex large-scale network environments.

[0085] The existing centralized and hybrid solutions do not consider the mapping relationship between computing nodes and export gateways, and the controller and cloud management system are not aware of the mapping relationship. Therefore, this application proposes a specific method to achieve topology correspondence through the above embodiments.

[0086] When there is a traffic engineering (TE) requirement, the traditional network controller sends the weights of multiple network paths to the ingress gateway, which selects the specified path to implement the TE function. For best effort (BE) requirements, the ingress gateway makes its own decision on forwarding. Most of the traffic is BE traffic, and TE is used for some high-priority or value-added services.

[0087] Analogous to the above solution, in the centralized computing routing scenario, the controller sends the path and destination node selection strategy to the ingress gateway in the TE scenario; however, in the BE scenario, the ingress gateway is not aware of the computing node status information, and cannot support the BE solution of computing network joint scheduling. For service requests that require joint scheduling, they can only be sent to the controller, which will bring performance burden to the controller;

[0088] Based on the above-mentioned embodiments, the present application also provides some embodiments for issuing weights based on a global view. That is to say, the routing decision is made by the ingress gateway or the scheduling domain. The controller does not issue a complete computing network view, but instead issues the weight value of the ingress gateway of each computing power node corresponding to the same anycast IP. The computing power weight information is combined on the ingress gateway side to implement BE forwarding.

[0089] Optionally, in some embodiments, the method further includes:

[0090] Based on the computing network view, determine the weight information of the computing power nodes, and the weight information is used to make computing power routing decisions;

[0091] Optionally, the weight information includes the egress gateway, the computing node ingress gateway IP, and the weight value corresponding to the computing node.

[0092] The weight information of the computing power node is sent to the entry gateway or the scheduling domain, wherein the scheduling domain includes at least one entry gateway.

[0093] Optionally, in an embodiment of the present application, the generation and distribution of weight information are divided into two granularities, namely, weight generation and distribution based on the entry gateway as the granularity and weight generation and distribution based on the scheduling domain as the granularity. In computing power routing, the number of computing power nodes covered by an entry gateway may be small, and effective load balancing cannot be formed. Therefore, it is possible to consider multiple entry gateways forming a domain to share the load among multiple entry gateways. At this time, it is necessary to generate and distribute weights based on the scheduling domain as the granularity. Optionally, the scheduling domain can be divided by geographical location area, and the nearest entry gateways are divided into the same scheduling domain; or it can be divided by service deployment, and several entry gateways deployed with the same service are divided into the same scheduling domain.

[0094] Optionally, in the embodiment of the present application, a complete computing network view is formed by a controller: (entry gateway, exit gateway, computing power node entry gateway IP, anycast IP and computing power status information), and then for each entry gateway or for each scheduling domain, the controller sends the weight information corresponding to the anycast IP (exit gateway, computing power node entry gateway IP, weight). After receiving the user request, the entry gateway or scheduling domain can select the computing power node according to the preset weight information, and then select the exit gateway, and the path between the entry gateway and the exit gateway is still BE.

[0095] In an embodiment of the present application, the network controller calculates the weights of the computing power nodes, generates a computing power node weight table for each ingress gateway / notification domain, and sends it to the ingress gateway. The weight information is generated by the controller and sent to the ingress gateway or scheduling domain, thereby reducing the computing overhead and storage overhead of the router, avoiding the transmission of computing power information in the routing system, reducing the burden on the routing system, and allowing for fast route calculation.

[0096] Optionally, the determining the weight information of the computing power nodes based on the computing network view further includes:

[0097] For each of the entry gateways, determine whether the number of computing nodes for each type of service is greater than or equal to 2;

[0098] If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the entry gateway;

[0099] If the number of computing nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing nodes;

[0100] or,

[0101] generating the scheduling domain;

[0102] For each of the scheduling domains, determine whether the number of computing nodes for each type of service is greater than or equal to 2;

[0103] If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the scheduling domain;

[0104] If the number of computing power nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing power nodes.

[0105] Optional, please refer to Figure 6Taking the entry gateway as the granularity, the entire process includes: the controller forms a complete computing network view; the information table is generated at the entry gateway as the granularity. For each entry gateway, if there are multiple computing nodes with the same anycast IP, the weight information of each computing node is generated according to the status information of the computing node, and the controller sends the weight information of each anycast IP (exit gateway, computing node entry gateway IP, weight); further, if for a certain entry gateway, there is only one computing node for a certain anycast IP, there is no need to generate and send weight information; reduce the amount of information sent.

[0106] Optionally, taking the scheduling domain as the granularity, the entire process includes: the controller forms a complete computing network view; the controller generates a scheduling domain; optionally, the scheduling domain can be divided by geographical location, and the nearest entry gateways are divided into the same scheduling domain; or it can be divided by service deployment, and several entry gateways deployed with the same service are divided into the same scheduling domain; an information table is generated with the scheduling domain as the granularity, and for each scheduling domain, if there are multiple computing nodes with the same anycast IP, the weight information of each computing node is generated according to the status information of the computing node, and the controller sends the weight information of each anycastIP (entry gateway, exit gateway, computing node entry gateway IP, weight); if there is only one computing node for a certain anycast IP in the scheduling domain, there is no need to generate and send weight information; reduce the amount of information sent.

[0107] Optionally, when delivering weights, the preset weight values ​​of computing power nodes can be delivered by extending BGP / Network Configuration Protocol (Netconf) / Path Computation Element Communication Protocol (PCEP).

[0108] Taking the extended PCEP as an example, the controller initiates the path computation (Path Computation Initiate, PCInitiate), which is actively sent by the path computation element (Path Computation Element, PCE) (undertaken by the controller) to the path computation client (Path Computation Client, PCC) head node (undertaken by the ingress gateway) to actively initialize the computing power node weight information (weight information generated by the controller).

[0109] It is worth noting that in this application Figure 5 and Figure 7 These are all examples, so the mapping relationship between some computing power nodes and gateways is omitted. Figure 7This is an example diagram for determining the weight information for each of the entry gateways. Due to the lack of some mapping relationships, the determination of the weight information is not only based on the content on the current example diagram. For example, in the computing power routing entry gateway IP2, the example diagram shows that only anycast3 is included, and anycast3 only includes the computing power node entry gateway IP5, but the weight is not NA. In fact, this is because the other computing power node entry gateways of anycast3 of the computing power routing entry gateway IP2 are omitted in the example diagram. Therefore, the weight is displayed as 1. Similarly, the weights of other computing power node entry gateways are the same.

[0110] The embodiments of the present application provide two refined weight delivery embodiments, which can be applicable to more scenarios, reduce the amount of information delivered, increase the transmission speed, and, in the case of only one computing power node, there is no need to generate and deliver weight information, which can further reduce the amount of information delivered.

[0111] Optionally, after sending the weight information of the computing power node to the ingress gateway or the scheduling domain, the method further includes:

[0112] If the computing power node information or service deployment changes, determine whether it is necessary to regenerate the weight information of the computing power node and issue it, wherein the service deployment is associated with the computing power node information and the computing power node inlet gateway information connected to the egress gateway;

[0113] If the weight needs to be regenerated and issued, the computing network view is regenerated based on the changed computing node information or the service deployment, and the computing node inlet gateway information connected to the egress gateway;

[0114] Based on the regenerated computing network view, the weight information of the computing power node is regenerated and sent to the entry gateway or scheduling domain.

[0115] Optionally, in some embodiments, when the status information of the computing power node changes or the service deployment changes, it is necessary to first determine whether the weight information of the computing power node needs to be regenerated and issued. If necessary, the node weight needs to be updated through a path calculation update (Path Computation Update, PCupdate) message.

[0116] Specifically, the present application provides some embodiments that require regenerating the weights and issuing them. For example, when the computing power node status information reported by the cloud management changes beyond a certain threshold, the controller will regenerate the weight information and trigger an update; when a new computing power node is added or a new anycast IP is deployed, if the new node meets the conditions for generating weight information in the above embodiments (for example, for each entry gateway or each scheduling domain, the number of computing power nodes corresponding to the new service or the total number of computing power nodes after the new computing power node is added to the original service is greater than or equal to 2), then a new computing network view and the corresponding weight information of the computing power node are regenerated and issued.

[0117] The present application also provides some embodiments that do not require regenerating and distributing the weight information of the computing power node. For example, if the change in the computing power node status information does not exceed a certain threshold, no update is required. For another example, if a new service is added and there is only one corresponding computing power node, there is no need to regenerate a new computing network view and the corresponding weight information of the computing power node and distribute it.

[0118] It is worth noting that the above examples are only used to illustrate the weight update conditions in the embodiments of the present application, and the actual determination of whether to update the weight information is not limited to the above embodiments.

[0119] Please refer to Figure 8 , the embodiment of the present application also provides a method for computing power routing, including:

[0120] Step 21: receiving weight information of a computing power node, wherein the weight information is determined based on a computing network view, wherein the computing network view includes a mapping relationship between an ingress gateway, an egress gateway, and computing power node information;

[0121] Step 22: Make a computing power routing decision based on the weight information.

[0122] In an embodiment of the present application, each inlet gateway or scheduling domain receives the weight information of the computing power node, and the weight information is determined based on the computing network view, and the computing network view includes: the mapping relationship between the inlet gateway, the egress gateway and the computing power node information; and the computing power routing decision is made according to the weight information. In an embodiment of the present application, each inlet gateway or scheduling domain makes a computing power routing decision based on the weight information of the computing power node, and generates and sends the weight information through the controller, thereby reducing the computational overhead and storage overhead of the router, and avoiding the transmission of computing power information in the routing system, reducing the burden on the routing system, and can quickly calculate the route.

[0123] Please refer to Fig. 9 The embodiment of the present application further provides a computing power routing system 30, including:

[0124] Data module 31, used to collect inlet gateway information and computing node information of computing nodes connected to the egress gateway;

[0125] The computing network module 32 is used to determine the computing network view according to the collected inlet gateway information and computing node information of the computing nodes connected to the egress gateway, and the computing network view includes: the mapping relationship between the inlet gateway, the egress gateway and the computing node information;

[0126] The computing network view is used to make computing power routing decisions.

[0127] Optionally, the data module 31 includes:

[0128] A receiving module is used to receive the inlet gateway information of the computing power node connected to the egress gateway, wherein the inlet gateway information of the computing power node includes at least: the egress gateway and the IP address of the inlet gateway of the computing power node;

[0129] Receive computing power node information, the computing power node information includes: the computing power node entry gateway IP, computing power node status information.

[0130] Optionally, the computing power node entry gateway information further includes an anycast address IP, wherein the anycast address IP is used to identify the same type of services;

[0131] The computing network view also includes: a mapping relationship between the computing power node information and different services.

[0132] Optionally, the receiving module includes:

[0133] A transmission module, used to receive the computing power node entry gateway information based on an extended distributed routing protocol;

[0134] Among them, the extended distributed routing protocol is an extended border gateway protocol link state protocol, and the extended border gateway protocol link state protocol includes the path attributes of the extended border gateway protocol link state protocol. The path attributes are used to notify the topology information of the connected computing power nodes and at least carry the computing power node entry gateway IP.

[0135] Optionally, the computing power routing system 30 further includes:

[0136] A weight module, used to determine the weight information of the computing power nodes based on the computing network view, and the weight information is used to make computing power routing decisions;

[0137] The weight information of the computing power node is sent to the entry gateway or the scheduling domain, wherein the scheduling domain includes at least one entry gateway.

[0138] Optionally, the weight distribution module also includes:

[0139] A first issuing module is used to determine, for each of the ingress gateways, whether the number of computing nodes for each type of service is greater than or equal to 2;

[0140] If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the entry gateway;

[0141] If the number of computing nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing nodes;

[0142] or,

[0143] A second issuing module, configured to generate the scheduling domain;

[0144] For each of the scheduling domains, determine whether the number of computing nodes for each type of service is greater than or equal to 2;

[0145] If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the scheduling domain;

[0146] If the number of computing power nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing power nodes.

[0147] Optionally, the weight module also includes:

[0148] An update module, used to determine whether it is necessary to regenerate and distribute the weight information of the computing power node if the computing power node information or service deployment changes, wherein the service deployment is associated with the computing power node information and the computing power node inlet gateway information connected to the egress gateway;

[0149] If the weight needs to be regenerated and issued, the computing network view is regenerated based on the changed computing node information or the service deployment, and the computing node inlet gateway information connected to the egress gateway;

[0150] Based on the regenerated computing network view, the weight information of the computing power node is regenerated and sent to the entry gateway or scheduling domain.

[0151] The computing power routing system provided in the embodiment of the present application can achieve Figure 4 The various processes implemented by the method embodiment and achieving the same technical effect are not described here to avoid repetition.

[0152] The present application embodiment provides an electronic device 40, see Fig.10 As shown, Fig.10This is a principle block diagram of an electronic device 40 according to an embodiment of the present application, including a processor 41, a memory 42, and a program or instruction stored in the memory 42 and executable on the processor 41. When the program or instruction is executed by the processor, the steps in any computing power routing method of the present application are implemented.

[0153] An embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, each process of an embodiment of a computing power routing method as described above is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0154] Computer readable media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0155] The present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the above Figure 1 The various processes of any of the computing power routing method embodiments shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0156] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0157] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0158] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a service classification device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0159] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A computing power routing method, characterized in that: include: Collect the inlet gateway information and computing node information of the computing power node connected to the egress gateway; Determine a computing network view according to the collected inlet gateway information and computing node information of the computing node connected to the egress gateway, wherein the computing network view includes: a mapping relationship between the inlet gateway, the egress gateway, and the computing node information; The computing network view is used to make computing power routing decisions.

2. The method according to claim 1, characterized in that The collecting of the inlet gateway information and the computing node information of the computing node connected to the egress gateway includes: Receive the computing node inlet gateway information connected to the egress gateway, the computing node inlet gateway information at least including: the egress gateway, the computing node inlet gateway IP; Receive computing power node information, the computing power node information includes: the computing power node entry gateway IP, computing power node status information.

3. The method according to claim 2, characterized in that The computing power node entry gateway information also includes an anycast address IP, wherein the anycast address IP is used to identify the same type of service; The computing network view also includes: a mapping relationship between the computing power node information and different services.

4. The method according to claim 2, characterized in that: The receiving egress gateway connected computing node ingress gateway information includes: Based on the extended distributed routing protocol, receiving the computing power node entry gateway information; Among them, the extended distributed routing protocol is an extended border gateway protocol link state protocol, and the extended border gateway protocol link state protocol includes the path attributes of the extended border gateway protocol link state protocol. The path attributes are used to notify the topology information of the connected computing power nodes and at least carry the computing power node entry gateway IP.

5. The method according to any one of claims 1 to 4, characterized in that: Also includes: Based on the computing network view, determine the weight information of the computing power nodes, and the weight information is used to make computing power routing decisions; The weight information of the computing power node is sent to the entry gateway or the scheduling domain, wherein the scheduling domain includes at least one entry gateway.

6. The method according to claim 5, characterized in that The determining the weight information of the computing power nodes based on the computing network view also includes: For each of the entry gateways, determine whether the number of computing nodes for each type of service is greater than or equal to 2; If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the entry gateway; If the number of computing nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing nodes; or, generating the scheduling domain; For each of the scheduling domains, determine whether the number of computing nodes for each type of service is greater than or equal to 2; If the number of computing nodes corresponding to the service is greater than or equal to 2, generate weight information of computing nodes for each type of service corresponding to the scheduling domain; If the number of computing power nodes corresponding to the service is less than 2, there is no need to determine the weight information of the computing power nodes.

7. The method according to claim 5, characterized in that After sending the weight information of the computing power node to the ingress gateway or the scheduling domain, the method further includes: If the computing power node information or service deployment changes, determine whether it is necessary to regenerate the weight information of the computing power node and issue it, wherein the service deployment is associated with the computing power node information and the computing power node inlet gateway information connected to the egress gateway; If the weight needs to be regenerated and issued, the computing network view is regenerated based on the changed computing node information or the service deployment, and the computing node inlet gateway information connected to the egress gateway; Based on the regenerated computing network view, the weight information of the computing power node is regenerated and sent to the entry gateway or scheduling domain.

8. A method for computing power routing, characterized in that: include: Receive weight information of a computing power node, where the weight information is determined based on a computing network view, where the computing network view includes: a mapping relationship between an ingress gateway, an egress gateway, and computing power node information; Based on the weight information, a computing power routing decision is made.

9. A computing power routing system, characterized in that: include: The data module is used to collect the inlet gateway information and computing node information of the computing node connected to the egress gateway; A computing network module, used to determine a computing network view according to the collected inlet gateway information and computing node information of the computing nodes connected to the egress gateway, wherein the computing network view includes: a mapping relationship between the inlet gateway, the egress gateway and the computing node information; The computing network view is used to make computing power routing decisions.

10. An electronic device, characterized in that: include: 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, the steps of the computing power routing method as described in any one of claims 1 to 7 are implemented.

11. A computer-readable storage medium, characterized in that: A computer program is stored on a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the computing power routing method as described in any one of claims 1 to 7 are implemented.

12. A computer program product, characterized in that It comprises computer instructions, which, when executed by a processor, implement the steps of the computing power routing method as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Backhaul traffic drainage method and device in computing power network, equipment and storage medium

    CN115695284A

  • Computing power network state updating method and device, equipment and storage medium

    CN116760704A

  • Computing power information notification method, computing power network element node, device and storage medium

    CN117201382A

  • Information transmission method and device, related equipment and storage medium

    CN118612215A