Computing power routing determination method, communication device and product

Abstract

The embodiment of the invention provides a computing power routing determination method, communication equipment and a product. The method comprises the steps that first network equipment sends a telemetering subscription request for a target computing power instance to second network equipment; the first network device receives computing power information of the target computing power instance sent by the second network device in response to the telemetering subscription request; and the first network equipment determines a computing power routing result reaching the target computing power instance according to the computing power information and path quality information, and the path quality information is quality information of a segmented routing strategy instance between the first network equipment and the second network equipment. The embodiment of the invention aims to reduce the configuration complexity while meeting the computing power routing calculation.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of communication technologies, and in particular, to a method for determining computing power routing, a communication device, and a product. Background Art

[0002] Traditional network routing only selects by perceiving network-level metrics, while computing power routing requires dual perception of the network and computing resources to implement routing dimension elevation. Facing the challenges of the ubiquity and high dynamics of computing power resources (including native computing power such as CPUs and GPUs, and service computing power such as AI algorithm services), the industry is researching related perception and notification technologies to support computing power gateways, routers, and computing network controllers to achieve integrated computing network perception and routing.

[0003] In related technologies, collecting computing power resources by docking with a cloud management platform or extending the Border Gateway Protocol (BGP) involves network scale upgrade and high configuration complexity. How to meet the computing power routing calculation while reducing the configuration complexity is an issue that needs to be discussed and solved urgently. Summary of the Invention

[0004] Embodiments of the present application provide a method for determining computing power routing, a communication device, and a product, aiming to meet the computing power routing calculation while reducing the configuration complexity.

[0005] In a first aspect, an embodiment of the present application provides a method for determining computing power routing, which is applied to a first network device. The method includes: sending a telemetry subscription request for a target computing power instance to a second network device; receiving the computing power information of the target computing power instance sent by the second network device in response to the telemetry subscription request; and determining a computing power routing result for reaching the target computing power instance according to the computing power information and path quality information, where the path quality information is the quality information of a segment routing policy instance between the first network device and the second network device.

[0006] In a second aspect, an embodiment of the present application provides a method for determining computing power routing, which is applied to a second network device. The method includes: receiving a telemetry subscription request for a target computing power instance sent by a first network device; and in response to the telemetry subscription request, sending the computing power information of the target computing power instance to the first network device, where the computing power information is used by the first network device to determine a computing power routing result for reaching the target computing power instance according to the computing power information and path quality information, and the path quality information is the quality information of a segment routing policy instance between the first network device and the second network device.

[0007] In a third aspect, an embodiment of the present application provides a communication device, including: at least one processor; at least one memory for storing at least one program; when the at least one program is executed by the at least one processor, the computing power routing determination method described in the first aspect or the second aspect is implemented.

[0008] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-executable instructions for executing the computing power routing determination method described in the first aspect or the second aspect.

[0009] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program or computer instructions, characterized in that the computer program or the computer instructions are stored in a computer-readable storage medium, and a processor of a computer device reads the computer program or the computer instructions from the computer-readable storage medium, and the processor executes the computer program or the computer instructions, so that the computer device executes the computing power routing determination method described in the first aspect or the second aspect.

[0010] In an embodiment of the present application, a first network device subscribes to the computing power information of a computing power instance corresponding to a required target service from a second network device through a telemetry mechanism. The first network device performs computing power routing calculation based on the obtained computing power information of the target computing power instance and path quality information. By subscribing to computing power information through the telemetry mechanism, the requirements for the announcement and routing calculation of computing power information can be met without modifying the existing routing protocol and without upgrading and modifying the network scale. Through the above method, the deployment and configuration processes can be simplified, with the characteristics of simplified operation and low invasiveness, effectively avoiding the problem of great impact on the existing network caused by the need to modify the protocol or network scale in the prior art, and significantly reducing the system pressure brought by the introduction of computing power information with high dynamics and high ubiquity into the routing protocol.

[0011] Other features and advantages of the present application will be described in the subsequent specification, and, in part, will be obvious from the specification, or will be understood by implementing the present application. The objectives and other advantages of the present application can be realized and obtained through the structures specifically pointed out in the specification, the claims, and the drawings. Description of the Drawings

[0012] Figure 1 A deployment diagram of computing power awareness and announcement provided for an example of related technologies;

[0013] Figure 2 A schematic diagram of the network architecture to which the computing power routing determination method provided by an embodiment of the present application is applied;

[0014] Figure 3 Flowchart of the computing power routing determination method provided by an embodiment of this application;

[0015] Figure 4 Configuration instance diagram of the telemetry subscription request of the client (first network device) provided by an example of this application;

[0016] Figure 5 Schematic diagram of the message content of the SATMP extended computing power information acquisition request provided by an example of this application;

[0017] Figure 6 Schematic diagram of the message content of the response information of the STAMP extended computing power information acquisition request provided by an example of this application;

[0018] Figure 7 Schematic diagram of the computing power instance configuration Telemetry server instance provided by an example of this application;

[0019] Figure 8 Schematic diagram of the first table entry instance provided by an example of this application;

[0020] Figure 9 Schematic diagram of the defined YANG model instance provided by an example of this application;

[0021] Figure 10 Schematic diagram of the subscription session table provided by an example of this application;

[0022] Figure 11 Schematic diagram of the third table entry instance provided by an example of this application;

[0023] Figure 12 Schematic diagram of the second table entry instance provided by an example of this application;

[0024] Figure 13 Schematic diagram of the computing power routing result table entry instance provided by an example of this application;

[0025] Figure 14 Flowchart of the computing power routing determination method provided by another embodiment of this application;

[0026] Figure 15 Schematic diagram of the structure of the communication device provided by an embodiment of this application. Detailed implementation manners

[0027] In order to make the objectives, technical solutions and advantages of this application clearer, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application.

[0028] It should be noted that although the functional modules are divided in the device schematic diagram and the logical sequence is shown in the flowchart, in some cases, the steps shown or described may be executed in a different module division from that in the device or a different order from that in the flowchart. Terms such as "first" and "second" in the description of the specification, claims, and the above-mentioned drawings are used to distinguish similar objects and do not necessarily describe a specific order or sequence.

[0029] In the description of the embodiments of the present application, unless otherwise clearly defined, words such as "set", "installed", and "connected" should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above words in the embodiments of the present application in combination with the specific content of the technical solution.

[0030] In the embodiments of the present application, words such as "furthermore", "exemplarily", or "optionally" are used to represent examples, illustrations, or explanations, and should not be construed as being more preferred or having more advantages than other embodiments or design solutions. The use of words such as "furthermore", "exemplarily", or "optionally" is intended to present relevant concepts in a specific manner.

[0031] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: Wideband Code Division Multiple Access (WCDMA) mobile communication systems, Evolved Universal Terrestrial Radio Access Network (E-UTRAN) systems, Next Generation Radio Access Network (NG-RAN) systems, Long Term Evolution (LTE) systems, Worldwide Interoperability For Microwave Access (WiMAX) communication systems, 5th Generation (5G) systems, such as New Radio Access Technology (NR), and future communication systems, such as 6G systems.

[0032] The technical solutions of the embodiments of the present application can be applied to various communication technologies, such as microwave communication, optical wave communication, millimeter wave communication, etc. The specific technologies and specific device forms adopted in the embodiments of the present application are not limited.

[0033] In related technologies, traditional network routing only makes selections by perceiving network-level metrics, while computing power routing needs to achieve dual perception of network and computing resources to implement routing dimension elevation. Facing the challenges of the ubiquity and high dynamicity of computing power resources (including native computing power such as CPU (Central Processing Unit), GPU (Graphics Processing Unit), and service computing power such as AI algorithm services), the industry is researching related perception and notification technologies to support computing power gateways, routers, and computing network controllers, and to adapt to cloud deployment to achieve integrated perception and routing of the computing network.

[0034] Figure 1 A deployment schematic diagram of computing power perception and notification provided for an example of related technologies. Refer to Figure 1 , there are mainly two existing computing power perception and notification schemes:

[0035] Solution 1: As shown in Opition1 in Figure 1 , the computing network controller docks with the northbound interface of the cloud management platform to obtain computing power information. The cloud management platform collects and outputs the computing power information. In order to perceive the gateway to which the computing power resources belong, it is also necessary to perform BGP (Border Gateway Protocol) extension.

[0036] Solution 2: As shown in Opition2 in Figure 1 , a BGP plugin is added to the computing power resources to collect computing power information and establish a BGP neighbor with the egress gateway (EGW) to exchange computing power information. In order to notify the ingress gateway (IGW) and the computing network controller of the computing power information, it is also necessary to extend the BGP address family or Pathattribute (path attribute) to achieve this.

[0037] The above two solutions also have the following problems:

[0038] Solution 1 collects data by docking with the cloud management platform and cannot directly provide the specific EGW gateway information to which the computing power instance is attached. It also requires separate reporting or manual configuration with relatively high complexity. The collection frequency and capabilities completely depend on the capabilities of the existing cloud management platform. In addition, if distributed computing is to be supported, it is also necessary to notify the ingress gateway of the computing power information by extending the north-south protocol.

[0039] Solution 2: 1) When a soft router is deployed on the computing power side: It is necessary to consider the impact of the firewall on the computing side, and the deployment and configuration are complex. The resource overhead of running the routing protocol is large, and hardware acceleration cannot be directly achieved. In addition, only computing side metrics can be obtained, and the full-segment metrics from the EGW to the computing side cannot be obtained, resulting in incomplete coverage; 2) Expanding the BGP address family or Path attribute in the network domain: The computing power information changes frequently and the standards are not unified, which has a large impact on the network routing control plane and strong invasiveness. When deploying an RR (Route Reflector) reflector in the bearer network, the reflector software also needs to be modified to support the distribution of new address families and PA (Path attribute) attributes, which involves the upgrade and deployment of key nodes in the bearer network.

[0040] In summary, collecting computing power resources by docking with the cloud management platform or expanding the border network management protocol will involve the upgrade of the network scale and the problem of high configuration complexity. How to meet the computing power routing calculation while reducing the configuration complexity is an urgent problem to be discussed and solved currently.

[0041] Based on this, the embodiment of the present application provides a method for determining computing power routing. The first network device uses the telemetry mechanism to subscribe to the computing power information of the target computing power instance required from the second network device. The first network device performs computing power routing calculation based on the obtained computing power information and path quality information of the target computing power instance. By subscribing to the computing power information through the telemetry mechanism, the requirements of computing power information notification and routing calculation can be met without modifying the existing routing protocol and without upgrading and modifying the network scale. Through the above method, the deployment and configuration process can be simplified, with the characteristics of simplified operation and low invasiveness, effectively avoiding the problem of large impact on the existing network caused by modifying the protocol or network scale in the prior art, and significantly reducing the system pressure brought by introducing computing power information with high dynamicity and high ubiquity into the routing protocol.

[0042] The following further elaborates on the embodiment of the present application in conjunction with the accompanying drawings.

[0043] Figure 2 It is a schematic diagram of the network architecture applied to the method for determining computing power routing provided by an embodiment of the present application. As Figure 2 shown, the network architecture at least includes but is not limited to: terminal 210, computing network controller 220, entrance computing power gateway 230, exit computing power gateway 240, and resource pool 250.

[0044] Among them, the terminal 210 is a device that uses the computing-network routing service. The terminal 210 includes various device forms such as desktop computers, laptops, PDAs (Personal Digital Assistants), mobile phones, vehicle-mounted terminals, home theater terminals, dedicated terminals, etc. It can also be a single device or a collection of multiple devices. For example, multiple devices are connected through a local area network and share a display device for collaborative work, jointly constituting a terminal. The terminal 210 is communicatively connected to the ingress computing power gateway 230 to obtain computing power services.

[0045] The computing-network controller 220 is mainly responsible for collecting information on computing power resources and network resources, performing network path calculation and distribution, and is responsible for unified orchestration, dynamic scheduling, and collaborative optimization of computing power resources (such as servers, GPUs, edge nodes) and network resources (such as switches, routers, links). The computing-network controller 220 can be but is not limited to being deployed in servers or virtual machines. The computing-network controller 220 can be communicatively connected to the ingress computing power gateway 230 and send computing power routing information to the ingress computing power gateway 230. The computing-network controller 220 can also be communicatively connected to the egress computing power gateway 240 and obtain the computing power information of relevant computing power instances in the resource pool 250 from the egress computing power gateway 240.

[0046] The ingress computing power gateway 230 interfaces with the terminal and realizes addressing and routing forwarding of the computing-network integrated service in terms of business. In this application, it mainly receives the computing power announcements pushed by telemetry and performs distributed computing power routing calculation; the ingress computing power gateway 230 is mainly responsible for processing the process of external requests accessing the computing power network, that is, the computing power resource application, data interaction, or service access initiated from the terminal 210 to the inside of the computing power network.

[0047] The egress computing power gateway 240 interfaces with the resource pool 250. In terms of business, it mainly forwards the computing power request message to the corresponding computing power instance. In this application, it is mainly responsible for collecting and processing the computing power information from the resource pool 250, and the method can be active or passive, and notifies the ingress computing power gateway 230 and / or the computing-network controller 220 of the computing power information through the method of telemetry dynamic subscription; the egress computing power gateway 240 mainly processes the process of the computing power instance resources inside the computing power network providing computing power services to the external network, that is, the computing power instances inside the computing power network send data, results, or response requests to the outside. The ingress computing power gateway 230 and the egress computing power gateway 240 are communicatively connected through the underlying network (Underlay).

[0048] The resource pool 250 deploys multiple computing power instances (INS1, INS2,... INSn) for providing computing power services. In some examples, the computing power instances in the resource pool 250 are divided according to services. The resource pool 250 can contain multiple different services. One service can correspond to multiple computing power instances, and the same computing power instance can only belong to one service.

[0049] The egress computing power gateway 240 collects corresponding computing power information from each computing power instance in the resource pool 250 and provides it to the computing network controller 220 and / or the ingress computing power gateway 230.

[0050] For Figure 2 example, exemplarily, the ingress computing power gateway 230 sends a telemetry subscription request for a target computing power instance to the egress computing power gateway 240; wherein, the target computing power instance is a computing power instance in the resource pool 250 for providing a target computing power service.

[0051] The ingress computing power gateway 230 receives the computing power information of the target computing power instance sent by the egress computing power gateway 240 in response to the telemetry subscription request (as shown by the dotted line between the egress computing power gateway 240 and the ingress computing power gateway 230 in Figure 2 ).

[0052] The ingress computing power gateway 230 determines a computing power routing result to the target computing power instance according to the computing power information and the path quality information, wherein the path quality information is the quality information of the segmented routing policy instance between the ingress computing power gateway 230 and the egress computing power gateway 240.

[0053] The ingress computing power gateway 230 performs transmission scheduling on the data sent by the terminal 210 according to the obtained computing power routing result.

[0054] In another example, also taking Figure 2 as an example, the computing network controller 220 sends a telemetry subscription request for a target computing power instance to the egress computing power gateway 240; wherein, the target computing power instance is a computing power instance in the resource pool 250 for providing a target computing power service.

[0055] The computing network controller 220 receives the computing power information of the target computing power instance sent by the egress computing power gateway 240 in response to the telemetry subscription request (as shown by the dotted line between the egress computing power gateway 240 and the computing network controller 220 in Figure 2 ).

[0056] The computing network controller 220 determines a computing power routing result to the target computing power instance according to the computing power information and the path quality information, wherein the path quality information is the quality information of the segmented routing policy instance between the ingress computing power gateway 230 and the egress computing power gateway 240.

[0057] The computing network controller 220 sends the computing power routing result to the ingress computing power gateway 230 (as shown by the dotted line in Figure 2(as shown by the solid line between the ingress computing power gateway 230 and the computing network controller 220). The ingress computing power gateway 230 performs transmission scheduling on the data sent by the terminal 210 according to the obtained computing power routing result. In this example, the computing network controller 220 can obtain the computing power information of all computing power instances from the egress computing power gateway 240, and perform centralized computing network routing calculation by combining the path quality information between each ingress computing power gateway 230 and each egress computing power gateway 240, and then send the obtained computing power routing result to the ingress computing power gateway 230 for scheduling, which has more advantages for global network balance.

[0058] In the above example, the ingress computing power gateway and / or the computing power controller use the telemetry mechanism to subscribe to the computing power information of the required target computing power instances from the egress computing power gateway. The ingress computing power gateway and / or the computing power controller perform computing power routing calculation according to the obtained computing power information of the target computing power instances and the path quality information. By subscribing to the computing power information through the telemetry mechanism, it is possible to meet the requirements of computing power information notification and routing calculation without modifying the existing routing protocol and without upgrading and modifying the network scale. Through the above method, the deployment and configuration process can be simplified, with the characteristics of simplified operation and low invasiveness, effectively avoiding the problem of great impact on the existing network caused by modifying the protocol or network scale in the prior art, and significantly reducing the system pressure brought by introducing the computing power information with high dynamicity and high ubiquity into the routing protocol.

[0059] Figure 3 It is a flowchart of a computing power routing determination method provided by an embodiment of the present application. As Figure 3 shown, the computing power routing determination method can be but is not limited to being applied to a first network device, such as an ingress computing power gateway, a computing network controller, etc., or an ingress computing power gateway 230 and a computing network controller 220 as Figure 2 shown. In this embodiment, the computing power routing determination method at least includes but is not limited to:

[0060] Step 310: Send a telemetry subscription request for a target computing power instance to a second network device;

[0061] Step 320: Receive the computing power information of the target computing power instance sent by the second network device in response to the telemetry subscription request;

[0062] Step 330: Determine a computing power routing result to reach the target computing power instance according to the computing power information and the path quality information, where the path quality information is the quality information of the segment routing policy instance between the first network device and the second network device.

[0063] In step 310, the second network device is responsible for actively or passively collecting and processing computing power information from the resource pool 250, and is also responsible for notifying the first network device of the computing power information through telemetry dynamic subscription. The second network device is an egress computing power gateway (EGW).

[0064] Telemetry refers to the process of collecting status data (such as CPU utilization, link bandwidth, temperature, alarm information, etc.) from network devices, servers, sensors, and other devices in real time and transmitting it to the data collection device / system. Telemetry dynamic subscription means that the data collector dynamically specifies the data type, collection frequency, transmission target and other parameters to be collected according to demand, rather than a fixed collection mode.

[0065] The telemetry subscription request is the request information generated by the first network device and sent to the second network device based on the telemetry dynamic subscription mechanism according to the demand. For example, the demand can be the service type that currently needs to perform computing power routing calculation, the target computing power instance that the first network device expects to collect, etc.

[0066] Exemplarily, the dynamic mode of Telemetry is adopted, and the second network device is always configured as the server, and the first network device is configured as the client (i.e., the data collector). The telemetry configuration is defined on the client and the server respectively, wherein the client telemetry configuration includes at least a service identifier, a server IP, an event mode, or a collection interval mode. The service identifier is used to indicate the type of computing service, the server IP is the node address of the second network device, the event mode refers to the client requiring the server to push information based on a preset event or condition trigger, and the collection interval mode refers to the client requiring the server to push information based on a preset time interval.

[0067] The first network device (client) actively initiates a telemetry subscription request to the second network device according to the configured server IP to establish a connection with the second network device as the server, and carries the service identifier, server IP, event mode or collection interval mode information in the request.

[0068] For example, Figure 4 The client (first network device) telemetry subscription request configuration example diagram provided in an example of this application is as follows: Figure 4 As shown, configure the telemetry client on the IGW or computing network controller, specifically including: 1) specifying the GRPC service IP address 11::11 and the service port 57400; 2) creating a subscription request, in which the request carries the encoding method (GPB), the time interval or event indication, and the service identification ID list that needs attention (1111::1, 2222::2, 3333::3.......); 3) initiating a client connection to start the subscription.

[0069] Configure the telemetry server on the EGW, specifically including: 1) Start the GRPC service with the service port being 57400; 2) Add the relevant information of the subscription destination group DG1, specifying the corresponding client IP, port, and encoding method (GPB).

[0070] 3) Add the sensor group SG1, specifying the collection object as the computing power information YANG model; 4) Add SG1 and DG1 to the dynamic subscription group. The EGW receives the client subscription request as the server, generates the computing power information required by the client according to the subscription request, and pushes it to the client.

[0071] In step 320, the target computing power instance refers to the computing power instance that the first network device expects to collect computing power information, or the computing power instance corresponding to the computing power service for which the first network device expects to calculate the computing power route. The target computing power instance can also be the computing power instance corresponding to the target service subscribed by the first network device. The computing power information is the information such as the location and status of the corresponding computing power instance.

[0072] In some examples, the first network device subscribes to the computing power information of the computing power service (i.e., the target service) from the second network device. The second network device determines the computing power instances included in the target service based on the target service indicated by the subscription request, that is, obtains the target computing power instances.

[0073] Exemplarily, when the server (the second network device) receives the telemetry subscription request sent by the client (the first network device), it will select the computing power information of the computing power instance corresponding to the telemetry subscription request from the computing power information collected by itself and push it to the client. Specifically, the server will generate a subscription session table according to the telemetry subscription request. The subscription session table includes service identification, client IP, event mode, or collection interval mode information. The server selects the computing power information of the computing power instance corresponding to the service identification in the subscription session table from the collected computing power information, uses the client IP as the target IP of the Telemetry information, and obtains the Telemetry information to be pushed. The Telemetry information contains the computing power information of the target computing power instance. The server then pushes the Telemetry information to the client according to the event mode or collection interval mode information in the subscription session table according to the corresponding notification method (notification based on event trigger or notification based on time interval).

[0074] In step 330, the path quality information refers to the quality information of each network path between the first network device and the second network device.

[0075] Segment Routing (SR) is a new network routing technology based on the source routing concept. By splitting the end-to-end path into multiple "segments", the source node (or controller) pre-plans the path and assigns segment identifiers (Segment IDs, SIDs) to data packets, enabling flexible control of network traffic.

[0076] A segment routing policy instance is a routing path instance determined between a first network device and a second network device based on segment routing technology, that is, each network path between the first network device and the second network device.

[0077] The computing power routing result refers to the end-to-end optimal computing power instance that can meet the user's needs and the corresponding routing path.

[0078] Based on the collected computing power information and path quality information, filter the computing power instances and segment routing policy instances between the first network device and the second network device that can meet the user's needs or service quality requirements and other requirements, and combine them to form a routing path from the terminal to the computing power instance, that is, the computing power routing result.

[0079] In the embodiment of the above steps 310 to 330, the first network device uses the telemetry mechanism to subscribe to the computing power information of the required target computing power instance from the second network device. The first network device performs computing power routing calculation based on the obtained computing power information and path quality information of the target computing power instance. By subscribing to computing power information through the telemetry mechanism, it is possible to meet the requirements of computing power information notification and routing calculation without modifying the existing routing protocol and without upgrading and modifying the network scale. Through the above method, the deployment and configuration process can be simplified, with the characteristics of simplified operation and low intrusion, effectively avoiding the problem of great impact on the existing network caused by the need to modify the protocol or network scale in the prior art, and significantly reducing the system pressure brought by the introduction of high-dynamic and high-ubiquitous computing power information into the routing protocol.

[0080] The above is the overall description of steps 310 to 330. The following will describe the specific implementation process of steps 310 to 330 in detail.

[0081] In one embodiment, the computing power information is collected by the second network device from the computing power instances. The steps for the second network device to collect the computing power information are as follows:

[0082] The second network device sends a computing power information acquisition request to each computing power instance, where the computing power information acquisition request carries an indication of the collection information list;

[0083] The second network device receives the computing power information responded by each computing power instance, where the computing power information is obtained by the computing power instance collecting the current metric data according to the collection information list and encapsulating it.

[0084] In this embodiment, the computing power information acquisition request is a command message used to request the computing power instance to send the computing power information. Exemplarily, the computing power information acquisition request can be a request for collecting computing power information initiated by extending the OAM (Operations, Administration, and Maintenance) protocol. The OAM protocol can include three-layer OAM, including but not limited to ping, BFD (Bidirectional Forwarding Detection), TWAMP (Two-Way Active Measurement Protocol), OWAMP (One-Way Active Measurement Protocol), STAMP (Simple Two-Way Active Measurement Protocol), etc.

[0085] The computing power information acquisition request carries a collection information list indication, and the collection information list indication is used to indicate which information the computing power instance needs to collect and the data encapsulation format of the collected information. Exemplarily, the collection information list indication can adopt the form of TLV indication; the TLV indication (TLV Indication) generally refers to specific status, event, or control information transmitted through the TLV (Type-Length-Value) structure, and is used to instruct the receiving party to perform a certain operation or handle a specific situation.

[0086] Exemplarily, the collected computing power information includes but not limited to service identifier, computing power instance IP, CPU occupancy rate, memory occupancy rate, IO bandwidth, processing delay, etc.; among them, the service identifier is used to characterize the service type provided by the computing power instance, and is registered by the computing resource management system to the network domain (usually the computing network control) to provide the computing network integrated service, and the network domain returns it to the computing resource management system to obtain.

[0087] Exemplarily, taking the OAM protocol and STAMP as an example, the active collection of computing power information by the second network device is described:

[0088] ① The egress computing power gateway sends an active OAM carrying the computing power information acquisition request:

[0089] The egress computing power gateway sends an OAM (operation, administration, and maintenance) request message carrying a computing power resource collection command, i.e., a computing power information acquisition request. Among them, the source address of the STAMP OAM message is the corresponding interface address IP1 of the egress computing power gateway connecting to the computing power instance, and the destination address is the address IP2 of the computing power instance. Each computing power instance has a unique IP address.

[0090] ② The OAM request carries a collection information list indication:

[0091] Specific STAMP request protocol extension implementation example reference Figure 5 . The TLV mechanism is adopted, where type=9 identifies the computing power request TLV, length=4 identifies the command list byte length, and the Value command list uses a bit list to specify the type of computing power resource indicators that need to be collected. Different bits are set to 1 to represent the collection of corresponding information, such as CPU occupancy, memory occupancy, inbound bandwidth, outbound bandwidth, processing latency, etc. These commands support arbitrary combinations and incremental expansion, and can flexibly request computing power information of a specific combination according to actual needs, so as to achieve accurate perception of the computing power status of the computing power instance.

[0092] ③The computing instance receives the OAM message and responds:

[0093] After the STAMP OAM module deployed in the computing power instance receives the OAM message carrying the collection information list indication, it obtains the required local computing power status data according to the command in the message. The local status data can be real-time monitoring data stored in the computing power instance's own monitoring software. The OAM module encapsulates the acquired computing power status data into the OAM response message. In addition, it also carries a service identifier in the response message to identify the service type to which the computing power instance belongs, and applies and responds to the corresponding egress computing power gateway. In the response message, the destination address is the source address IP1 of the request message, and the source address is the destination address IP2 of the request message, ensuring that the response message can be accurately returned to the egress computing power gateway that initiated the request, thereby completing the collection and feedback process of the computing power information. For details, refer to the structure of the response information based on the STAMP extended computing power information acquisition request. Figure 6 , where type=10 represents the computing power collection response, length represents the TLV length, RET code represents the return code, Reserved is a reserved extension, Service ID is a service identifier, and Optional sub-TLVs is a sub-TLV that carries the computing power status information required in the request list. Each status can be carried by a corresponding sub-TLV.

[0094] In the above embodiment, the second network device collects the required computing power information from the computing power instances in an active request manner, and the collection is more flexible and accurate, facilitating the modification of the computing power information acquisition request for different computing power information required in different situations.

[0095] In another embodiment, the second network device can also collect computing power information through the following steps:

[0096] The second network device receives the computing power information sent by each computing power instance according to the static subscription configuration; wherein, the static subscription configuration is used to instruct the computing power instance to push the computing power information to the second network device.

[0097] In this embodiment, the static subscription configuration refers to setting the data subscription relationship in a predefined manner, enabling the data source to send specific types of data to the target according to fixed rules. Compared with telemetry dynamic subscription, the subscription rules of static subscription will not be automatically adjusted dynamically after configuration.

[0098] Exemplarily, passive collection is achieved by deploying the Telemetry module on the computing power instance. For example, the monitoring module deployed on the resource pool where the computing power instance is located collects the corresponding computing power information, and the monitoring module actively pushes it to the EGW connected to the computing power instance according to the preset rules by using the Telemetry mechanism.

[0099] Exemplarily, the specific process for the EGW to collect computing power information in a passive manner is as follows:

[0100] ① Deploy Telemetry on the computing power instance and configure it as a client:

[0101] Using the existing cloud resource monitoring and collection mechanism, deploy the Telemetry protocol in the computing power instance resource pool and configure it in client mode. By deploying the Telemetry protocol, detailed metrics of each computing power instance can be collected in real time, including but not limited to key performance metrics such as CPU occupancy, memory occupancy, IO bandwidth, and processing latency. These metrics can comprehensively reflect the current operating state and available computing power resources of the computing power instance. For specific configuration examples, refer to Figure 7 , where the source-ip identifies the IP address of the target device for pushing, usually the EGW interface IP.

[0102] ② Configure the EGW as a telemetry server:

[0103] Configure the Telemetry server on the egress computing power gateway so that the egress computing power gateway can receive the computing power information of each computing power instance belonging to this gateway. The computing power instance, as a client, establishes a communication connection with the EGW node through the Telemetry protocol.

[0104] ③The computing power instance pushes the collected computing power information to the EGW through Telemetry according to the EGW configuration:

[0105] The computing power instance actively sends the collected computing power information to the corresponding EGW node through the Telemetry mechanism. The information format pushed by Telemetry is jointly defined by the computing side and the network side, including but not limited to key performance indicators such as CPU occupancy rate, memory occupancy rate, IO bandwidth, and processing delay, as well as the service identification information corresponding to the computing power instance. Among them, the service identification source is applied by the cloud provider from the operator according to the service type and is used to identify different service types.

[0106] The above embodiment realizes the collection of computing power information by the second network device in a passive manner through static subscription, with high stability and strong predictability.

[0107] In one embodiment, the EGW receives the computing power information data collected through the active / passive method provided in the above embodiment, and determines the VPN SID (Virtual Private Network Service Instance Identification) information according to the local interface connecting to the computing power instance. Among them, the VPN SID includes END.DT (Data Terminal Endpoint) or END.DX (Data Exchange Endpoint), which is usually obtained through static configuration or dynamic allocation in the EGW. The EGW forms a first table entry together with the collection result (as Figure 8 shown). The specific table entry content includes: the IP address of the computing power instance, service identification, connection interface, VPN SID, and the status information of the computing power instance (i.e., the computing power instance metrics, including CPU occupancy rate, memory occupancy rate, IO bandwidth, and processing delay).

[0108] Among them, the service identification: identifies the service type provided by the computing power instance. The service identification can be a digital identification or an anycast IP address; the computing power instance IP: corresponds to the unique index of the computing power instance;

[0109] Connection interface: the interface related to the EGW docking the computing power instance;

[0110] VPN SID: The END.DX allocated by the EGW for the relevant interface connecting to the corresponding computing power instance. In specific cases, it can also be END.DT, which is ultimately used to carry the forwarding behavior indication in the service request. Usually, each VPN SID is allocated or statically configured by the EGW;

[0111] Computing power instance metrics: include but not limited to key performance indicators such as CPU occupancy rate, memory occupancy rate, IO bandwidth, and processing delay.

[0112] In one embodiment, the EGW can optimize the collected computing power information data according to the computing network constraint conditions to reduce the pressure on the network domain. Here, SLA stands for Service Level Agreement, which is a service level agreement. Exemplarily, after the EGW passes the computing power information data in the form of the first entry shown in Figure 8 , the EGW can optimize all records with the same service identifier and select the entries that finally need to be announced. The specific steps are as follows: The EGW collects all records for collection, as specifically shown in Figure 8 . It filters all records related to the same service identifier; it optimizes according to the rules predefined for a specific service identifier. For example, it selects a computing power instance where the mem occupancy corresponding to the service identifier 1111::1 is less than 32% and the delay is less than 6 ms. Finally, it determines that the entry corresponding to the computing power instance IP 10:230:152::201 needs to be announced.

[0113] In one embodiment, the computing power information is encapsulated by a second network device based on the telemetry data structure model for the address information and status information of the target computing power instance.

[0114] In this embodiment, the telemetry data structure model is a standardized data structure model defined using the YANG model mechanism. The YANG (Yet Another Next Generation) model is a data modeling language used to define the configuration and status data structures of network devices. Exemplarily, the YANG model provides a standardized format for the data related to the computing power information collected by Telemetry, enabling the data receiving end to more easily parse and understand this data. Regardless of which network device the data comes from, as long as it follows the same YANG model, the structure and meaning of the data are consistent, which greatly simplifies the complexity of data parsing. As shown in Figure 9 , in order to announce the computing power information in the network domain, a telemetry data structure model of the computing power information as shown in the figure is established. This telemetry data structure model at least includes a service identifier, a server IP, an EGW node IP, a VPN SID, and a computing power instance METRIC indicator, thus ensuring the interoperability of announcing the computing power information in the network domain.

[0115] Exemplarily, the EGW performs Telemetry encapsulation on the computing power information entries of the optimized computing power instances. The encapsulation mainly follows the GRPC encapsulation protocol and the YANG model defined as in the above example. The specific steps are as follows: Combine with the Figure 9 defined YANG model to encapsulate and carry the Telemetry message; among them, according to the Telemetry subscription request configuration shown in Figure 4 , it is required that the data model uses GPB encoding and the interface protocol uses GRPC. For Figure 8For the content of the first table entry shown, the computing power information is encapsulated into a Telemetry message, and the computing power information of multiple computing power instances can be carried in one message to ensure the efficiency of pushing. Ensure that the Telemetry message format complies with GRPC and YANG standards, facilitating client parsing and processing.

[0116] Exemplarily, the computing power information pushed by the EGW at one time includes at least the computing power instance IP, service identifier, VPN SID, EGW node IP (the second network device node address), and computing power instance status information. The EGW and the IGW and / or the computing network controller perform dynamic subscription and pushing of computing power information through the Telemetry mechanism. The entire process does not involve the participation and processing of routing protocols such as IGP / BGP, with high flexibility and little impact on existing protocols. In some examples, the EGW can encapsulate the content that multiple first table entries need to announce (i.e., the computing power information data that needs to be announced to the first network device) in one message for pushing to improve the announcement efficiency.

[0117] In one embodiment, the telemetry subscription request further carries an event pattern; correspondingly, step 320 includes:

[0118] Receiving the computing power information of the target computing power instance sent by the second network device in response to meeting the event condition after receiving the telemetry subscription request, where the event condition is determined by the second network device according to the event pattern.

[0119] In this embodiment, the event condition refers to the triggering condition for triggering the second network device to push the computing power information to the first network device. The event condition is extracted and confirmed by the second network device from the event pattern-related data carried in the telemetry subscription request sent by the received first network device. For example, the event pattern indicates immediate notification when the computing power state changes. Correspondingly, the event condition is that there is a change in the computing power state in the computing power information collected by the second network device. When there is a change in the computing power state in the computing power information collected by the second network device, the second network device pushes the computing power information to the first network device.

[0120] The above embodiment can reduce unnecessary data transmission and save resources by triggering the second network device to push the computing power information to the first network device based on an event.

[0121] In one embodiment, the telemetry subscription request further carries a time interval pattern; correspondingly, step 320 includes:

[0122] Receiving the computing power information of the target computing power instance sent by the second network device at the interval time corresponding to the time interval pattern after receiving the telemetry subscription request.

[0123] In this embodiment, the time interval mode is the indication content carried in the telemetry subscription request for instructing the second network device to push computing power information at a defined time interval.

[0124] In one embodiment, the EGW receives, as a server, a telemetry subscription request sent by a first network device acting as a client, and generates a subscription session table according to the subscription information carried in the telemetry subscription request. Figure 10 The following is a schematic diagram of the subscription session table provided by an example of this application. As Figure 10 shown, where the client IP is the first network device node address; the notification mode is determined according to the event mode or time interval mode field carried in the telemetry subscription request. When the notification mode field is 0, it means that the event condition mechanism is used to trigger the notification. For the same service identifier, there can be multiple client IPs, indicating that multiple IGWs or computing network controllers have jointly subscribed to the computing power information of the service type corresponding to the service identifier.

[0125] Specifically, the EGW sends the computing power information to the corresponding clients in sequence according to the Figure 10 subscription session table shown as follows. The specific steps are as follows: for the computing power information entry of the preferred computing power instance, the EGW obtains its service identifier, and queries the IP addresses and notification modes of all its corresponding clients in the subscription session table according to this service identifier; according to the notification mode, if it is the event mode, when the preset variable meets the event condition, the EGW initiates the push of the computing power information, and if it is the time interval mode, the push is initiated at a fixed interval. According to the client IP, the encapsulated computing power information is sent to the ingress computing power gateway and / or the computing network controller in sequence. In the Telemetry message, the source address uses the EGW node IP, and the destination address uses the addresses corresponding to the client IP column in the subscription session table respectively.

[0126] In one embodiment, before step 330, it further includes:

[0127] Determine all segment routing policy instances between the first network device and the second network device;

[0128] Obtain the network quality metrics corresponding to each segment routing policy instance;

[0129] Integrate the path quality information according to each segment routing policy instance and the corresponding network quality metrics.

[0130] In this embodiment, the network quality metric is an index that can reflect the network performance of the corresponding segment routing policy instance, such as available bandwidth (ABW), delay, packet loss rate (los), etc.

[0131] The segment routing policy instance can be characterized by corresponding key values. The segment routing policy key values include the first network device node address (i.e., the node IP of the IGW or the computing and networking controller), the second network device node address (the EGW node IP), and color; where color is used to characterize the service attribute or service level requirement, helping the network device to dynamically select or construct a forwarding path that meets specific quality requirements according to the service type.

[0132] Integrate the segment routing policy key values and the corresponding network quality metrics to form a third entry (i.e., path quality information), providing network-side quality metric data support for computing power routing calculation. As Figure 11 shown, the third entry specifically includes: POLICY instance, the name of the segment routing policy instance, which is convenient for quickly distinguishing each instance during operation and maintenance (not necessary); IGW node IP, that is, the first network device node address; EGW node IP, that is, the second network device node address; Color attribute; network quality metrics, including available bandwidth, delay, and packet loss rate.

[0133] In one embodiment, the computing power information includes at least one of the following: service identifier, virtual private network service instance identifier, second network device node address, computing power instance address, and computing power instance metrics.

[0134] Exemplarily, the ingress computing power gateway or the computing and networking controller acts as a client, receives the computing power information of the computing power instance from the egress computing power gateway, and forms a second entry, providing computing-side metric data support for computing power routing calculation. As Figure 12 shown, specifically, the second entry includes: computing power instance IP address, service identifier, EGW node IP (i.e., the second network device node IP), VPN SID, computing power instance metrics (including CPU occupancy rate, memory occupancy rate, IO bandwidth, processing delay), and other information.

[0135] In one embodiment, the computing power information includes the second network device node address, the computing power instance address, and the computing power instance metrics; the path quality information includes the segment routing policy key values and the network quality metrics. Step 330 includes:

[0136] Determine the second network device node address according to the computing power information;

[0137] According to the first network device node address and the second network device node address, determine all segment routing policy key values and the corresponding network quality metrics from the path quality information;

[0138] Determine the target computing power instance and the target segment routing policy instance according to the preset computing and networking service level agreement constraint conditions, computing power instance metrics, and network quality metrics;

[0139] Determine the computing power routing result to the target computing power instance according to the computing power instance address corresponding to the target computing power instance and the target segmented routing policy instance.

[0140] In this embodiment, the target computing power instance refers to a computing power instance whose corresponding computing power instance metrics can meet the constraints of the computing network service level agreement. The target segmented routing policy instance refers to a segmented routing policy instance whose corresponding network quality metrics can meet the constraints of the computing network service level agreement.

[0141] After the ingress computing power gateway or the computing network controller collects the computing power information (the second entry) and the path quality information (the third entry), it determines the corresponding computing network service level agreement constraint conditions according to the service identifier, and combines the computing power information in the second entry and the path quality information in the third entry to calculate the computing power routing result. Specifically, for each service identifier, determine all the computing power instances corresponding to the service identifier and their corresponding EGW node IPs; then, according to the IGW node IP of the current first network device itself and the determined EGW node IPs, determine all the segmented routing policy instances (SR-policy) from IGW to EGW. According to the computing network SLA constraint conditions, select the end-to-end optimal computing power instance and SR-policy from the determined computing power instances and segmented routing policy instances to form the computing power routing result entry (the table shown in the instance reference Figure 13 ).

[0142] Exemplarily, taking the ingress computing power gateway as an example, the method for determining the computing power routing in the embodiment of the present application is described as a whole:

[0143] ① The ingress computing power gateway obtains computing power information: The ingress computing power gateway (IGW), as a telemetry client, receives the computing power information of the announced computing power instances from the egress computing power gateway (EGW) to provide basic data for subsequent computing power routing calculation and resource scheduling. The specific computing power information includes: service identifier, computing power instance IP, EGW node IP, VPN SID, computing power instance metrics, where the EGW node IP is extracted from the source address of the telemetry packet IP header, and other field information is obtained from the telemetry YANG model fields.

[0144] ② The ingress computing power gateway obtains path quality information: The IGW obtains the network quality metric information of all SR-policy network paths from the IGW to the EGW to form path quality information, which provides network-level performance data support for the calculation of computing power routing. The specific path quality information includes: POLICY instance name, IGW node IP, EGW node IP, Color, network quality metrics (available bandwidth, delay, packet loss, etc.).

[0145] ③Service identifier-oriented computing power routing calculation: For a service identifier, based on the computing network SLA requirements corresponding to the service identifier, that is, the computing network SLA constraint conditions, the computing power routing result is obtained by using the computing power information and path quality information obtained above. The specific method for determining the computing power routing is as follows:

[0146] The IGW is based on the Figure 12 shown computing power instances and related computing power information and the Figure 11 shown SR-policy and network quality index information, filters all computing power instance entries corresponding to the target service identifier, and finds all SR-policy entries from the IGW to the EGW node according to the IGW and EGW node IPs corresponding to each computing power instance entry.

[0147] In the previous step, the network quality index of the SR-policy entry is combined with the computing power instance index in the computing power information. According to the end-to-end computing network SLA constraint conditions corresponding to the target service identifier, the computing power instances and SR-policy that meet the requirements are filtered out to obtain the computing power routing result corresponding to the target service identifier.

[0148] The computing power routing result is saved to the IGW to guide subsequent data forwarding. The content of the computing power routing result may include: service identifier, POLICY instance, computing power instance IP, VPN SID.

[0149] It should be noted that when the first network device is a computing network controller or an architecture of an ingress computing power gateway + computing network controller, the corresponding specific process is the same as that of the above embodiments and will not be described separately.

[0150] Figure 14 This is a flowchart of the method for determining computing power routing provided in another embodiment of the present application. As Figure 14 shown, this method for determining computing power routing can be but is not limited to being applied to a second network device, such as an egress computing power gateway, or the Figure 2 shown egress computing power gateway 240. In this embodiment, this method for determining computing power routing at least includes but is not limited to:

[0151] Step 1410: Receive a telemetry subscription request for a target computing power instance sent by the first network device;

[0152] Step 1420: In response to the telemetry subscription request, send the computing power information of the target computing power instance to the first network device. The computing power information is used for the first network device to determine the computing power routing result to the target computing power instance according to the computing power information and path quality information, where the path quality information is the quality information of the segment routing policy instance between the first network device and the second network device.

[0153] In one embodiment, the method for determining computing power routing further includes:

[0154] Send a computing power information acquisition request to each computing power instance, where the computing power information acquisition request carries an indication of the collection information list;

[0155] Receive the computing power information responded by each computing power instance, where the computing power information is obtained by the computing power instance collecting the current index data according to the indication of the collection information list and encapsulating it.

[0156] In one embodiment, the computing power routing determination method further includes:

[0157] Receive the computing power information sent by each computing power instance according to the static subscription configuration;

[0158] Among them, the static subscription configuration is used to instruct the computing power instance to push the computing power information to the second network device.

[0159] In one embodiment, the computing power routing determination method further includes:

[0160] Encapsulate the address information and status information of the target computing power instance according to the telemetry data structure model to obtain the computing power information.

[0161] In one embodiment, the telemetry subscription request also carries an event mode; step 1420 includes:

[0162] After receiving the telemetry subscription request, determine the event condition according to the event mode;

[0163] In response to meeting the event condition, send the computing power information of the target computing power instance to the first network device.

[0164] In one embodiment, the telemetry subscription request also carries a time interval mode; step 1420 includes:

[0165] After receiving the telemetry subscription request, determine the interval time according to the time interval mode;

[0166] Send the computing power information of the target computing power instance to the first network device according to the interval time.

[0167] It should be noted that Figure 14 The provided computing power routing determination method and Figure 3 The provided computing power routing determination method correspond to each other on the second network device side (export computing power gateway) and the first network device side (import computing power gateway / network computing controller). Therefore Figure 14 The further specific implementation details and beneficial effects of the provided computing power routing determination method correspond to the above embodiments and will not be elaborated here.

[0168] Figure 15 It is a schematic structural diagram of a communication device provided by an embodiment of the present application. As Figure 15As shown, the communication device 2000 includes a memory 2100 and a processor 2200. The number of the memory 2100 and the processor 2200 can be one or more. Figure 15 In Figure 15 , a memory 2101 and a processor 2201 are taken as an example; the memory 2101 and the processor 2201 in the network device can be connected through a bus or other means. Figure 15 In Figure 15 , taking the connection through a bus as an example.

[0169] The memory 2101, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as program instructions / modules corresponding to the methods provided in any embodiment of the present application. The processor 2201 realizes the computing power routing determination method provided in any of the above embodiments by running the software programs, instructions, and modules stored in the memory 2101.

[0170] The memory 2101 mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function. In addition, the memory 2101 can include high-speed random access memory, and can also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some instances, the memory 2101 further includes a memory remotely set relative to the processor 2201, and these remote memories can be connected to the device through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0171] An embodiment of the present application further provides a computer-readable storage medium storing computer-executable instructions for executing the computing power routing determination method provided in any embodiment of the present application.

[0172] An embodiment of the present application further provides a computer program product including a computer program or computer instructions. The computer program or computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer program or computer instructions from the computer-readable storage medium, and the processor executes the computer program or computer instructions, so that the computer device executes the computing power routing determination method provided in any embodiment of the present application.

[0173] The system architecture and application scenarios described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that with the evolution of the system architecture and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

[0174] Those of ordinary skill in the art will understand that all or some of the steps in the methods disclosed above, and the functional modules / units in systems and devices, can be implemented as software, firmware, hardware, or a suitable combination thereof.

[0175] In a hardware implementation, the division of functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a single physical component may have multiple functions, or a function or step may be performed by several physical components working together. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transitory medium). As is well known to those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that a communication medium typically contains computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium.

[0176] The terms "component", "module", "system", etc. used in this specification are used to denote computer-related entities, hardware, firmware, combinations of hardware and software, software, or software in execution. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, an executable file, an execution thread, a program, or a computer. By way of illustration, both an application running on a computing device and the computing device can be components. One or more components may reside within a process or execution thread, and a component may be located on one computer or distributed between two or more computers. In addition, these components may execute from various computer-readable media storing various data structures. A component may communicate, for example, by signals according to one or more data packets (such as data from two components interacting with each other from a local system, a distributed system, or a network, such as via the Internet interacting with other systems).

[0177] Some embodiments of the present application have been described above with reference to the accompanying drawings, which does not limit the scope of the rights of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the present application shall fall within the scope of the rights of the present application.

Claims

1. A computing power routing determination method, characterized in that Applied to a first network device, the method includes: Sending a telemetry subscription request for a target computing power instance to a second network device; Receiving the computing power information of the target computing power instance sent by the second network device in response to the telemetry subscription request; Determining a computing power routing result to the target computing power instance according to the computing power information and path quality information, where the path quality information is the quality information of a segment routing policy instance between the first network device and the second network device.

2. The computing power routing determination method according to claim 1, wherein The computing power information is encapsulated by the second network device according to a telemetry data structure model for the address information and status information of the target computing power instance.

3. The computing power routing determination method according to claim 1, wherein The telemetry subscription request also carries an event pattern; The receiving the computing power information of the target computing power instance sent by the second network device in response to the telemetry subscription request includes: Receiving the computing power information of the target computing power instance sent by the second network device in response to meeting an event condition after receiving the telemetry subscription request, where the event condition is determined by the second network device according to the event pattern.

4. The computing power routing determination method according to claim 1, wherein The telemetry subscription request also carries a time interval pattern; The receiving the computing power information of the target computing power instance sent by the second network device in response to the telemetry subscription request includes: Receiving the computing power information of the target computing power instance sent by the second network device at an interval time corresponding to the time interval pattern after receiving the telemetry subscription request.

5. The computing power routing determination method according to claim 1, wherein Before determining the computing power routing result to the target computing power instance according to the computing power information and path quality information, the method further includes: Determining all the segment routing policy instances between the first network device and the second network device; Obtaining the network quality metrics corresponding to each segment routing policy instance; Integrating the path quality information according to each segment routing policy instance and the corresponding network quality metrics.

6. The computing power routing determination method according to claim 1, wherein The computing power information includes the second network device node address, the computing power instance address, and the computing power instance metrics; The path quality information includes the segment routing policy key value and the network quality metrics; The determining the computing power routing result to the target computing power instance according to the computing power information and path quality information includes: Determining the second network device node address according to the computing power information; Determining the segment routing policy key value and the corresponding network quality metrics from the path quality information according to the first network device node address and the second network device node address; Determining a target computing power instance and a target segment routing policy instance according to a preset computing network service level agreement constraint condition, the computing power instance metrics, and the network quality metrics; Determining the computing power routing result to the target computing power instance according to the computing power instance address corresponding to the target computing power instance and the target segment routing policy instance.

7. A computing power routing determination method, characterized in that Applied to a second network device, the method includes: Receiving a telemetry subscription request for a target computing power instance sent by a first network device; In response to the telemetry subscription request, send the computing power information of the target computing power instance to the first network device, where the computing power information is used by the first network device to determine a computing power routing result to the target computing power instance according to the computing power information and path quality information, and the path quality information is the quality information of a segment routing policy instance between the first network device and the second network device.

8. The computing power routing determination method according to claim 7, wherein The method further includes: Send a computing power information acquisition request to each computing power instance, where the computing power information acquisition request carries an indication of a collection information list; Receive the computing power information responded by each of the computing power instances, where the computing power information is obtained by the computing power instance collecting current metric data according to the indication of the collection information list and encapsulating it.

9. The computing power routing determination method according to claim 7, wherein The method further includes: Receive the computing power information sent by each of the computing power instances according to a static subscription configuration; where the static subscription configuration is used to instruct the computing power instance to push the computing power information to the second network device.

10. The computing power routing determination method according to claim 7, characterized in that, The method further includes: Encapsulate the address information and status information of the target computing power instance according to a telemetry data structure model to obtain the computing power information.

11. The computing power routing determination method according to claim 7, wherein The telemetry subscription request further carries an event pattern; The step of, in response to the telemetry subscription request, sending the computing power information of the target computing power instance to the first network device includes: After receiving the telemetry subscription request, determine an event condition according to the event pattern; In response to the event condition being satisfied, send the computing power information of the target computing power instance to the first network device.

12. The computing power routing determination method according to claim 7, wherein The telemetry subscription request further carries a time interval pattern; The step of, in response to the telemetry subscription request, sending the computing power information of the target computing power instance to the first network device includes: After receiving the telemetry subscription request, determine an interval time according to the time interval pattern; Send the computing power information of the target computing power instance to the first network device at the interval time.

13. A communication device, characterized in that, Includes: At least one processor; At least one memory for storing at least one program; When at least one of the programs is executed by at least one of the processors, the method according to any one of claims 1 to 12 is implemented.

14. A computer program product, comprising a computer program or computer instructions, characterized in that, The computer program or the computer instruction is stored in a computer-readable storage medium, and the processor of the computer device reads the computer program or the computer instruction from the computer-readable storage medium, and the processor executes the computer program or the computer instruction, so that the computer device executes the method according to any one of claims 1 to 12.

Images